Difference between revisions of "Timeline of ChatGPT"

From Timelines
Jump to: navigation, search
 
(307 intermediate revisions by 2 users not shown)
Line 1: Line 1:
This is a '''timeline of {{w|ChatGPT}}''', an AI-based chatbot developed by {{w|OpenAI}}. ChatGPT falls into the category of generative AI, a type of artificial intelligence that can be used to create text or images. It works by utilizing a large language model, which enables it to understand and produce dialogue using significant amounts of data.<ref name="Kharpal">{{cite web |last1=Kharpal |first1=Arjun |title=Chinese tech giant Alibaba working on a ChatGPT rival; shares jump |url=https://www.cnbc.com/2023/02/08/chinese-tech-giant-alibaba-working-on-a-chatgpt-rival-shares-jump.html |website=CNBC |access-date=4 March 2023 |language=en}}</ref> Since its launch in November 2022, as of March 2023 ChatGPT has become the fastest-growing consumer app in internet history, reaching 100 million users within two months.<ref name="Bhaimiya"/> The tool contains 570 gigabytes of text data. Training the model took only 34 days, and the tool costs around $100,000 per day to run on Microsoft’s Azure Cloud. Despite being generated by an AI, an average of 53% of people can't tell that ChatGPT content is not human-written.<ref>{{cite web |url=https://www.tooltester.com/en/blog/chatgpt-statistics/#:~:text=ChatGPT%20web%20traffic%20statistics,-ChatGPT%20received%20an&text=The%20site's%20traffic%20has%20grown,62.37%25%20from%20January%20to%20February. |access-date=25 March 2023}}</ref>
+
{{focused coverage period|end-date = September 2023}}
 +
This is a '''timeline of {{w|ChatGPT}}''', an AI-based chatbot developed by {{w|OpenAI}}. ChatGPT falls into the category of generative AI, a type of artificial intelligence that can be used to create text or images. It works by utilizing a large language model, which enables it to understand and produce dialogue using significant amounts of data.<ref name="Kharpal">{{cite web |last1=Kharpal |first1=Arjun |title=Chinese tech giant Alibaba working on a ChatGPT rival; shares jump |url=https://www.cnbc.com/2023/02/08/chinese-tech-giant-alibaba-working-on-a-chatgpt-rival-shares-jump.html |website=CNBC |access-date=4 March 2023 |language=en}}</ref> Since its launch in November 2022, as of March 2023 ChatGPT has become the fastest-growing consumer app in internet history, reaching 100 million users within two months.<ref name="Bhaimiya"/>  
  
 
== Sample questions ==
 
== Sample questions ==
  
The following are some interesting questions that can be answered by reading this timeline:  
+
The following are some interesting questions that can be answered by reading this timeline:
 +
 
 +
* What are some important events preluding the release of ChatGPT?
 +
** Sort the full timeline by "Event type" and look for the group of rows with value "Prelude".
 +
** You will see some significant events preluding the release of ChatGPT.
 +
* When was ChatGPT launched? What are other related services?
 +
** Sort the full timeline by "Event type" and look for the group of rows with value "Product launch".
 +
** You will see a number of events describing the launch and expansion of ChatGPT into updated versions. 
 +
** describes the launch and expansion of ChatGPT
 +
* What are some events describing the performance of ChatGPT?
 +
** Sort the full timeline by "Event type" and look for the group of rows with value "Performance".
 +
** You will see a number of studies assessing ChatGPT on different skills.
 +
* What are some notable cases of adoption?
 +
** Sort the full timeline by "Event type" and look for the group of rows with value "Adoption".
 +
** You will see examples showing how ChatGPT is being used in a variety of ways. You will also see a number of studies analizing the adoption of ChatGPT in some fields.
 +
* What are some reviews on the behavior of ChatGPT?
 +
** Sort the full timeline by "Event type" and look for the group of rows with value "System behavior".
 +
** You will see a number of reviews discussing topics such as outages, data exposure, hallucination, and also ChatGPT's political Ideology.
 +
* What are some events discussing actual or potential impact of ChatGPT and similar products on different aspects of society?
 +
** Sort the full timeline by "Event type" and look for the group of rows with value "Impact".
 +
** You will see a number of areas of impact and discussed issues, such as potential ChatGPT political engagement, warning against humanizing chatbots, the transformative role in education, impact on job roles, ocupations and industries, and other ethical concerns.
 +
* What are some examples of expansion of ChatGPT into other territories and languages?
 +
** Sort the full timeline by "Event type" and look for the group of rows with value "Expansion".
 +
* What are some cases of competition from other companies developing similar products?
 +
** Sort the full timeline by "Event type" and look for the group of rows with value "Competition".
 +
** You will see some comparisons between ChatGPT and other services, as well as prospective launches of large language model-based chatbots by other companies.
 +
* What are some cases of ban or restriction of use of ChatGPT by official entities?
 +
** Sort the full timeline by "Event type" and look for the group of rows with value "Restriction".
 +
** You will see some notable restrictions, like the temporary ban in Italy, as well as some cautious company bans.
 +
* Other events are described under the following types: "Criticism", "Extension", "Growth", "Maintenance", "Privacy", "Reception", "Recognition", and "Revenue".
  
 
==Big picture==
 
==Big picture==
  
 
{| class="wikitable"
 
{| class="wikitable"
! Time period !! Development summary !! More details
+
! Time period (approximately) !! Development summary !! More details
 +
|-
 +
| December 2022 || Successful launch || After being launched on November 30, by December ChatGPT already becomes widely used and dominates chat rooms as the next innovative AI language technology.<ref>{{cite web |last1=DiBenedetto |first1=Chase |title=ChatGPT's surprisingly human voice came with a human cost |url=https://mashable.com/article/chat-gpt-open-ai-workers-exploitation |website=Mashable |access-date=7 March 2023 |language=en |date=18 January 2023}}</ref>
 +
|-
 +
| January 2023 || Record adoption || ChatGPT reaches 100 million monthly users, a faster rate of adoption than {{w|Instagram}} or {{w|TikTok}}.<ref>{{cite web |title=The AI Arms Race Is On. Start Worrying |url=https://time.com/6255952/ai-impact-chatgpt-microsoft-google/ |website=Time |access-date=6 March 2023 |language=en |date=16 February 2023}}</ref> The chatbot makes headlines for its ability to pass prestigious graduate-level exams.<ref>{{cite web |last1=Kelly |first1=Samantha Murphy |title=5 jaw-dropping things GPT-4 can do that ChatGPT couldn't {{!}} CNN Business |url=https://edition.cnn.com/2023/03/16/tech/gpt-4-use-cases/index.html |website=CNN |access-date=23 March 2023 |language=en |date=16 March 2023}}</ref>
 +
|-
 +
| February 2023 || Valuation || ChatGPT receives an estimated 1 billion visitors, an average of over 35 million visitors per day.<ref name="tooltester">{{cite web |title=ChatGPT Statistics 2023 |url=https://www.tooltester.com/en/blog/chatgpt-statistics/ |website=tooltester |access-date=22 March 2023}}</ref> The valuation of ChatGPT surpasses $29 billion.<ref name="Rogers">{{cite web |last1=Rogers |first1=Paul |title=56 ChatGPT Statistics For 2023 (Users, Facts & Trends) |url=https://www.answeriq.com/chatgpt-statistics/ |website=Answer iQ |access-date=17 June 2023 |date=8 March 2023}}</ref>
 +
|-
 +
| March 2023 || ChatGPT Plus launch|| ChatGPT Plus is released as a subscription-based service, providing faster responses and priority access to new features compared to the free version of ChatGPT.<ref>{{cite web |title=OpenAI adds new data controls to ChatGPT and previews business version |url=https://siliconangle.com/2023/04/25/openai-adds-new-data-controls-chatgpt-previews-business-version/ |website=SiliconANGLE |access-date=7 May 2023 |date=25 April 2023}}</ref> ChatGPT experiencea a data breach when a bug exposea users' personal information.<ref>{{cite web |title=From the headlines: ChatGPT and other AI text-generating risks |url=https://www.scmagazine.com/resource/emerging-technology/from-the-headlines-chatgpt-and-other-ai-text-generating-risks |website=SC Media |access-date=17 June 2023 |language=en |date=4 May 2023}}</ref><ref>{{cite web |last1=Team |first1=Editorial |title=Why ChatGPT is a Cyber Threat to Businesses |url=https://insidebigdata.com/2023/05/02/why-chatgpt-is-a-cyber-threat-to-businesses/ |website=insideBIGDATA |access-date=17 June 2023 |date=2 May 2023}}</ref>   
 +
|-
 +
| April 2023 || First notable ban || ChatGPT attracts about 1.76 billion visits worldwide, up 12.6% from March.<ref>{{cite web|title=ChatGPT’s Growth Begins to Flatten, up 12.6% from March to April|url=https://www.similarweb.com/blog/insights/ai-news/chatgpt-growth-flattens/}}</ref> Italy bans on the chatbot over worries about its potential adverse effects on the community.<ref>{{cite web |last1=Saltzman |first1=Marc |last2=Apr. 21 |first2=readersdigest caUpdated: |title=What is ChatGPT, Anyway? |url=https://www.readersdigest.ca/home-garden/tips/what-is-chatgpt/ |website=Reader's Digest Canada |access-date=7 May 2023 |date=21 April 2023}}</ref> By this time, users, privacy experts, and governments have raised criticisms regarding the data retention policies of ChatGPT.<ref name=privacy/>
 
|-
 
|-
| December 2022 || || After being launched on November 30, by December ChatGPT already becomes widely used and dominates chat rooms as the next innovative AI language technology.<ref>{{cite web |last1=DiBenedetto |first1=Chase |title=ChatGPT's surprisingly human voice came with a human cost |url=https://mashable.com/article/chat-gpt-open-ai-workers-exploitation |website=Mashable |access-date=7 March 2023 |language=en |date=18 January 2023}}</ref>
+
| May 2023 || Platform expansion || OpenAI launches an iOS app for ChatGPT.
 
|-
 
|-
| January 2023 || || ChatGPT reaches 100 million monthly users, a faster rate of adoption than {{w|Instagram}} or {{w|TikTok}}.<ref>{{cite web |title=The AI Arms Race Is On. Start Worrying |url=https://time.com/6255952/ai-impact-chatgpt-microsoft-google/ |website=Time |access-date=6 March 2023 |language=en |date=16 February 2023}}</ref> The chatbot makes headlines for its ability to pass prestigious graduate-level exams.<ref>{{cite web |last1=Kelly |first1=Samantha Murphy |title=5 jaw-dropping things GPT-4 can do that ChatGPT couldn't {{!}} CNN Business |url=https://edition.cnn.com/2023/03/16/tech/gpt-4-use-cases/index.html |website=CNN |access-date=23 March 2023 |language=en |date=16 March 2023}}</ref>
+
| June 2023 || User decline || ChatGPT sees its first-ever user decline.<ref>{{cite web |title=ChatGPT saw its first-ever user decline in June |url=https://www.engadget.com/chatgpt-saw-its-first-ever-user-decline-in-june-163043304.html?guccounter=1&guce_referrer=aHR0cHM6Ly93d3cuZ29vZ2xlLmNvbS8&guce_referrer_sig=AQAAAA_deDjNh9IokDYAoVXIGkvm_OYf8uRbupS85EgTY234vlgP2_HrWlx8LRvFhI_uqWUF1u85DDGTjPMDa_Y4IsCcU8ASo_9uFUf9h25l4ZgYlNNTMX99TGAlWAcM0Gh-egCfel6imrFpf20e3j1Z2J9Uu83Fm9hV2Y4-t7myJfFf |website=Engadget |access-date=25 July 2023}}</ref>
 
|-
 
|-
| February 2023 || || ChatGPT receives an estimated 1 billion visitors, an average of over 35 million visitors per day.<ref name="tooltester">{{cite web |title=ChatGPT Statistics 2023 |url=https://www.tooltester.com/en/blog/chatgpt-statistics/ |website=tooltester |access-date=22 March 2023}}</ref> ChatGPT plus is released.  
+
| July 2023 || Platform expansion || ChatGPT app launches for [[w:Android (operating system)|Android]].
 
|-
 
|-
| March 2023 || || {{w|GPT-4}} is released.
+
| August 2023 || Academic adoption || ChatGPT witnesses a modest surge in traffic within the United States, which aligns with the commencement of the academic year.
 
|-
 
|-
 
|}
 
|}
Line 23: Line 63:
  
 
{| class="sortable wikitable"
 
{| class="sortable wikitable"
! Year !! Month and date !! Event type !! Details
+
! Year !! Month and date !! Caption !! Event type !! Details
 +
|-
 +
| 2001 || || SmarterChild: pioneering AI chatbot || Prelude || Early chatbot SmarterChild gains popularity among preteens for its witty responses on AIM and MSN. It is a precursor to modern AI bots like {{w|Siri}} and [[w:Amazon Alexa|Alexa]], even inspiring investors to fund Siri. SmarterChild's parent company, ActiveBuddy, sees its potential and allows other businesses to create AI-driven chatbots. Unlike future modern AI bots like ChatGPT, SmarterChild has a personality and engages in humorous banter with users. The founders of SmarterChild would follow developments in AI over the years, with differing views on its future. SmarterChild is considered by some to be a milestone in AI chatbot evolution.<ref>{{cite web |last1=Silberling |first1=Amanda |title=Twenty years ago, AIM chatbot SmarterChild out-snarked ChatGPT |url=https://techcrunch.com/2023/07/25/twenty-years-ago-aim-chatbot-smarterchild-out-snarked-chatgpt/ |website=TechCrunch |access-date=25 July 2023 |date=25 July 2023}}</ref>
 +
|-
 +
| 2015 || December 11 || OpenAI, non-profit advancing human-friendly AI || Prelude || OpenAI is introduced as a {{w|non-profit organization}} focused on advancing digital intelligence in a way that benefits humanity as a whole. Its goal is to create {{w|artificial intelligence}} that is an extension of human will and distributed broadly and evenly. OpenAI's researchers are encouraged to publish their work and collaborate with others across many institutions. The organization receives funding from several donors, including {{w|Sam Altman}}, {{w|Elon Musk}}, and {{w|Amazon Web Services}}. Its leaders include {{w|Ilya Sutskever}} and {{w|Greg Brockman}}, who are experts in {{w|machine learning}}.<ref>{{cite web |title=Introducing OpenAI |url=https://openai.com/blog/introducing-openai |website=openai.com |access-date=15 March 2023}}</ref>
 +
|-
 +
| 2016 || April 27 || Introducing OpenAI Gym, RL algorithms & environments || Prelude || OpenAI Gym is introduced as a toolkit for developing and comparing {{w|reinforcement learning}} (RL) {{w|algorithm}}s that consist of a suite of environments from simulated robots to Atari games, and a site for comparing and reproducing results. The environments are written in Python and are compatible with algorithms written in any framework, such as {{w|Tensorflow}} and Theano. The platform aims to fix the lack of standardization of environments used in publications and the need for better benchmarks by providing a diverse suite of environments that range from easy to difficult and involve many different kinds of data. OpenAI Gym provides a site for comparing and reproducing results. They start out by maintaining a curated list of contributions that say something interesting about algorithmic capabilities.<ref>{{cite web |title=OpenAI Gym Beta |url=https://openai.com/research/openai-gym-beta |website=openai.com |access-date=15 March 2023}}</ref>
 +
|-
 +
| 2016 || December 5 || OpenAI Universe, training general intelligence in AI || Prelude || OpenAI releases Universe as a software platform designed to measure and train general intelligence in AI systems using a variety of games, websites, and applications. Universe allows an AI to operate a computer like a human does, using screen pixels and a virtual keyboard and mouse, allowing it to be trained on any task that a human can perform with a computer. The platform is capable of launching any program as a Gym environment and upon its launch includes over a thousand environments, including games like slither.io and GTA V, and browser tasks. Universe aims to develop a single AI agent that can apply past experience to quickly master new, difficult environments, moving towards {{w|general intelligence}}. Universe is an {{w|open-source}} platform that can be used for research or integrated with other games or applications.<ref>{{cite web |title=Universe |url=https://openai.com/research/universe |website=openai.com |access-date=15 March 2023}}</ref>
 +
|-
 +
| 2018 || June 11 || GPT-1 launch || Prelude || OpenAI unveils GPT-1, its initial version of GPT. This natural language processing model boasts 117 million parameters and is trained on a vast 4.5GB text dataset. What distinguishes GPT-1 from other NLP models at this time is its ability to undergo a substantial part of its training process without direct human oversight.<ref name="tooltester"/><ref name="www.skippet">{{cite web |last1= |title=GPT-4: another breakthrough in AI? |url=https://www.skippet.com/post/gpt-4-another-breathrough-ai-compared-to-other-models#:~:text=It%20was%20the%2011th%20of,GB%20of%20training%20data%20text. |website=www.skippet.com |access-date=1 September 2023}}</ref> 
 +
|-
 +
| 2019 || February 14 || || || The second version of GPT (Generative Pre-trained Transformer), known as {{w|GPT-2}}, makes its debut. The model, characterized by its remarkable size and proficiency in generating language, garners considerable interest from both the AI community and the media. With an impressive 1.5 billion parameters, it stays as the most extensive and powerful language model during its introduction, signaling a trend toward increasingly larger models in the years ahead.<ref name="techtimes.com">{{cite web |title=ChatGPT Timeline: Evolution and Rise of AI, Impact, Threat, and Opportunities |url=https://www.techtimes.com/articles/287927/20230228/chatgpt-timeline-evolution-rise-ai-impact-threat-opportunities.htm |website=techtimes.com |access-date=18 December 2023}}</ref>
 +
|-
 +
| 2019 || July 22 || || Partnership || The partnership between {{w|Microsoft}} and {{w|OpenAI}} is disclosed, marking the commencement of their collaborative efforts in advancing AI technology. In this association, Microsoft not only pledges a substantial $1 billion to OpenAI but also commits to joint efforts in developing innovative AI tools and services. Subsequently, Microsoft actively plays a key role in supporting OpenAI's research and development endeavors, particularly those linked to GPT (Generative Pre-trained Transformer) and ChatGPT.<ref>{{cite web |last1=Center |first1=Microsoft News |title=OpenAI forms exclusive computing partnership with Microsoft to build new Azure AI supercomputing technologies |url=https://news.microsoft.com/2019/07/22/openai-forms-exclusive-computing-partnership-with-microsoft-to-build-new-azure-ai-supercomputing-technologies/#:~:text=SAN%20FRANCISCO%20and%20REDMOND%2C%20Wash,in%20large%2Dscale%20AI%20systems. |website=Stories |access-date=18 December 2023 |date=22 July 2019}}</ref><ref>{{cite web |title=Microsoft invests in and partners with OpenAI to support us building beneficial AGI |url=https://openai.com/blog/microsoft-invests-in-and-partners-with-openai |website=openai.com |access-date=18 December 2023}}</ref><ref>{{cite web |last1=Vincent |first1=James |title=Microsoft invests $1 billion in OpenAI to pursue holy grail of artificial intelligence |url=https://www.theverge.com/2019/7/22/20703578/microsoft-openai-investment-partnership-1-billion-azure-artificial-general-intelligence-agi |website=The Verge |access-date=18 December 2023 |language=en |date=22 July 2019}}</ref><ref>{{cite web |title=Microsoft invests $1 billion in OpenAI, which is going all-in on Azure |url=https://www.zdnet.com/article/microsoft-invests-1-billion-in-openai-which-is-going-all-in-on-azure/#google_vignette |website=ZDNET |access-date=18 December 2023 |language=en}}</ref><ref>{{cite web |last1=Rowley |first1=Jason D. |title=Microsoft Commits $1 Billion To OpenAI Partnership To Build Its AI Systems On Azure |url=https://news.crunchbase.com/business/microsoft-commits-1-billion-to-openai-partnership-to-build-its-ai-systems-on-azure/ |website=Crunchbase News |access-date=18 December 2023 |date=22 July 2019}}</ref>
 +
|-
 +
| 2020 || June 11 || GPT-3 launch || Prelude || OpenAI introduces GPT-3, which has over a thousand times the number of parameters compared to its initial predecessor and the capability to process a vast 570GB of plain text.<ref name="www.skippet"/>
 +
|-
 +
| 2022 || November 30 || OpenAI launches GPT-3.5 featuring ChatGPT, an interactive conversational AI trained with RLHF and safety measures || Product launch || GPT-3.5 is released.<ref name="tooltester"/> ChatGPT is introduced as a conversational AI model that can interact with users and respond to follow-up questions, challenge incorrect premises, and reject inappropriate requests. It is trained using {{w|Reinforcement Learning}} from Human Feedback (RLHF) and fine-tuned from the GPT-3.5 model. While it has limitations such as occasionally providing incorrect or nonsensical answers and being sensitive to input phrasing, it is designed with safety mitigations and the ability to learn from user feedback. ChatGPT is available for free during the research preview on chat.openai.com.<ref>{{cite web |title=Introducing ChatGPT |url=https://openai.com/blog/chatgpt |website=openai.com |access-date=16 March 2023}}</ref>
 
|-
 
|-
| 2015 || December 11 || Prelude || OpenAI is introduced as a non-profit organization focused on advancing digital intelligence in a way that benefits humanity as a whole. Its goal is to create AI that is an extension of human will and distributed broadly and evenly. OpenAI's researchers are encouraged to publish their work and collaborate with others across many institutions. The organization has received funding from several donors, including {{w|Sam Altman}}, {{w|Elon Musk}}, and {{w|Amazon Web Services}}. Its leaders include {{w|Ilya Sutskever}} and {{w|Greg Brockman}}, who are experts in machine learning. OpenAI hopes to build a leading research institution that prioritizes a good outcome for all over its own self-interest.<ref>{{cite web |title=Introducing OpenAI |url=https://openai.com/blog/introducing-openai |website=openai.com |access-date=15 March 2023}}</ref>
+
| 2022 || December 5 || The ''New York Times'' comments on ChatGPT  || Recognition || Kevin Roose of ''The New York Times'' labels ChatGPT as "the best artificial intelligence chatbot ever released to the general public". ChatGPT demonstrates exceptional flexibility, wit, and problem-solving abilities, attracting over a million testers within days of its launch. While it can handle various tasks, from writing jokes and computer code to aiding programmers and analyzing school assignments, it is not without flaws and may occasionally provide incorrect responses. ''{{w|The New York Times}}'' also reports that it has been "rumored" that the next version of the AI, {{w|GPT-4}}, will be launched sometime in 2023.<ref name=":4">{{Cite web |last=Roose |first=Kevin |date=December 5, 2022 |title=The Brilliance and Weirdness of ChatGPT |url=https://www.nytimes.com/2022/12/05/technology/chatgpt-ai-twitter.html |access-date= December 26, 2022 |website=The New York Times |language=en-US |quote=Like those tools, ChatGPT — which stands for "generative pre-trained transformer" — landed with a splash. |archive-date=January 18, 2023 |archive-url=https://web.archive.org/web/20230118134332/https://www.nytimes.com/2022/12/05/technology/chatgpt-ai-twitter.html}}</ref>  
 
|-
 
|-
| 2016 || April 27 || Prelude || OpenAI Gym is introduced as a toolkit for developing and comparing reinforcement learning (RL) algorithms that consist of a suite of environments from simulated robots to Atari games, and a site for comparing and reproducing results. The environments are written in Python and are compatible with algorithms written in any framework, such as Tensorflow and Theano. The platform aims to fix the lack of standardization of environments used in publications and the need for better benchmarks by providing a diverse suite of environments that range from easy to difficult and involve many different kinds of data. OpenAI Gym provides a site for comparing and reproducing results. They are starting out by maintaining a curated list of contributions that say something interesting about algorithmic capabilities.<ref>{{cite web |title=OpenAI Gym Beta |url=https://openai.com/research/openai-gym-beta |website=openai.com |access-date=15 March 2023}}</ref>  
+
| 2022 || December 5 || || Restriction || Popular coding question-and-answer site {{w|Stack Overflow}} temporarily bans the sharing of AI-generated answers produced by ChatGPT. The site's moderators say the ban is necessary as ChatGPT's answers are often incorrect and this can mislead users. ChatGPT ability to generate plausible but ultimately false responses is one of several well-known failings of AI text generation models.<ref>{{cite web |last1=Vincent |first1=James |title=AI-generated answers temporarily banned on coding Q&A site Stack Overflow |url=https://www.theverge.com/2022/12/5/23493932/chatgpt-ai-generated-answers-temporarily-banned-stack-overflow-llms-dangers |website=The Verge |access-date=6 March 2023 |date=5 December 2022}}</ref>
 
|-
 
|-
| 2016 || December 5 || Prelude || OpenAI releases Universe as a software platform designed to measure and train general intelligence in AI systems using a variety of games, websites, and applications. Universe allows an AI to operate a computer like a human does, using screen pixels and a virtual keyboard and mouse, allowing it to be trained on any task that a human can perform with a computer. The platform is capable of launching any program as a Gym environment and upon its launch includes over a thousand environments, including games like slither.io and GTA V, and browser tasks. Universe aims to develop a single AI agent that can apply past experience to quickly master new, difficult environments, moving towards general intelligence. Universe is an open-source platform that can be used for research or integrated with other games or applications.<ref>{{cite web |title=Universe |url=https://openai.com/research/universe |website=openai.com |access-date=15 March 2023}}</ref>
+
| 2022 || December 27 || Article explores ChatGPT's political potential || Impact || A {{w|BBC}} article discusses the potential of using chatbots, particularly ChatGPT, for political purposes. It explores the chatbot's ability to engage in simulated conversations, including Prime Minister's Questions, and highlights its polite and non-inflammatory responses. The chatbot's performance in writing an MP's maiden speech is also praised, with an example of a real MP using the chatbot to create a speech. However, it notes that the chatbot is not allowed to express political opinions or engage in activism. The article concludes by discussing the theoretical possibility of a robot becoming a prime minister, contingent on advancements in artificial intelligence and decision-making capabilities.<ref>{{cite news |title=Could a chatbot answer Prime Minister's Questions? |url=https://www.bbc.com/news/uk-politics-64053550 |first=Kate |last=Whannel |access-date= December 30, 2022 |work=BBC News |date=December 27, 2022 |archive-date=January 17, 2023 |archive-url=https://web.archive.org/web/20230117060725/https://www.bbc.com/news/uk-politics-64053550 }}</ref>
 
|-
 
|-
| 2017 || August 11 || Prelude || OpenAI announces having created a bot that learned to play Dota 2 through self-play and is capable of beating top professionals in 1v1 matches. The bot uses reinforcement learning and does not rely on imitation learning or tree search. The game is complex, with hidden information, requiring players to plan, attack, trick, and deceive their opponents. The AI's actions-per-minute are comparable to that of an average human player, but it has learned to predict opponents' movements, improvise in unfamiliar situations, and influence other players' units to help it succeed. OpenAI's next goal is to create a team of Dota 2 bots that can compete and collaborate with the top human teams.<ref>{{cite web |title=Dota 2 |url=https://openai.com/research/dota-2 |website=openai.com |access-date=15 March 2023}}</ref>
+
| 2023 || January 5 || Paper analyzes ChatGPT's political ideology || System behavior || A paper examines the political ideology of ChatGPT, and finds that it has a pro-environmental, left-libertarian orientation based on its responses to 630 political statements from voting advice applications and the political compass test. ChatGPT would impose taxes on flights, restrict rent increases, and legalize abortion. The findings are consistent across multiple languages and different experiment variations. The study highlights the potential impact of politically biased conversational AI on society.<ref>{{cite journal |last1=Hartmann |first1=Jochen |last2=Schwenzow |first2=Jasper |last3=Witte |first3=Maximilian |title=The political ideology of conversational AI: Converging evidence on ChatGPT's pro-environmental, left-libertarian orientation |date=2023 |doi=10.48550/arXiv.2301.01768 |url=https://arxiv.org/abs/2301.01768}}</ref>
 
|-
 
|-
| 2018 || June 11 || || GPT-1 is released.<ref name="tooltester"/>
+
| 2023 || January 27 || {{w|Elon Musk}} plans alternative to ChatGPT || Competition || It is reported that South African business magnate {{w|Elon Musk}} plans to form a new AI research lab to develop an alternative to OpenAI's ChatGPT, after having approached AI researchers, including Igor Babuschkin. At this time, the project is still in its early stages, with no concrete plan to develop specific products.<ref>{{cite web |title=Elon Musk recruits team to develop OpenAI’s ChatGPT rival: Reports |url=https://english.alarabiya.net/News/world/2023/02/28/Elon-Musk-recruits-team-to-develop-OpenAI-s-ChatGPT-rival-Reports |website=Al Arabiya English |access-date=4 March 2023 |language=en |date=28 February 2023}}</ref>
 
|-
 
|-
| 2019 || February 14 || || OpenAI publishes a blog post discussing the release of GPT-2, a large-scale unsupervised language model with 1.5 billion parameters, which can generate coherent paragraphs of text, achieve state-of-the-art performance on many language modeling benchmarks, and perform rudimentary reading comprehension, machine translation, question answering, and summarization without task-specific training. Due to concerns about malicious applications, the trained model is not released, but a smaller model and a technical paper are released for research purposes. GPT-2 is trained to predict the next word in 40GB of internet text, using a diverse dataset, and can generate conditional synthetic text samples of unprecedented quality.<ref>{{cite web |title=Better language models and their implications |url=https://openai.com/research/better-language-models |website=openai.com |access-date=23 March 2023}}</ref>
+
| 2023 || January 27 || Article outlines six critical challenges for ChatGPT || Impact || An article highlights the six major challenges facing ChatGPT at the time. These include capacity issues caused by overwhelming traffic, instances of plagiarism and cheating among students who utilize the chatbot, concerns regarding racism, sexism, and bias in its responses, accuracy problems resulting in incorrect information, the controversial training methods involving disturbing content and exploitation of low-cost workers, and the absence of a mobile app, leading to scams and unauthorized copycat applications. According to the article, these issues raise important ethical considerations and question the long-term viability of the chatbot in its current form.<ref>{{cite web |title=The 6 biggest problems with ChatGPT right now |url=https://www.digitaltrends.com/computing/the-6-biggest-problems-with-chatgpt-right-now/ |website=Digital Trends |access-date=24 May 2023 |language=en |date=27 January 2023}}</ref>
 
|-
 
|-
| 2019 || May 7 || || Adam King on {{w|Twitter}} launches “TalktoTransformer.com”, a site where people can have an interface to play with OpenAI's new text-generating language model.<ref>{{cite web |title=https://twitter.com/adamdanielking/status/1125831730848571392?lang=en |url=https://twitter.com/adamdanielking/status/1125831730848571392?lang=en |website=Twitter |access-date=23 March 2023 |language=en}}</ref>
+
| 2023 || January 30 || Article examines ChatGPT and Google Translate's translation abilities || Competition || An article compares the translation capabilities of ChatGPT and Google Translate. It explains that a good translation goes beyond swapping words and should convey the meaning of the entire text, preserve tone and context, and handle challenges like language differences and colloquialisms. The article tests both tools with colloquialisms and finds that while Google Translate often provides literal translations, ChatGPT offers additional assistance by providing interpretations and meanings.
 
|-
 
|-
| 2019 || May 9 || || French company {{w|Hugging Face}} releases a conversational AI demo based on GPT-2 models. The company believes that open-source and knowledge sharing should be the default, and that the progress of deep learning for NLP is at risk without open-source. While they are firm believers in open-source, they also understand that technology is not neutral and specific action must be taken to ensure a positive impact. The company analyzed the potential malicious use of the technology, and concluded that it could be used for drastic improvement of spam-bots or mass catfishing and identity fraud. To mitigate these risks, Hugging Face provides users with some simple tricks to recognize the conversational AI from a human conversation.<ref>{{cite web |last1=Delangue |first1=Clément |title=Ethical analysis of the open sourcing of a state-of-the-art conversational AI |url=https://medium.com/huggingface/ethical-analysis-of-the-open-sourcing-of-a-state-of-the-art-conversational-ai-852113c324b2 |website=HuggingFace |access-date=23 March 2023 |language=en |date=9 May 2019}}</ref>
+
| 2023 || January 31 || AI Text Classifier to detect assignment cheating with ChatGPT || Product launch || OpenAI launches a new tool, AI Text Classifier, to address concerns about cheating with ChatGPT. The tool aims to help teachers detect if a student or AI generated a particular assignment. While OpenAI acknowledges the tool's limitations, some school districts consider using ChatGPT as a teaching tool by this time. OpenAI also highlights the tool's potential in detecting disinformation campaigns and misuse of AI. In the meantime, higher education institutions debate responsible AI use, with some implementing guidelines to address ethical concerns.<ref>{{cite web |title=Faced with criticism it's a haven for cheaters, ChatGPT adds tool to catch them |url=https://www.cbc.ca/news/business/chatgpt-academic-cheating-1.6732115 |website=cbc.ca |access-date=24 May 2023}}</ref>
 
|-
 
|-
| 2019 || August 20 || || OpenAI announces the release of its 774 million parameter {{w|GPT-2}} language model, along with an open-source legal agreement to make it easier for organizations to initiate model-sharing partnerships with each other. They also publish a technical report about their experience in coordinating with the wider AI research community on publication norms. Through their research, they found that coordination for language models is difficult but possible, synthetic text generated by language models can be convincing to humans, and detecting malicious use of language models is a genuinely difficult research problem that requires both statistical detection and human judgment.<ref name="GPT-2: 6-month follow-up">{{cite web |title=GPT-2: 6-month follow-up |url=https://openai.com/research/gpt-2-6-month-follow-up |website=openai.com |access-date=23 March 2023}}</ref>
+
| 2023 || January 31 || || Growth || ChatGPT experiences its highest level of traffic with 28 million visitors in a single day.<ref name="Rogers"/>  
 
|-
 
|-
| 2019 || September 17 || Prelude || OpenAI publishes an article describing a new simulation environment that allows agents to learn and improve their ability to play hide-and-seek, ultimately leading to the emergence of complex tool use strategies. In the simulation, agents can move, see, sense, grab, and lock objects in place. There are no explicit incentives for the agents to interact with objects other than the hide-and-seek objective. Agents are rewarded based on the outcome of the game. As agents train against each other in hide-and-seek, up to six distinct strategies emerge, leading to increasingly complex tool use. The self-supervised emergent complexity in this simple environment further suggests that multi-agent co-adaptation may one day produce extremely complex and intelligent behavior.<ref>{{cite web |title=Emergent tool use from multi-agent interaction |url=https://openai.com/research/emergent-tool-use |website=openai.com |access-date=15 March 2023}}</ref>
+
| 2023 || February 1 || ChatGPT Plus subscription plan is launched || Product launch || OpenAI launches a pilot subscription plan called ChatGPT Plus, which offers faster response times, priority access to new features and improvements, and general access to ChatGPT even during peak times. The subscription costs $20/month and is available to customers worldwide. OpenAI still offers free access to ChatGPT but hopes to support its free access availability to as many people as possible through the subscription plan. OpenAI announces plans to refine and expand the offering based on user feedback and needs, and that it is actively exploring options for lower-cost plans, business plans, and data packs for more availability.<ref>{{cite web |title=Introducing ChatGPT Plus |url=https://openai.com/blog/chatgpt-plus |website=openai.com |access-date=16 March 2023}}</ref>
 
|-
 
|-
| 2019 || October 15 || Prelude || OpenAI reports on having trained a pair of neural networks that can solve the {{w|Rubik's Cube}} with a human-like robotic hand. The neural networks were trained entirely in simulation, using the same reinforcement learning code as OpenAI Five paired with a new technique called Automatic Domain Randomization (ADR). ADR creates diverse environments in simulation that can capture the physics of the real world, enabling the transfer of neural networks learned in simulation to be applied to the real world. The system can handle situations it never saw during training, such as being prodded by a stuffed giraffe. The breakthrough shows that reinforcement learning isn’t just a tool for virtual tasks, but can solve physical-world problems requiring unprecedented dexterity.<ref>{{cite web |title=Solving Rubik’s Cube with a robot hand |url=https://openai.com/research/solving-rubiks-cube |website=openai.com |access-date=15 March 2023}}</ref>
+
| 2023 || February 1 || Opinions on ChatGPT-3 among academics and netizens || Reception || A study explores the opinions and perceptions of academicians, {{w|netizen}}s, and information professionals towards ChatGPT-3, using {{w|social media}} comments and a survey. The research suggests that while ChatGPT-3 can be useful for research and writing, it is not entirely accurate and should be cross-checked. It also suggests that most academicians are beginning to accept ChatGPT-3, despite some resistance. The study is useful for academics, content developers, and librarians.<ref>{{cite journal |last1=A |first1=Subaveerapandiyan |last2=A |first2=Vinoth |last3=Tiwary |first3=Neelam |title=Netizens, Academicians, and Information Professionals' Opinions About AI With Special Reference To ChatGPT |journal=arXiv:2302.07136 [cs] |date=1 February 2023 |doi=10.48550/arXiv.2302.07136 |url=https://arxiv.org/abs/2302.07136}}</ref>
 
|-
 
|-
| 2019 || November 5 || Prelude || OpenAI releases the largest version of GPT-2, the 1.5B parameter version, along with code and model weights to aid detection of outputs of GPT-2 models. OpenAI releases the model as a test case for a full staged release process for future powerful models, hoping to continue the conversation with the AI community on responsible publication. OpenAI conducted some tests and research on the GPT-2 model and found that humans find GPT-2 outputs convincing, it can be fine-tuned for misuse, detection of synthetic text is challenging, they have not found evidence of misuse so far, and standards are needed for studying bias.<ref>{{cite web |title=GPT-2: 1.5B release |url=https://openai.com/research/gpt-2-1-5b-release |website=openai.com |access-date=15 March 2023}}</ref>
+
| 2023 || February 2 || ChatGPT achieves unprecedented growth || Growth || It is reported that ChatGPT reached an estimated 100 million monthly active users in January, just two months after its launch, making it the fastest-growing consumer application in history, according to a UBS study. The growing usage, while imposing substantial computing cost on OpenAI, also provides valuable feedback to help train the chatbot's responses.<ref>{{cite web |last1=Hu |first1=Krystal |title=ChatGPT sets record for fastest-growing user base - analyst note |url=https://www.reuters.com/technology/chatgpt-sets-record-fastest-growing-user-base-analyst-note-2023-02-01/ |website=Reuters |access-date=14 March 2023 |language=en |date=2 February 2023}}</ref>
 
|-
 
|-
| 2020 || May 28 || || GPT-3 is released in Beta.<ref name="tooltester"/>
+
| 2023 || February 7 || Google unveils Bard, its AI chatbot || Competition || Google presents its own AI chatbot called Bard, which is similar to ChatGPT. Google CEO Sundar Pichai says that Bard is more powerful and reliable than ChatGPT, as it draws on information from the web to provide fresh, high-quality responses. Bard is part of a series of tools based on generative artificial intelligence that Google launches to the market at this time.<ref>{{cite web |last1=Pascual |first1=M. G. |title=Google presents Bard, its answer to ChatGPT |url=https://english.elpais.com/science-tech/2023-02-07/google-presents-bard-its-answer-to-chatgpt.html |website=EL PAÍS English Edition |access-date=4 March 2023 |language=en-us |date=7 February 2023}}</ref>  
 
|-
 
|-
| 2021 || January 5 || || OpenAI introduces CLIP (Contrastive Language-Image Pre-training), a neural network that learns visual concepts from natural language supervision and can be applied to any visual classification benchmark. CLIP is trained on a variety of images with natural language supervision available on the internet and can be instructed in natural language to perform a variety of classification benchmarks without directly optimizing for the benchmark's performance. This approach improves the model's robustness and can match the performance of traditional models on benchmarks without using any labeled examples. CLIP's performance is more representative of how it will fare on datasets that measure accuracy in different settings. CLIP builds on previous work on zero-shot transfer, natural language supervision, and multimodal learning and uses a simple pre-training task to achieve competitive zero-shot performance on a variety of image classification datasets.<ref>{{cite web |title=CLIP: Connecting text and images |url=https://openai.com/research/clip |website=openai.com |access-date=24 March 2023}}</ref>
+
| 2023 || February 8 || Alibaba developing rival AI Chatbot || Competition || Chinese multinational technology company {{w|Alibaba Group}} announces that it is developing a rival to OpenAI's ChatGPT AI chatbot. The move comes as several other tech companies, including Google and Baidu, are also developing AI chatbot technology to rival ChatGPT.<ref name="Kharpal"/>
 
|-
 
|-
| 2022 || November 30 || Initial launch || GPT-3.5 is released.<ref name="tooltester"/> ChatGPT is introduced as a conversational AI model that can interact with users and respond to follow-up questions, challenge incorrect premises, and reject inappropriate requests. It was trained using {{w|Reinforcement Learning}} from Human Feedback (RLHF) and fine-tuned from the GPT-3.5 model. While it has limitations such as occasionally providing incorrect or nonsensical answers and being sensitive to input phrasing, it is designed with safety mitigations and the ability to learn from user feedback. ChatGPT is available for free during the research preview on chat.openai.com.<ref>{{cite web |title=Introducing ChatGPT |url=https://openai.com/blog/chatgpt |website=openai.com |access-date=16 March 2023}}</ref>
+
| 2023 || February 8 || Generation of academic essays undetectable by plagiarism tools || Performance || A study explores the potential of ChatGPT, a popular AI chatbot, in generating academic essays that can evade plagiarism detection tools. The results suggest that ChatGPT can produce high-quality content on various topics with high originality. The study highlights the need for institutions to take appropriate measures to address potential plagiarism issues arising from the use of AI technology in education.<ref>{{cite journal |last1=Khalil |first1=Mohammad |last2=Er |first2=Erkan |title=Will ChatGPT get you caught? Rethinking of Plagiarism Detection |journal=arXiv:2302.04335 [cs] |date=8 February 2023 |doi=10.48550/arXiv.2302.04335 |url=https://arxiv.org/abs/2302.04335}}</ref>
 
|-
 
|-
| 2022 || December 5 || Reception (positive) || Kevin Roose of ''The New York Times'' labels ChatGPT as "the best artificial intelligence chatbot ever released to the general public". ''{{w|The New York Times}}'' also reports that it has been "rumored" that the next version of the AI, {{w|GPT-4}}, will be launched sometime in 2023.<ref name=":4">{{Cite web |last=Roose |first=Kevin |date=December 5, 2022 |title=The Brilliance and Weirdness of ChatGPT |url=https://www.nytimes.com/2022/12/05/technology/chatgpt-ai-twitter.html |access-date= December 26, 2022 |website=The New York Times |language=en-US |quote=Like those tools, ChatGPT — which stands for "generative pre-trained transformer" — landed with a splash. |archive-date=January 18, 2023 |archive-url=https://web.archive.org/web/20230118134332/https://www.nytimes.com/2022/12/05/technology/chatgpt-ai-twitter.html}}</ref>  
+
| 2023 || February 8 || Framework to assess interactive language learning models is exemplified by ChatGPT || Performance || A paper proposes a framework for evaluating interactive language learning models (LLMs) such as ChatGPT using publicly available data sets. The authors evaluate ChatGPT's performance on 23 data sets covering eight different NLP tasks and find that ChatGPT outperforms other LLMs on most tasks, but has a lower accuracy in reasoning and suffers from hallucination problems. The authors also find that ChatGPT's interactive feature enables human collaboration to improve its performance. They release a codebase for evaluation set extraction.<ref>{{cite journal |last1=Bang |first1=Yejin |last2=Cahyawijaya |first2=Samuel |last3=Lee |first3=Nayeon |last4=Dai |first4=Wenliang |last5=Su |first5=Dan |last6=Wilie |first6=Bryan |last7=Lovenia |first7=Holy |last8=Ji |first8=Ziwei |last9=Yu |first9=Tiezheng |last10=Chung |first10=Willy |last11=Do |first11=Quyet V. |last12=Xu |first12=Yan |last13=Fung |first13=Pascale |title=A Multitask, Multilingual, Multimodal Evaluation of ChatGPT on Reasoning, Hallucination, and Interactivity |journal=arXiv:2302.04023 [cs] |date=28 February 2023 |doi=10.48550/arXiv.2302.04023 |url=https://arxiv.org/abs/2302.04023}}</ref>
 
|-
 
|-
| 2022 || December 5 || Ban || Popular coding question-and-answer site Stack Overflow temporarily bans the sharing of AI-generated answers produced by ChatGPT. The site's moderators say the ban is necessary as ChatGPT's answers are often incorrect and this can mislead users. ChatGPT ability to generate plausible but ultimately false responses is one of several well-known failings of AI text generation models.<ref>{{cite web |last1=Vincent |first1=James |title=AI-generated answers temporarily banned on coding Q&A site Stack Overflow |url=https://www.theverge.com/2022/12/5/23493932/chatgpt-ai-generated-answers-temporarily-banned-stack-overflow-llms-dangers |website=The Verge |access-date=6 March 2023 |date=5 December 2022}}</ref>
+
| 2023 || February 9 || ChatGPT-3's impact on student essay writing || Performance || A study compares students' essay writing performance with or without the use of ChatGPT-3 as a writing assistant tool. Eighteen students participate in the study, and essay scores, GPTs, writing time, authenticity, and content similarity were compared between the two groups. The study finds that using ChatGPT-3 did not improve essay quality, as the control group outperforms the experimental group in most parameters. The average grade is C for both groups, and none of the predictors affect essay scores. The text unauthenticity is slightly higher in the experimental group, but the similarity among essays is generally low.<ref>{{cite journal |last1=Basic |first1=Zeljana |last2=Banovac |first2=Ana |last3=Kruzic |first3=Ivana |last4=Jerkovic |first4=Ivan |title=Better by you, better than me, chatgpt3 as writing assistance in students essays |journal=arXiv:2302.04536 [cs] |date=9 February 2023 |doi=10.48550/arXiv.2302.04536 |url=https://arxiv.org/abs/2302.04536}}</ref>
 
|-
 
|-
| 2022 || December 27 || Policy || According to the BBC, as of date, ChatGPT is not allowed to "express political opinions or engage in political activism".<ref>{{cite news |title=Could a chatbot answer Prime Minister's Questions? |url=https://www.bbc.com/news/uk-politics-64053550 |first=Kate |last=Whannel |access-date= December 30, 2022 |work=BBC News |date=December 27, 2022 |archive-date=January 17, 2023 |archive-url=https://web.archive.org/web/20230117060725/https://www.bbc.com/news/uk-politics-64053550 }}</ref>
+
| 2023 || February 12 || Semantic communication scheme with ordered importance (SCOI) || Performance || A letter introduces a new semantic communication scheme with ordered importance (SCOI) that utilizes ChatGPT as a consulting assistant. In SCOI, the transmitter queries ChatGPT to determine the importance order of each word in the message, and then uses an unequal error protection transmission strategy to ensure reliable transmission of essential words. This approach is compatible with existing source-channel separation designs and can be easily incorporated into current communication systems. Experimental results show that SCOI outperforms existing communication schemes in terms of transmission bit error rate and semantic loss measured by ChatGPT.<ref>{{cite journal |last1=Guo |first1=Shuaishuai |last2=Wang |first2=Yanhu |last3=Li |first3=Shujing |last4=Saeed |first4=Nasir |title=Semantic Communications with Ordered Importance using ChatGPT |journal=arXiv:2302.07142 [eess] |date=12 February 2023 |doi=10.48550/arXiv.2302.07142 |url=https://arxiv.org/abs/2302.07142}}</ref>
 
|-
 
|-
| 2023 || January 5 || || A paper examines the political ideology of ChatGPT, and finds that it has a pro-environmental, left-libertarian orientation based on its responses to 630 political statements from voting advice applications and the political compass test. ChatGPT would impose taxes on flights, restrict rent increases, and legalize abortion. The findings are consistent across multiple languages and different experiment variations. The study highlights the potential impact of politically biased conversational AI on society.<ref>{{cite journal |last1=Hartmann |first1=Jochen |last2=Schwenzow |first2=Jasper |last3=Witte |first3=Maximilian |title=The political ideology of conversational AI: Converging evidence on ChatGPT's pro-environmental, left-libertarian orientation |date=2023 |doi=10.48550/arXiv.2301.01768 |url=https://arxiv.org/abs/2301.01768}}</ref>
+
| 2023 || February 14 || ChatGPT-generated hints vs. human tutors in algebra || Performance || A paper evaluates the effectiveness of ChatGPT-generated hints compared to those created by human tutors in two algebra topics. The study finds that 70% of hints produced by ChatGPT passed quality checks and both human and ChatGPT hints resulted in positive learning gains. However, the learning gains from human-created hints are significantly higher than ChatGPT hints in both topics. The authors discuss the limitations of the study and suggest future research directions. The problem and hint content used in the experiment is provided for replication purposes. The study suggests that while ChatGPT hints can be useful, human-created hints are still significantly more effective in promoting learning gains.<ref>{{cite journal |last1=Pardos |first1=Zachary A. |last2=Bhandari |first2=Shreya |title=Learning gain differences between ChatGPT and human tutor generated algebra hints |journal=arXiv:2302.06871 [cs] |date=14 February 2023 |url=https://arxiv.org/abs/2302.06871 |access-date=8 March 2023}}</ref>
 
|-
 
|-
| 2023 || January 27 || || It is reported that {{w|Elon Musk}} plans to form a new AI research lab to develop an alternative to OpenAI's ChatGPT, after having approached AI researchers, including Igor Babuschkin. At this time, the project is still in its early stages, with no concrete plan to develop specific products.<ref>{{cite web |title=Elon Musk recruits team to develop OpenAI’s ChatGPT rival: Reports |url=https://english.alarabiya.net/News/world/2023/02/28/Elon-Musk-recruits-team-to-develop-OpenAI-s-ChatGPT-rival-Reports |website=Al Arabiya English |access-date=4 March 2023 |language=en |date=28 February 2023}}</ref>
+
| 2023 || February 14 || India's first ChatGPT-powered chatbot || Competition || Indian fintech company Velocity launches India's first ChatGPT-powered chatbot named "Lexi." The AI development is integrated with its existing analytics tool, Velocity Insights, which offers e-commerce giants AI-powered business solutions. The integration enables users to leverage the same interface they use to drive business decisions. The Ministry of Electronics and IT also plans to integrate ChatGPT with WhatsApp to help Indian farmers understand and learn about several government schemes.<ref>{{cite web |title=India gets its first ChatGPT-powered AI chatbot, ‘Lexi’: All you need to know |url=https://www.livemint.com/technology/tech-news/india-gets-its-first-chatgpt-powered-ai-chatbot-lexi-all-you-need-to-know-11676364056207.html |website=livemint.com|access-date=4 March 2023 |language=en |date=14 February 2023}}</ref>  
 
|-
 
|-
| 2023 || February 1 || || OpenAI launches a pilot subscription plan called ChatGPT Plus, which offers faster response times, priority access to new features and improvements, and general access to ChatGPT even during peak times. The subscription costs $20/month and is available to customers worldwide. OpenAI still offers free access to ChatGPT but hopes to support its free access availability to as many people as possible through the subscription plan. OpenAI announces plans to refine and expand the offering based on user feedback and needs, and that it is actively exploring options for lower-cost plans, business plans, and data packs for more availability.<ref>{{cite web |title=Introducing ChatGPT Plus |url=https://openai.com/blog/chatgpt-plus |website=openai.com |access-date=16 March 2023}}</ref>
+
| 2023 || February 23 || ChatGPT as a translation tool || Performance || An article discusses the potential of ChatGPT as a translation tool and compares it to {{w|Google Translate}}. It highlights that ChatGPT is primarily a text-based system, while Google Translate offers live speech and image translation capabilities. The article presents an example comparing the translations of a tweet by both tools, where ChatGPT performs similarly to Google Translate with a minor difference in translation accuracy. It also mentions that ChatGPT seems to excel in translating content into English rather than the reverse. However, Google Translate has invested more in training for specific language combinations. The article concludes that while ChatGPT shows promise in translating text, language professionals can benefit from AI translation tools by using them judiciously alongside review and editorial processes to ensure quality and speed.<ref>{{cite web |last1=team |first1=The Marketing editorial |title=How Good Is ChatGPT at Translating Content? |url=https://www.translateplus.com/blog/chatgpt-translating-content/#:~:text=This%20article%20from%20Makeuseof.com,Spanish%20and%20German%20into%20English. |website=translate plus |access-date=24 May 2023 |date=23 February 2023}}</ref>
 
|-
 
|-
| 2023 || February 1 || Study || A study explores the opinions and perceptions of academicians, netizens, and information professionals towards ChatGPT-3, using social media comments and a survey. The research suggests that while ChatGPT-3 can be useful for research and writing, it is not entirely accurate and should be cross-checked. It also suggests that most academicians are beginning to accept ChatGPT-3, despite some resistance. The study is useful for academics, content developers, and librarians.<ref>{{cite journal |last1=A |first1=Subaveerapandiyan |last2=A |first2=Vinoth |last3=Tiwary |first3=Neelam |title=Netizens, Academicians, and Information Professionals' Opinions About AI With Special Reference To ChatGPT |journal=arXiv:2302.07136 [cs] |date=1 February 2023 |doi=10.48550/arXiv.2302.07136 |url=https://arxiv.org/abs/2302.07136}}</ref>
+
| 2023 || February 15 || ChatGPT in human-centered design || Performance || A study explores the use of ChatGPT in a human-centered design process. A hypothetical design project is conducted, where ChatGPT is utilized to generate personas, simulate interviews with fictional users, create new design ideas, simulate usage scenarios and conversations between an imaginary prototype and fictional users, and evaluate user experience. The results show that ChatGPT effectively performs the tasks assigned to it as a designer, user, or product, providing mostly appropriate responses, but with some drawbacks such as forgotten information, partial responses, and a lack of output diversity. The study discusses the potential benefits and limitations of using conversational LLMs in design and suggests directions for future research in this area.<ref>{{cite journal |last1=Kocaballi |first1=A. Baki |title=Conversational AI-Powered Design: ChatGPT as Designer, User, and Product |journal=arXiv:2302.07406 [cs] |date=14 February 2023 |doi=10.48550/arXiv.2302.07406 |url=https://arxiv.org/abs/2302.07406}}</ref>
 
|-
 
|-
| 2023 || February 2 || Growth || It is reported that ChatGPT reached an estimated 100 million monthly active users in January, just two months after its launch, making it the fastest-growing consumer application in history, according to a UBS study. The growing usage, while imposing substantial computing cost on OpenAI, also provides valuable feedback to help train the chatbot's responses.<ref>{{cite web |last1=Hu |first1=Krystal |title=ChatGPT sets record for fastest-growing user base - analyst note |url=https://www.reuters.com/technology/chatgpt-sets-record-fastest-growing-user-base-analyst-note-2023-02-01/ |website=Reuters |access-date=14 March 2023 |language=en |date=2 February 2023}}</ref>
+
| 2023 || February 16 || Warning on ChatGPT's hallucination || System behavior || Google's senior vice president, {{w|Prabhakar Raghavan}}, issues a warning about ChatGPT, expressing concerns that the technology could provide "completely fictitious" answers to questions, which he refers to as "[[w:Hallucination (artificial intelligence)|hallucination]]." He adds that the huge language models behind this technology make it impossible for humans to monitor every conceivable behavior of the system. Around this time, teachers in the United Kingdom also consider how to prevent the use of ChatGPT to cheat on school coursework.<ref>{{cite web |last1=Kindred |first1=Alahna |title=Google issues urgent warning to anybody using AI chatbot ChatGPT |url=https://www.mirror.co.uk/news/technology-science/google-boss-issues-urgent-warning-29229679 |website=mirror |access-date=4 March 2023 |language=en |date=16 February 2023}}</ref>
 
|-
 
|-
| 2023 || February 7 || Competition || Google presents its own AI chatbot called Bard, which is similar to ChatGPT. Google CEO Sundar Pichai says that Bard is more powerful and reliable than ChatGPT, as it draws on information from the web to provide fresh, high-quality responses. Bard is part of a series of tools based on generative artificial intelligence that Google launches to the market at this time.<ref>{{cite web |last1=Pascual |first1=M. G. |title=Google presents Bard, its answer to ChatGPT |url=https://english.elpais.com/science-tech/2023-02-07/google-presents-bard-its-answer-to-chatgpt.html |website=EL PAÍS English Edition |access-date=4 March 2023 |language=en-us |date=7 February 2023}}</ref>  
+
| 2023 || February 16 || Paper evaluates ChatGPT for text summarization tasks || Performance || A paper explores the performance of ChatGPT for query or aspect-based text summarization. The authors conduct an evaluation of ChatGPT's performance on four benchmark datasets and find that its performance is comparable to traditional fine-tuning methods in terms of Rouge scores. The authors also highlight some unique differences between ChatGPT-generated summaries and human references, providing valuable insights into the capabilities of ChatGPT for diverse text summarization tasks. The findings call for further research to systematically examine the characteristics of ChatGPT-generated summaries through extensive human evaluation.<ref>{{cite journal |last1=Yang |first1=Xianjun |last2=Li |first2=Yan |last3=Zhang |first3=Xinlu |last4=Chen |first4=Haifeng |last5=Cheng |first5=Wei |title=Exploring the Limits of ChatGPT for Query or Aspect-based Text Summarization |journal=arXiv:2302.08081 [cs] |date=15 February 2023 |doi=10.48550/arXiv.2302.08081 |url=https://arxiv.org/abs/2302.08081 |access-date=7 March 2023}}</ref>
 
|-
 
|-
| 2023 || February 8 || Competition || Chinese multinational technology company {{w|Alibaba Group}} announces that it is developing a rival to OpenAI's ChatGPT AI chatbot. The move comes as several other tech companies, including Google and Baidu, are also developing AI chatbot technology to rival ChatGPT.<ref name="Kharpal"/>
+
| 2023 || February 16 || OpenAI's blog post on ChatGPT's behavior and improvements || System behavior || OpenAI shares a blog post describing how ChatGPT's behavior is shaped, as well as the company's plans for improving the behavior of the system, allowing more user customization, and getting more public input into decision-making. OpenAI explains that unlike ordinary software, their models are massive {{w|neural network}}s that learn from a broad range of data, not programmed explicitly. The models undergo a two-step process: pre-training and fine-tuning. OpenAI claims it is committed to robustly addressing biases in AI systems and being transparent about their intentions and progress. The company reports it is researching how to make the fine-tuning process more understandable and controllable, and building on external advances such as rule-based rewards and Constitutional AI.<ref>{{cite web |title=How should AI systems behave, and who should decide? |url=https://openai.com/blog/how-should-ai-systems-behave |website=openai.com |access-date=16 March 2023}}</ref>
 
|-
 
|-
| 2023 || February 8 || Study || A study explores the potential of ChatGPT, a popular AI chatbot, in generating academic essays that can evade plagiarism detection tools. The results suggest that ChatGPT can produce high-quality content on various topics with high originality. The study highlights the need for institutions to take appropriate measures to address potential plagiarism issues arising from the use of AI technology in education.<ref>{{cite journal |last1=Khalil |first1=Mohammad |last2=Er |first2=Erkan |title=Will ChatGPT get you caught? Rethinking of Plagiarism Detection |journal=arXiv:2302.04335 [cs] |date=8 February 2023 |doi=10.48550/arXiv.2302.04335 |url=https://arxiv.org/abs/2302.04335}}</ref>
+
| 2023 || February 17 || Exploring ChatGPT in statistical process control || Performance || A paper explores the potential benefits and limitations of using generative AI models like ChatGPT in {{w|Statistical Process Control}} (SPC) practice, education, and research. The study investigates ChatGPT's ability to provide code, explain concepts, and create knowledge related to SPC by analyzing responses to structured prompts. The results indicate that ChatGPT performs well for structured tasks but struggles with more nuanced tasks, emphasizing the need for proper validation and the use of multiple methods to ensure accurate results when utilizing generative AI models in SPC.<ref>{{cite journal |last1=Megahed |first1=Fadel M. |last2=Chen |first2=Ying-Ju |last3=Ferris |first3=Joshua A. |last4=Knoth |first4=Sven |last5=Jones-Farmer |first5=L. Allison |title=How Generative AI models such as ChatGPT can be (Mis)Used in SPC Practice, Education, and Research? An Exploratory Study |journal=arXiv:2302.10916 [cs] |date=17 February 2023 |url=https://arxiv.org/abs/2302.10916 |access-date=7 March 2023}}</ref><ref>{{cite journal |last1=Megahed |first1=Fadel M. |last2=Chen |first2=Ying-Ju |last3=Ferris |first3=Joshua A. |last4=Knoth |first4=Sven |last5=Jones-Farmer |first5=L. Allison |title=How Generative AI models such as ChatGPT can be (Mis)Used in SPC Practice, Education, and Research? An Exploratory Study |journal=arXiv:2302.10916 [cs] |date=17 February 2023 |doi=10.48550/arXiv.2302.10916 |url=https://arxiv.org/abs/2302.10916}}</ref>
 
|-
 
|-
| 2023 || February 8 || Study || A paper proposes a framework for evaluating interactive language learning models (LLMs) such as ChatGPT using publicly available data sets. The authors evaluate ChatGPT's performance on 23 data sets covering eight different NLP tasks and find that ChatGPT outperforms other LLMs on most tasks, but has a lower accuracy in reasoning and suffers from hallucination problems. The authors also find that ChatGPT's interactive feature enables human collaboration to improve its performance. They release a codebase for evaluation set extraction.<ref>{{cite journal |last1=Bang |first1=Yejin |last2=Cahyawijaya |first2=Samuel |last3=Lee |first3=Nayeon |last4=Dai |first4=Wenliang |last5=Su |first5=Dan |last6=Wilie |first6=Bryan |last7=Lovenia |first7=Holy |last8=Ji |first8=Ziwei |last9=Yu |first9=Tiezheng |last10=Chung |first10=Willy |last11=Do |first11=Quyet V. |last12=Xu |first12=Yan |last13=Fung |first13=Pascale |title=A Multitask, Multilingual, Multimodal Evaluation of ChatGPT on Reasoning, Hallucination, and Interactivity |journal=arXiv:2302.04023 [cs] |date=28 February 2023 |doi=10.48550/arXiv.2302.04023 |url=https://arxiv.org/abs/2302.04023}}</ref>
+
| 2023 || February 19 || Comparing ChatGPT and BERT on GLUE Benchmark || Performance || A study compares the understanding ability of ChatGPT and fine-tuned BERT on the GLUE benchmark. The study finds that ChatGPT performs well on inference tasks but struggles with paraphrase and similarity tasks. ChatGPT achieves comparable performance with BERT on {{w|sentiment analysis}} and question-answering tasks. The study also shows that the understanding ability of ChatGPT can be improved with advanced prompting strategies.<ref>{{cite journal |last1=Zhong |first1=Qihuang |last2=Ding |first2=Liang |last3=Liu |first3=Juhua |last4=Du |first4=Bo |last5=Tao |first5=Dacheng |title=Can ChatGPT Understand Too? A Comparative Study on ChatGPT and Fine-tuned BERT |journal=arXiv:2302.10198 [cs] |date=2 March 2023 |doi=10.48550/arXiv.2302.10198 |url=https://arxiv.org/abs/2302.10198}}</ref>
 
|-
 
|-
| 2023 || February 9 || Study || A study compares students' essay writing performance with or without the use of ChatGPT-3 as a writing assistant tool. Eighteen students participated in the study, and essay scores, GPTs, writing time, authenticity, and content similarity were compared between the two groups. The study found that using ChatGPT-3 did not improve essay quality, as the control group outperformed the experimental group in most parameters. The average grade was C for both groups, and none of the predictors affected essay scores. The text unauthenticity was slightly higher in the experimental group, but the similarity among essays was generally low.<ref>{{cite journal |last1=Basic |first1=Zeljana |last2=Banovac |first2=Ana |last3=Kruzic |first3=Ivana |last4=Jerkovic |first4=Ivan |title=Better by you, better than me, chatgpt3 as writing assistance in students essays |journal=arXiv:2302.04536 [cs] |date=9 February 2023 |doi=10.48550/arXiv.2302.04536 |url=https://arxiv.org/abs/2302.04536}}</ref>
+
| 2023 || February 20 || Perception and impact of ChatGPT in social media and scientific discourse || Reception || A paper analyzes over 300,000 tweets and 150 scientific papers to investigate how ChatGPT is perceived and discussed after its release in November 2022. The analysis shows that ChatGPT is generally viewed positively with high quality, but its perception has slightly decreased since its debut. Positive emotions dominate in social media, with joy being the most frequent, while negative sentiment and surprise are on the rise. ChatGPT is perceived more negatively in languages other than English. Scientific papers characterize ChatGPT as a great opportunity across various fields, including the medical domain, but also as a threat concerning ethics, with mixed assessments for education. The authors make their data available to contribute to shaping the public debate and informing future development.<ref>{{cite journal |last1=Leiter |first1=Christoph |last2=Zhang |first2=Ran |last3=Chen |first3=Yanran |last4=Belouadi |first4=Jonas |last5=Larionov |first5=Daniil |last6=Fresen |first6=Vivian |last7=Eger |first7=Steffen |title=ChatGPT: A Meta-Analysis after 2.5 Months |journal=arXiv:2302.13795 [cs] |date=20 February 2023 |doi=10.48550/arXiv.2302.13795 |url=https://arxiv.org/abs/2302.13795 |access-date=6 March 2023}}</ref>
 
|-
 
|-
| 2023 || February 12 || Study || A letter introduces a new semantic communication scheme with ordered importance (SCOI) that utilizes ChatGPT as a consulting assistant. In SCOI, the transmitter queries ChatGPT to determine the importance order of each word in the message, and then uses an unequal error protection transmission strategy to ensure reliable transmission of essential words. This approach is compatible with existing source-channel separation designs and can be easily incorporated into current communication systems. Experimental results show that SCOI outperforms existing communication schemes in terms of transmission bit error rate and semantic loss measured by ChatGPT.<ref>{{cite journal |last1=Guo |first1=Shuaishuai |last2=Wang |first2=Yanhu |last3=Li |first3=Shujing |last4=Saeed |first4=Nasir |title=Semantic Communications with Ordered Importance using ChatGPT |journal=arXiv:2302.07142 [eess] |date=12 February 2023 |doi=10.48550/arXiv.2302.07142 |url=https://arxiv.org/abs/2302.07142}}</ref>
+
| 2023 || February 20 || ChatIE, zero-shot information extraction using ChatGPT || Performance || A study explores the use of ChatGPT in zero-shot {{w|information extraction}} (IE) to build IE systems from unannotated text. The authors transform the zero-shot IE task into a multi-turn question-answering problem with a two-stage framework (ChatIE) and evaluate it on three IE tasks: entity-relation triple extract, named entity recognition, and event extraction. ChatIE achieves impressive performance and even surpasses some full-shot models on several datasets, indicating the potential of prompt-based methods to reduce the time and effort that data labeling takes.<ref>{{cite journal |last1=Wei |first1=Xiang |last2=Cui |first2=Xingyu |last3=Cheng |first3=Ning |last4=Wang |first4=Xiaobin |last5=Zhang |first5=Xin |last6=Huang |first6=Shen |last7=Xie |first7=Pengjun |last8=Xu |first8=Jinan |last9=Chen |first9=Yufeng |last10=Zhang |first10=Meishan |last11=Jiang |first11=Yong |last12=Han |first12=Wenjuan |title=Zero-Shot Information Extraction via Chatting with ChatGPT |journal=arXiv:2302.10205 [cs] |date=20 February 2023 |doi=10.48550/arXiv.2302.10205 |url=https://arxiv.org/abs/2302.10205 |access-date=7 March 2023}}</ref>
 
|-
 
|-
| 2023 || February 14 || Study || A paper evaluates the effectiveness of ChatGPT-generated hints compared to those created by human tutors in two algebra topics. The study found that 70% of hints produced by ChatGPT passed quality checks and both human and ChatGPT hints resulted in positive learning gains. However, the learning gains from human-created hints were significantly higher than ChatGPT hints in both topics. The authors discuss the limitations of the study and suggest future research directions. The problem and hint content used in the experiment is provided for replication purposes. The study suggests that while ChatGPT hints can be useful, human-created hints are still significantly more effective in promoting learning gains.<ref>{{cite journal |last1=Pardos |first1=Zachary A. |last2=Bhandari |first2=Shreya |title=Learning gain differences between ChatGPT and human tutor generated algebra hints |journal=arXiv:2302.06871 [cs] |date=14 February 2023 |url=https://arxiv.org/abs/2302.06871 |access-date=8 March 2023}}</ref>
+
| 2023 || February 21 || Assessing ChatGPT across NLP tasks, capabilities and limitations || Performance || A paper evaluates the capabilities of ChatGPT on 25 diverse analytical {{w|natural language processing}} tasks, including sentiment analysis, emotion recognition, offensiveness and stance detection, natural language inference, word sense disambiguation, linguistic acceptability, and question answering. The study finds that ChatGPT's average loss in quality is about 25% for zero-shot and few-shot evaluation, and the more difficult the task, the higher the ChatGPT loss. The paper also discusses ChatGPT's bias and the need for validation procedures for such systems.<ref>{{cite journal |last1=Kocoń |first1=Jan |last2=Cichecki |first2=Igor |last3=Kaszyca |first3=Oliwier |last4=Kochanek |first4=Mateusz |last5=Szydło |first5=Dominika |last6=Baran |first6=Joanna |last7=Bielaniewicz |first7=Julita |last8=Gruza |first8=Marcin |last9=Janz |first9=Arkadiusz |last10=Kanclerz |first10=Kamil |last11=Kocoń |first11=Anna |last12=Koptyra |first12=Bartłomiej |last13=Mieleszczenko-Kowszewicz |first13=Wiktoria |last14=Miłkowski |first14=Piotr |last15=Oleksy |first15=Marcin |last16=Piasecki |first16=Maciej |last17=Radliński |first17=Łukasz |last18=Wojtasik |first18=Konrad |last19=Woźniak |first19=Stanisław |last20=Kazienko |first20=Przemysław |title=ChatGPT: Jack of all trades, master of none |journal=arXiv:2302.10724 [cs] |date=21 February 2023 |doi=10.48550/arXiv.2302.10724 |url=https://arxiv.org/abs/2302.10724 |access-date=7 March 2023}}</ref>
 
|-
 
|-
| 2023 || February 14 || || Indian fintech company Velocity launches India's first ChatGPT-powered chatbot named "Lexi." The AI development is integrated with its existing analytics tool, Velocity Insights, which offers e-commerce giants AI-powered business solutions. The integration enables users to leverage the same interface they use to drive business decisions. The Ministry of Electronics and IT also plans to integrate ChatGPT with WhatsApp to help Indian farmers understand and learn about several government schemes.<ref>{{cite web |title=India gets its first ChatGPT-powered AI chatbot, ‘Lexi’: All you need to know |url=https://www.livemint.com/technology/tech-news/india-gets-its-first-chatgpt-powered-ai-chatbot-lexi-all-you-need-to-know-11676364056207.html |website=livemint.com|access-date=4 March 2023 |language=en |date=14 February 2023}}</ref>  
+
| 2023 || February 22 || Robustness evaluation of ChatGPT, adversarial and out-of-distribution perspectives || Performance || A paper evaluates the robustness of ChatGPT from the adversarial and out-of-distribution perspective. The authors employ AdvGLUE and ANLI benchmarks to assess adversarial robustness and {{w|Flipkart}} review and DDXPlus medical diagnosis datasets for out-of-distribution (OOD) evaluation. ChatGPT shows consistent advantages on most adversarial and OOD classification and translation tasks, but the absolute performance is still far from perfection, indicating that robustness remains a significant threat. ChatGPT shows astounding performance in understanding dialogue-related texts and tends to provide informal suggestions for medical tasks. The paper concludes with in-depth discussions of possible research directions.<ref>{{cite journal |last1=Wang |first1=Jindong |last2=Hu |first2=Xixu |last3=Hou |first3=Wenxin |last4=Chen |first4=Hao |last5=Zheng |first5=Runkai |last6=Wang |first6=Yidong |last7=Yang |first7=Linyi |last8=Huang |first8=Haojun |last9=Ye |first9=Wei |last10=Geng |first10=Xiubo |last11=Jiao |first11=Binxin |last12=Zhang |first12=Yue |last13=Xie |first13=Xing |title=On the Robustness of ChatGPT: An Adversarial and Out-of-distribution Perspective |journal=arXiv:2302.12095 [cs] |date=2 March 2023 |doi=10.48550/arXiv.2302.12095 |url=https://arxiv.org/abs/2302.12095 |access-date=7 March 2023}}</ref>
 
|-
 
|-
| 2023 || February 15 || Study || A study explores the use of ChatGPT in a human-centered design process. A hypothetical design project was conducted, where ChatGPT was utilized to generate personas, simulate interviews with fictional users, create new design ideas, simulate usage scenarios and conversations between an imaginary prototype and fictional users, and evaluate user experience. The results show that ChatGPT effectively performed the tasks assigned to it as a designer, user, or product, providing mostly appropriate responses, but with some drawbacks such as forgotten information, partial responses, and a lack of output diversity. The study discusses the potential benefits and limitations of using conversational LLMs in design and suggests directions for future research in this area.<ref>{{cite journal |last1=Kocaballi |first1=A. Baki |title=Conversational AI-Powered Design: ChatGPT as Designer, User, and Product |journal=arXiv:2302.07406 [cs] |date=14 February 2023 |doi=10.48550/arXiv.2302.07406 |url=https://arxiv.org/abs/2302.07406}}</ref>
+
| 2023 || February 23 || Impact of prompt knowledge on ChatGPT's health advice || Performance || A paper discusses the impact of prompt knowledge on the correctness of answers generated by ChatGPT in the context of consumers seeking health advice from the model. The authors compare the effectiveness of ChatGPT in generating answers when it leverages the model's knowledge alone versus in combination with prompt knowledge. They show that the knowledge provided in the prompt can overturn the knowledge encoded in the model, leading to lower answer correctness. The study has implications for the development of more robust and transparent question-answering systems based on generative pre-trained language models.<ref>{{cite journal |last1=Zuccon |first1=Guido |last2=Koopman |first2=Bevan |title=Dr ChatGPT, tell me what I want to hear: How prompt knowledge impacts health answer correctness |journal=arXiv:2302.13793 [cs] |date=23 February 2023 |doi=10.48550/arXiv.2302.13793 |url=https://arxiv.org/abs/2302.13793 |access-date=6 March 2023}}</ref>
 
|-
 
|-
| 2023 || February 16 || Notable comment || Google's senior vice president, Prabhakar Raghavan, issues a warning about ChatGPT, expressing concerns that the technology could provide "completely fictitious" answers to questions, which he refers to as "hallucination." He adds that the huge language models behind this technology make it impossible for humans to monitor every conceivable behavior of the system. Around this time, teachers in the United Kingdom also consider how to prevent the use of ChatGPT to cheat on school coursework.<ref>{{cite web |last1=Kindred |first1=Alahna |title=Google issues urgent warning to anybody using AI chatbot ChatGPT |url=https://www.mirror.co.uk/news/technology-science/google-boss-issues-urgent-warning-29229679 |website=mirror |access-date=4 March 2023 |language=en |date=16 February 2023}}</ref>
+
| 2023 || February 23 || ChatGPT evaluation on mathematical word problems || Performance || A paper presents the first independent evaluation of ChatGPT on [[w:Word problem (mathematics education)|mathematical word problems]] (MWPs) from the dataset DRAW-1K. The authors find that ChatGPT's performance varies dramatically based on the requirement to show its work, with a failure rate of 20% when it provides work and 84% when it does not. The paper also highlights several factors relating to the number of unknowns and operations that increase the probability of failure, with the probability of failure increasing linearly with the number of addition and subtraction operations.<ref>{{cite journal |last1=Shakarian |first1=Paulo |last2=Koyyalamudi |first2=Abhinav |last3=Ngu |first3=Noel |last4=Mareedu |first4=Lakshmivihari |title=An Independent Evaluation of ChatGPT on Mathematical Word Problems (MWP) |journal=arXiv:2302.13814 [cs] |date=27 February 2023 |doi=10.48550/arXiv.2302.13814 |url=https://arxiv.org/abs/2302.13814}}</ref>
 
|-
 
|-
| 2023 || February 16 || Study || A paper explores the performance of ChatGPT for query or aspect-based text summarization. The authors conducted an evaluation of ChatGPT's performance on four benchmark datasets and found that its performance is comparable to traditional fine-tuning methods in terms of Rouge scores. The authors also highlight some unique differences between ChatGPT-generated summaries and human references, providing valuable insights into the capabilities of ChatGPT for diverse text summarization tasks. The findings call for further research to systematically examine the characteristics of ChatGPT-generated summaries through extensive human evaluation.<ref>{{cite journal |last1=Yang |first1=Xianjun |last2=Li |first2=Yan |last3=Zhang |first3=Xinlu |last4=Chen |first4=Haifeng |last5=Cheng |first5=Wei |title=Exploring the Limits of ChatGPT for Query or Aspect-based Text Summarization |journal=arXiv:2302.08081 [cs] |date=15 February 2023 |doi=10.48550/arXiv.2302.08081 |url=https://arxiv.org/abs/2302.08081 |access-date=7 March 2023}}</ref>
+
| 2023 || February 26 || DevBots in software engineering, ChatGPT's role in architecting || Performance || A paper proposes the use of Software Development Bots (DevBots) trained on large language models to aid in software architecture-centric software engineering (ACSE). The paper specifically focuses on the use of ChatGPT for collaboration between a novice software architect and the AI model in analyzing, synthesizing, and evaluating a services-driven software application. Preliminary results indicate that ChatGPT can mimic an architect's role but requires human oversight and decision support for collaborative architecting. The authors suggest further research into harnessing empirical evidence and exploring socio-technical aspects of architecting with ChatGPT.<ref>{{cite journal |last1=Ahmad |first1=Aakash |last2=Waseem |first2=Muhammad |last3=Liang |first3=Peng |last4=Fehmideh |first4=Mahdi |last5=Aktar |first5=Mst Shamima |last6=Mikkonen |first6=Tommi |title=Towards Human-Bot Collaborative Software Architecting with ChatGPT |journal=arXiv:2302.14600 [cs] |date=26 February 2023 |doi=10.48550/arXiv.2302.14600 |url=https://arxiv.org/abs/2302.14600}}</ref>
 
|-
 
|-
| 2023 || February 16 || || OpenAI shares a blog post on how ChatGPT's behavior is shaped, and the company's plans for improving the behavior of the system, allowing more user customization, and getting more public input into decision-making. OpenAI explains that unlike ordinary software, their models are massive neural networks that learn from a broad range of data, not programmed explicitly. The models undergo a two-step process: pre-training and fine-tuning. OpenAI claims it is committed to robustly addressing biases in AI systems and being transparent about their intentions and progress. The company reports it is researching how to make the fine-tuning process more understandable and controllable, and building on external advances such as rule-based rewards and Constitutional AI.<ref>{{cite web |title=How should AI systems behave, and who should decide? |url=https://openai.com/blog/how-should-ai-systems-behave |website=openai.com |access-date=16 March 2023}}</ref>
+
| 2023 || February 27 || Overview of ChatGPT, technology, applications, and ethical considerations || Study || A paper provides an overview of the technology behind ChatGPT. It discusses the historical background of chatbots and the potential applications of ChatGPT in various domains such as healthcare, education, and research. However, the paper also highlights the privacy and ethical concerns surrounding ChatGPT and the limitations of the current version. The paper includes an interview with ChatGPT and presents its responses to various questions.<ref>{{cite journal |last1=Shahriar |first1=Sakib |last2=Hayawi |first2=Kadhim |title=Let's have a chat! A Conversation with ChatGPT: Technology, Applications, and Limitations |date=2023 |doi=10.48550/arXiv.2302.13817 |url=https://arxiv.org/abs/2302.13817}}</ref>
 
|-
 
|-
| 2023 || February 17 || Study || A paper explores the potential benefits and limitations of using generative AI models like ChatGPT in {{w|Statistical Process Control}} (SPC) practice, education, and research. The study investigates ChatGPT's ability to provide code, explain concepts, and create knowledge related to SPC by analyzing responses to structured prompts. The results indicate that ChatGPT performs well for structured tasks but struggles with more nuanced tasks, emphasizing the need for proper validation and the use of multiple methods to ensure accurate results when utilizing generative AI models in SPC.<ref>{{cite journal |last1=Megahed |first1=Fadel M. |last2=Chen |first2=Ying-Ju |last3=Ferris |first3=Joshua A. |last4=Knoth |first4=Sven |last5=Jones-Farmer |first5=L. Allison |title=How Generative AI models such as ChatGPT can be (Mis)Used in SPC Practice, Education, and Research? An Exploratory Study |journal=arXiv:2302.10916 [cs] |date=17 February 2023 |url=https://arxiv.org/abs/2302.10916 |access-date=7 March 2023}}</ref><ref>{{cite journal |last1=Megahed |first1=Fadel M. |last2=Chen |first2=Ying-Ju |last3=Ferris |first3=Joshua A. |last4=Knoth |first4=Sven |last5=Jones-Farmer |first5=L. Allison |title=How Generative AI models such as ChatGPT can be (Mis)Used in SPC Practice, Education, and Research? An Exploratory Study |journal=arXiv:2302.10916 [cs] |date=17 February 2023 |doi=10.48550/arXiv.2302.10916 |url=https://arxiv.org/abs/2302.10916}}</ref>
+
| 2023 || February 28 || Empirical study, ChatGPT's performance in cross-lingual summarization || Performance || A paper presents an empirical study that investigates the performance of ChatGPT in {{w|cross-lingual summarization}} (CLS), which involves generating a summary in a different target language from a source language document. The authors use various prompts to guide ChatGPT to perform zero-shot CLS from different paradigms and provide a preliminary evaluation of its generated summaries. The study finds that ChatGPT initially prefers to produce lengthy summaries with more detailed information, but with the help of interactive prompts, it can significantly improve its CLS performance by balancing informativeness and conciseness. Experimental results on three widely-used CLS datasets show that ChatGPT outperforms the advanced GPT 3.5 model. The paper provides qualitative case studies to demonstrate the superiority of ChatGPT on CLS.<ref>{{cite journal |last1=Wang |first1=Jiaan |last2=Liang |first2=Yunlong |last3=Meng |first3=Fandong |last4=Li |first4=Zhixu |last5=Qu |first5=Jianfeng |last6=Zhou |first6=Jie |title=Cross-Lingual Summarization via ChatGPT |journal=arXiv:2302.14229 [cs] |date=27 February 2023 |url=https://arxiv.org/abs/2302.14229 |access-date=6 March 2023}}</ref>
 
|-
 
|-
| 2023 || February 19 || Study || A study compares the understanding ability of ChatGPT and fine-tuned BERT on the GLUE benchmark. The study finds that ChatGPT performs well on inference tasks but struggles with paraphrase and similarity tasks. ChatGPT achieves comparable performance with BERT on sentiment analysis and question-answering tasks. The study also shows that the understanding ability of ChatGPT can be improved with advanced prompting strategies.<ref>{{cite journal |last1=Zhong |first1=Qihuang |last2=Ding |first2=Liang |last3=Liu |first3=Juhua |last4=Du |first4=Bo |last5=Tao |first5=Dacheng |title=Can ChatGPT Understand Too? A Comparative Study on ChatGPT and Fine-tuned BERT |journal=arXiv:2302.10198 [cs] |date=2 March 2023 |doi=10.48550/arXiv.2302.10198 |url=https://arxiv.org/abs/2302.10198}}</ref>
+
| 2023 || February || Colombian judge utilizes ChatGPT in autism medical rights case || Adoption || A Colombian judge reportedly uses ChatGPT to help rule on a case about the medical rights of a child with autism. The case is about whether the child's insurance should cover the bills for his medical treatments. The judge uses ChatGPT in accordance with a [[w:Judiciary of Colombia|law in Colombia]] that allows the use of virtual tools in some cases. Opinions are divided, with some professionals opposing the use of AI in a legal setting, while others see it as a useful tool.<ref name="Bhaimiya">{{cite web |last1=Bhaimiya |first1=Sawdah |title=A Colombian judge used ChatGPT to help rule on a case about the medical rights of a child with autism, reports say |url=https://www.businessinsider.com/chatgpt-colombian-judge-used-in-ruling-child-medical-rights-report-2023-2 |website=Business Insider |access-date=4 March 2023}}</ref>  
 
|-
 
|-
| 2023 || February 20 || Study || A paper analyzes over 300,000 tweets and 150 scientific papers to investigate how ChatGPT is perceived and discussed after its release in November 2022. The analysis shows that ChatGPT is generally viewed positively with high quality, but its perception has slightly decreased since its debut. Positive emotions dominate in social media, with joy being the most frequent, while negative sentiment and surprise are on the rise. ChatGPT is perceived more negatively in languages other than English. Scientific papers characterize ChatGPT as a great opportunity across various fields, including the medical domain, but also as a threat concerning ethics, with mixed assessments for education. The authors make their data available to contribute to shaping the public debate and informing future development.<ref>{{cite journal |last1=Leiter |first1=Christoph |last2=Zhang |first2=Ran |last3=Chen |first3=Yanran |last4=Belouadi |first4=Jonas |last5=Larionov |first5=Daniil |last6=Fresen |first6=Vivian |last7=Eger |first7=Steffen |title=ChatGPT: A Meta-Analysis after 2.5 Months |journal=arXiv:2302.13795 [cs] |date=20 February 2023 |doi=10.48550/arXiv.2302.13795 |url=https://arxiv.org/abs/2302.13795 |access-date=6 March 2023}}</ref>
+
| 2023 || February || UAE small businesses embrace ChatGPT to boost productivity || Adoption || Small businesses in the {{w|United Arab Emirates}} turn to ChatGPT to reduce the amount of time spent on mundane tasks such as writing job descriptions and blog posts, thereby allowing them to concentrate on more creative aspects of their business. By this time, ChatGPT is used by large companies including [[W:Meta Platforms|Meta]] and {{w|Shopify}}. The ''{{w|Wall Street Journal}}'' speculates that the tool could eventually lead to job losses in some areas, although ChatGPT's flaws, including providing inaccurate responses, and the need for further regulation of AI technology so far prevent this for the time being.<ref>{{cite web |title=UAE small businesses embrace ChatGPT as tech allows productivity boost |url=https://english.alarabiya.net/News/gulf/2023/02/15/UAE-small-businesses-embrace-ChatGPT-as-tech-allows-productivity-boost |website=Al Arabiya English |access-date=4 March 2023 |language=en |date=15 February 2023}}</ref>
 
|-
 
|-
| 2023 || February 20 || Study || A study explores the use of ChatGPT in zero-shot {{w|information extraction}} (IE) to build IE systems from unannotated text. The authors transform the zero-shot IE task into a multi-turn question-answering problem with a two-stage framework (ChatIE) and evaluate it on three IE tasks: entity-relation triple extract, named entity recognition, and event extraction. ChatIE achieves impressive performance and even surpasses some full-shot models on several datasets, indicating the potential of prompt-based methods to reduce the time and effort that data labeling takes.<ref>{{cite journal |last1=Wei |first1=Xiang |last2=Cui |first2=Xingyu |last3=Cheng |first3=Ning |last4=Wang |first4=Xiaobin |last5=Zhang |first5=Xin |last6=Huang |first6=Shen |last7=Xie |first7=Pengjun |last8=Xu |first8=Jinan |last9=Chen |first9=Yufeng |last10=Zhang |first10=Meishan |last11=Jiang |first11=Yong |last12=Han |first12=Wenjuan |title=Zero-Shot Information Extraction via Chatting with ChatGPT |journal=arXiv:2302.10205 [cs] |date=20 February 2023 |doi=10.48550/arXiv.2302.10205 |url=https://arxiv.org/abs/2302.10205 |access-date=7 March 2023}}</ref>
+
| 2023 || February || ChatGPT's poor performance in Singapore PSLE raises concerns || Performance || ChatGPT is found to perform poorly when asked to answer questions from {{w|Singapore}}'s Primary School Leaving Exam (PSLE), with an average score of 16 out of 100 marks for mathematics and 21 for science. It is found to make errors even in simple addition and can not understand or answer any questions that reference diagrams or graphs. However, ChatGPT manages to pass the English tests and scores an average of 11 out of 20 marks across the three papers it took. The bot's poor performance in exams raises concerns about the possibility of AI bots being used for cheating.<ref>{{cite web |last1=Bharade |first1=Aditi |title=ChatGPT failed miserably in Singapore's 6th-grade tests, averaging 16% for math and 21% for science. Days later, it was getting answers right. |url=https://www.businessinsider.com/chatgpt-failed-singapore-sixth-grade-exams-psle-2023-2 |website=Business Insider |access-date=4 March 2023}}</ref>
 
|-
 
|-
| 2023 || February 21 || Study || A paper evaluates the capabilities of ChatGPT on 25 diverse analytical {{w|natural language processing}} tasks, including sentiment analysis, emotion recognition, offensiveness and stance detection, natural language inference, word sense disambiguation, linguistic acceptability, and question answering. The study found that ChatGPT's average loss in quality was about 25% for zero-shot and few-shot evaluation, and the more difficult the task, the higher the ChatGPT loss. The paper also discusses ChatGPT's bias and the need for validation procedures for such systems.<ref>{{cite journal |last1=Kocoń |first1=Jan |last2=Cichecki |first2=Igor |last3=Kaszyca |first3=Oliwier |last4=Kochanek |first4=Mateusz |last5=Szydło |first5=Dominika |last6=Baran |first6=Joanna |last7=Bielaniewicz |first7=Julita |last8=Gruza |first8=Marcin |last9=Janz |first9=Arkadiusz |last10=Kanclerz |first10=Kamil |last11=Kocoń |first11=Anna |last12=Koptyra |first12=Bartłomiej |last13=Mieleszczenko-Kowszewicz |first13=Wiktoria |last14=Miłkowski |first14=Piotr |last15=Oleksy |first15=Marcin |last16=Piasecki |first16=Maciej |last17=Radliński |first17=Łukasz |last18=Wojtasik |first18=Konrad |last19=Woźniak |first19=Stanisław |last20=Kazienko |first20=Przemysław |title=ChatGPT: Jack of all trades, master of none |journal=arXiv:2302.10724 [cs] |date=21 February 2023 |doi=10.48550/arXiv.2302.10724 |url=https://arxiv.org/abs/2302.10724 |access-date=7 March 2023}}</ref>
+
| 2023 || February 27 || International baccalaureate allows ChatGPT content in essays || Adoption || The {{w|International Baccalaureate}} says that students can use content created by ChatGPT in their essays, but they must make it clear when they are [[w:quotation|quoting]] its responses and treat it as they would any other source. ChatGPT has the ability to produce plausible responses to text prompts, including requests to write essays. The IB, which is taken by thousands of children every year in the UK at more than 120 schools, says that the rise of {{w|chatbot}} technology would mean that essay writing will feature less prominently in future qualifications.<ref>{{cite web |last1=Milmo |first1=Dan |last2=editor |first2=Dan Milmo Global technology |title=ChatGPT allowed in International Baccalaureate essays |url=https://www.theguardian.com/technology/2023/feb/27/chatgpt-allowed-international-baccalaureate-essays-chatbot |website=The Guardian |access-date=3 March 2023 |date=27 February 2023}}</ref>
 
|-
 
|-
| 2023 || February 22 || Study || A paper evaluates the robustness of ChatGPT from the adversarial and out-of-distribution perspective. The authors employ AdvGLUE and ANLI benchmarks to assess adversarial robustness and {{w|Flipkart}} review and DDXPlus medical diagnosis datasets for out-of-distribution (OOD) evaluation. ChatGPT shows consistent advantages on most adversarial and OOD classification and translation tasks, but the absolute performance is still far from perfection, indicating that robustness remains a significant threat. ChatGPT shows astounding performance in understanding dialogue-related texts and tends to provide informal suggestions for medical tasks. The paper concludes with in-depth discussions of possible research directions.<ref>{{cite journal |last1=Wang |first1=Jindong |last2=Hu |first2=Xixu |last3=Hou |first3=Wenxin |last4=Chen |first4=Hao |last5=Zheng |first5=Runkai |last6=Wang |first6=Yidong |last7=Yang |first7=Linyi |last8=Huang |first8=Haojun |last9=Ye |first9=Wei |last10=Geng |first10=Xiubo |last11=Jiao |first11=Binxin |last12=Zhang |first12=Yue |last13=Xie |first13=Xing |title=On the Robustness of ChatGPT: An Adversarial and Out-of-distribution Perspective |journal=arXiv:2302.12095 [cs] |date=2 March 2023 |doi=10.48550/arXiv.2302.12095 |url=https://arxiv.org/abs/2302.12095 |access-date=7 March 2023}}</ref>
+
| 2023 || March 1 || Warning against humanizing chatbots || Impact || {{w|Computational linguist}} Emily M. Bender at the {{w|University of Washington}}, warns about the dangers of humanizing language models like chatbots, stating that the machines are great at mimicry but poor at facts. Bender argues that the big question is how humans will handle themselves around these machines, as we have learned to make machines that can mindlessly generate text but have not learned how to stop imagining the mind behind it.<ref>{{cite web |last1=Weil |first1=Elizabeth |title=You Are Not a Parrot |url=https://nymag.com/intelligencer/article/ai-artificial-intelligence-chatbots-emily-m-bender.html |website=Intelligencer |access-date=4 March 2023 |language=en-us |date=1 March 2023}}</ref><ref>{{cite web |last1=Weil |first1=Elizabeth |title=You Are Not a Parrot |url=https://nymag.com/intelligencer/article/ai-artificial-intelligence-chatbots-emily-m-bender.html |website=Intelligencer |access-date=24 May 2023 |language=en-us |date=1 March 2023}}</ref>
 
|-
 
|-
| 2023 || February 23 || Study || A paper discusses the impact of prompt knowledge on the correctness of answers generated by ChatGPT in the context of consumers seeking health advice from the model. The authors compare the effectiveness of ChatGPT in generating answers when it leverages the model's knowledge alone versus in combination with prompt knowledge. They show that the knowledge provided in the prompt can overturn the knowledge encoded in the model, leading to lower answer correctness. The study has implications for the development of more robust and transparent question-answering systems based on generative pre-trained language models.<ref>{{cite journal |last1=Zuccon |first1=Guido |last2=Koopman |first2=Bevan |title=Dr ChatGPT, tell me what I want to hear: How prompt knowledge impacts health answer correctness |journal=arXiv:2302.13793 [cs] |date=23 February 2023 |doi=10.48550/arXiv.2302.13793 |url=https://arxiv.org/abs/2302.13793 |access-date=6 March 2023}}</ref>
+
| 2023 || March 1 || Generic evaluation framework for LLMs on personality assessment || Performance || A paper proposes a generic evaluation framework for Large Language Models (LLMs), specifically ChatGPT and InstructGPT, to assess human personalities based on Myers Briggs Type Indicator (MBTI) tests. The framework includes unbiased prompts, replacing the subject in question statements, and re-formulating question instructions. The study finds that ChatGPT is capable of assessing human personalities and achieves consistent and fairer assessments compared to InstructGPT, albeit with lower robustness against prompt biases. The proposed framework enables flexible assessments of personalities for different groups of people.<ref>{{cite journal |last1=Rao |first1=Haocong |last2=Leung |first2=Cyril |last3=Miao |first3=Chunyan |title=Can ChatGPT Assess Human Personalities? A General Evaluation Framework |date=2023 |doi=10.48550/arXiv.2303.01248}}</ref>
 
|-
 
|-
| 2023 || February 23 || Study || A paper presents the first independent evaluation of ChatGPT on [[w:Word problem (mathematics education)|mathematical word problems]] (MWPs) from the dataset DRAW-1K. The authors found that ChatGPT's performance varies dramatically based on the requirement to show its work, with a failure rate of 20% when it provides work and 84% when it does not. The paper also highlights several factors relating to the number of unknowns and operations that increase the probability of failure, with the probability of failure increasing linearly with the number of addition and subtraction operations.<ref>{{cite journal |last1=Shakarian |first1=Paulo |last2=Koyyalamudi |first2=Abhinav |last3=Ngu |first3=Noel |last4=Mareedu |first4=Lakshmivihari |title=An Independent Evaluation of ChatGPT on Mathematical Word Problems (MWP) |journal=arXiv:2302.13814 [cs] |date=27 February 2023 |doi=10.48550/arXiv.2302.13814 |url=https://arxiv.org/abs/2302.13814}}</ref>
+
| 2023 || March 1 || OpenAI launches ChatGPT API for business integration || Product launch || {{w|OpenAI}} launches an {{w|API}} for ChatGPT, enabling businesses to integrate this advanced technology into their applications, websites, and services. Priced at $0.002 per 1,000 tokens (equivalent to about 750 words), the API supports various "non-chat" applications. Powered by the robust AI model "gpt-3.5-turbo," the ChatGPT API offers versatile applications such as AI-powered tutoring, personalized shopping recommendations, and virtual assistance. This move to make ChatGPT more affordable attracts interest from companies like {{w|Microsoft}}, {{w|Shopify}}, {{w|Snapchat}}, and {{w|Instacart}}, highlighting its potential for chatbot development and language processing tasks.<ref>{{cite web |last1=Wiggers |first1=Kyle |title=OpenAI launches an API for ChatGPT, plus dedicated capacity for enterprise customers |url=https://techcrunch.com/2023/03/01/openai-launches-an-api-for-chatgpt-plus-dedicated-capacity-for-enterprise-customers/ |website=TechCrunch |access-date=4 March 2023 |date=1 March 2023}}</ref><ref>{{cite web |title=Introducing ChatGPT and Whisper APIs |url=https://openai.com/blog/introducing-chatgpt-and-whisper-apis |website=openai.com |access-date=16 March 2023}}</ref><ref>{{cite web |title=OpenAI announces ChatGPT API, Snapchat, Instacart and others already using it |url=https://www.gsmarena.com/openai_announces_chatgpt_api_snapchat_instacart_and_others_are_already_using_it-news-57751.php |website=GSMArena.com |access-date=3 March 2023}}</ref>
 
|-
 
|-
| 2023 || February 26 || || A paper proposes the use of Software Development Bots (DevBots) trained on large language models to aid in software architecture-centric software engineering (ACSE). The paper specifically focuses on the use of ChatGPT, a natural language processing tool, for collaboration between a novice software architect and the AI model in analyzing, synthesizing, and evaluating a services-driven software application. Preliminary results indicate that ChatGPT can mimic an architect's role but requires human oversight and decision support for collaborative architecting. The authors suggest further research into harnessing empirical evidence and exploring socio-technical aspects of architecting with ChatGPT.<ref>{{cite journal |last1=Ahmad |first1=Aakash |last2=Waseem |first2=Muhammad |last3=Liang |first3=Peng |last4=Fehmideh |first4=Mahdi |last5=Aktar |first5=Mst Shamima |last6=Mikkonen |first6=Tommi |title=Towards Human-Bot Collaborative Software Architecting with ChatGPT |journal=arXiv:2302.14600 [cs] |date=26 February 2023 |doi=10.48550/arXiv.2302.14600 |url=https://arxiv.org/abs/2302.14600}}</ref>
+
| 2023 || March 2 || ChatGPT's performance on physics conceptual questions analyzed || Performance || A paper presents an analysis of ChatGPT's ability to answer conceptual physics questions related to {{w|kinematics}} and {{w|Newtonian dynamics}} using the {{w|Force Concept Inventory}} (FCI). The study shows that ChatGPT can perform as well as or better than the median performance of a university student who has completed one semester of college physics, although its performance is uneven. The paper raises questions about the use of conceptual assessment tools like the FCI, the implications of ChatGPT's performance for students and physics instruction, and the potential use of ChatGPT as an out-of-classroom aid for physics pedagogy.<ref>{{cite journal |last1=West |first1=Colin G. |title=AI and the FCI: Can ChatGPT Project an Understanding of Introductory Physics? |date=2023 |doi=10.48550/arXiv.2303.01067}}</ref>
 
|-
 
|-
| 2023 || February 27 || Study || A paper provides an overview of the technology behind ChatGPT. It discusses the historical background of chatbots and the potential applications of ChatGPT in various domains such as healthcare, education, and research. However, the paper also highlights the privacy and ethical concerns surrounding ChatGPT and the limitations of the current version. The paper also includes an interview with ChatGPT and presents its responses to various questions.<ref>{{cite journal |last1=Shahriar |first1=Sakib |last2=Hayawi |first2=Kadhim |title=Let's have a chat! A Conversation with ChatGPT: Technology, Applications, and Limitations |date=2023 |doi=10.48550/arXiv.2302.13817 |url=https://arxiv.org/abs/2302.13817}}</ref>
+
| 2023 || March 2 || Assessing AI language models' impact on occupations and industries || Impact || A paper presents a methodology to assess the impact of AI language models like ChatGPT on different occupations, industries, and geographies. The authors find that occupations like telemarketers and post-secondary teachers in subjects such as English language, literature, foreign language, and history are the most exposed to language modeling. The industries most exposed to advances in language modeling are legal services and securities, commodities, and investments.<ref>{{cite journal |last1=Felten |first1=Ed |last2=Raj |first2=Manav |last3=Seamans |first3=Robert |title=How will Language Modelers like ChatGPT Affect Occupations and Industries? |date=2023 |doi=10.48550/arXiv.2303.01157}}</ref>
 
|-
 
|-
| 2023 || February 28 || Study || A paper presents an empirical study that investigates the performance of ChatGPT in cross-lingual summarization (CLS), which involves generating a summary in a different target language from a source language document. The authors use various prompts to guide ChatGPT to perform zero-shot CLS from different paradigms and provide a preliminary evaluation of its generated summaries. The study finds that ChatGPT initially prefers to produce lengthy summaries with more detailed information, but with the help of interactive prompts, it can significantly improve its CLS performance by balancing informativeness and conciseness. Experimental results on three widely-used CLS datasets show that ChatGPT outperforms the advanced GPT 3.5 model. The paper provides qualitative case studies to demonstrate the superiority of ChatGPT on CLS.<ref>{{cite journal |last1=Wang |first1=Jiaan |last2=Liang |first2=Yunlong |last3=Meng |first3=Fandong |last4=Li |first4=Zhixu |last5=Qu |first5=Jianfeng |last6=Zhou |first6=Jie |title=Cross-Lingual Summarization via ChatGPT |journal=arXiv:2302.14229 [cs] |date=27 February 2023 |url=https://arxiv.org/abs/2302.14229 |access-date=6 March 2023}}</ref>
+
| 2023 || March 3 || Microsoft introduces Kosmos-1, a multimodal AI model with AGI potential || Competition || Microsoft unveils a new AI model, Kosmos-1, which can respond to visual cues or images and is capable of tasks such as visual question answering and image captioning. The model is a Multimodal Large Language Model (MLLM), which allows the user to interact with it through text prompts, images, and videos. Kosmos-1 is seen as a key step toward artificial general intelligence, with researchers suggesting that multimodal perception is required to move beyond ChatGPT-like capabilities to AGI.<ref>{{cite web |last1=Nigam |first1=Saumya |last2=News |first2=India TV |title=Microsoft unveils Kosmos-1,a new AI model to race up with ChatGPT |url=https://www.indiatvnews.com/technology/news/microsoft-unveils-kosmos-1-a-new-ai-model-to-race-up-with-chatgpt-2023-03-03-851902 |website=www.indiatvnews.com |access-date=4 March 2023 |language=en |date=3 March 2023}}</ref>
 
|-
 
|-
| 2023 || February || || A Colombian judge reportedly uses ChatGPT to help rule on a case about the medical rights of a child with autism. The case is about whether the child's insurance should cover the bills for his medical treatments. The judge uses ChatGPT in accordance with a [[w:Judiciary of Colombia|law in Colombia]] that allows the use of virtual tools in some cases. Opinions are divided, with some professionals opposing the use of AI in a legal setting, while others see it as a useful tool.<ref name="Bhaimiya">{{cite web |last1=Bhaimiya |first1=Sawdah |title=A Colombian judge used ChatGPT to help rule on a case about the medical rights of a child with autism, reports say |url=https://www.businessinsider.com/chatgpt-colombian-judge-used-in-ruling-child-medical-rights-report-2023-2 |website=Business Insider |access-date=4 March 2023}}</ref>  
+
| 2023 || March 7 || Feasibility of ChatGPT for event extraction || Performance || A paper discusses the feasibility of using ChatGPT, which is a challenging task due to the lack of annotated data. While by this time ChatGPT demonstrates impressive results in other tasks, it presents challenges for event extraction, which requires complex instructions defining all event types and their schemas. The authors conduct experiments to explore the feasibility of using ChatGPT for event extraction and find that it has only 51.04% of the performance of a task-specific model in complex scenarios. The paper concludes that model is not robust enough and is highly sensitive to different prompt styles, which can result in a poor user experience.<ref>{{cite journal |last1=Gao |first1=Jun |last2=Zhao |first2=Huan |last3=Yu |first3=Changlong |last4=Xu |first4=Ruifeng |title=Exploring the Feasibility of ChatGPT for Event Extraction |journal=arXiv.org |date=7 March 2023 |doi=10.48550/arXiv.2303.03836 |url=https://arxiv.org/abs/2303.03836 |language=en}}</ref>
 
|-
 
|-
| 2023 || February || Adoption || Small businesses in the {{w|United Arab Emirates}} turn to ChatGPT to reduce the amount of time spent on mundane tasks such as writing job descriptions and blog posts, thereby allowing them to concentrate on more creative aspects of their business. By this time, ChatGPT is used by large companies including Meta and Shopify. The Wall Street Journal speculates that the tool could eventually lead to job losses in some areas, although ChatGPT's flaws, including providing inaccurate responses, and the need for further regulation of AI technology so far prevent this for the time being.<ref>{{cite web |title=UAE small businesses embrace ChatGPT as tech allows productivity boost |url=https://english.alarabiya.net/News/gulf/2023/02/15/UAE-small-businesses-embrace-ChatGPT-as-tech-allows-productivity-boost |website=Al Arabiya English |access-date=4 March 2023 |language=en |date=15 February 2023}}</ref>
+
| 2023 || March 9 || {{w|Stephen Wolfram}} publishes book on ChatGPT || Literature || British-American computer scientist {{w|Stephen Wolfram}} publishes ''What Is ChatGPT Doing ... and Why Does It Work?'', in which he explores ChatGPT's unexpected ability to write convincingly at a human level. Wolfram delves into the inner workings of ChatGPT, bridging modern neural net technology with ancient questions about language and human cognition, as posed by Aristotle over two millennia ago. The book offers insights into the AI's success and sheds light on the convergence of AI technology and age-old inquiries into human thought and language.<ref>{{cite book |last1=Wolfram |first1=Stephen |title=What Is ChatGPT Doing ... and Why Does It Work? |date=2023 |publisher=Wolfram Media, Incorporated |isbn=978-1-57955-081-3 |url=https://books.google.com.ar/books/about/What_Is_ChatGPT_Doing_and_Why_Does_It_Wo.html?id=1UOuzwEACAAJ&source=kp_book_description&redir_esc=y |language=en}}</ref><ref>{{cite web |title=What Is ChatGPT Doing ... and Why Does It Work? |url=https://www.wolfram-media.com/products/what-is-chatgpt-doing-and-why-does-it-work/ |website=www.wolfram-media.com |access-date=13 September 2023 |language=en}}</ref>
 
|-
 
|-
| 2023 || February || || ChatGPT performs poorly when asked to answer questions from Singapore's Primary School Leaving Exam (PSLE), with an average score of 16 out of 100 marks for mathematics and 21 for science. It is found to make errors even in simple addition and can not understand or answer any questions that reference diagrams or graphs. However, ChatGPT manages to pass the English tests and scores an average of 11 out of 20 marks across the three papers it took. The bot's poor performance in exams raises concerns about the possibility of AI bots being used for cheating.<ref>{{cite web |last1=Bharade |first1=Aditi |title=ChatGPT failed miserably in Singapore's 6th-grade tests, averaging 16% for math and 21% for science. Days later, it was getting answers right. |url=https://www.businessinsider.com/chatgpt-failed-singapore-sixth-grade-exams-psle-2023-2 |website=Business Insider |access-date=4 March 2023}}</ref>
+
| 2023 || March 12 || OpenChatKit 0.15, open-source chatbot alternative to ChatGPT || Competition || {{w|California}}-based {{w|API}}-provider Together Computer releases OpenChatKit 0.15, the first open-source chatbot alternative to ChatGPT, under the [[w:Apache License|Apache-2.0 license]]. The chat model has 20 billion parameters and was trained on 43 million instructions, and includes a generic chatbot and the necessary components to create specialized bots. The framework has four main parts, including a {{w|large language model}} tuned for chat, a set of customization recipes, a retrieval system, and a moderation model. The open-source foundation offered by OpenChatKit enables the development of domain-specific and general-purpose chatbots.<ref>{{cite web |last1=Shenwai |first1=Tanushree |title=Together Releases The First Open-Source ChatGPT Alternative Called OpenChatKit |url=https://www.marktechpost.com/2023/03/12/together-releases-the-first-open-source-chatgpt-alternative-called-openchatkit/ |website=MarkTechPost |access-date=14 March 2023 |date=12 March 2023}}</ref>  
 
|-
 
|-
| 2023 || February 25 || || A paper introduces ChatAug, a text data augmentation approach based on ChatGPT, to mitigate the challenge of limited sample sizes in natural language processing tasks, particularly in few-shot learning scenarios. The proposed approach rephrases each sentence in the training samples into multiple conceptually similar but semantically different samples to ensure the faithfulness and completeness of the generated data. The experiment results demonstrate that ChatAug outperforms state-of-the-art text data augmentation methods in terms of testing accuracy and distribution of the augmented samples.<ref>{{cite journal |last1=Dai |first1=Haixing |last2=Liu |first2=Zhengliang |last3=Liao |first3=Wenxiong |last4=Huang |first4=Xiaoke |last5=Wu |first5=Zihao |last6=Zhao |first6=Lin |last7=Liu |first7=Wei |last8=Liu |first8=Ninghao |last9=Li |first9=Sheng |last10=Zhu |first10=Dajiang |last11=Cai |first11=Hongmin |last12=Li |first12=Quanzheng |last13=Shen |first13=Dinggang |last14=Liu |first14=Tianming |last15=Li |first15=Xiang |title=ChatAug: Leveraging ChatGPT for Text Data Augmentation |journal=arXiv:2302.13007 [cs] |date=27 February 2023 |doi=10.48550/arXiv.2302.13007 |url=https://arxiv.org/abs/2302.13007 |access-date=6 March 2023}}</ref>
+
| 2023 || March 14 || GPT-4 introduction || Update || OpenAI announces GPT-4, the latest AI language model in its line of language models. GPT-4 can parse both text and image input but can only respond via text. The company cautions that GPT-4 retains some of the same problems as earlier models, such as the tendency to make up information and generate harmful text. By this time, OpenAI has partnered with companies including Duolingo, Stripe, and Khan Academy to integrate GPT-4 into their products, and it would also be available as an API for developers to build on. GPT-4 is available to the general public via OpenAI's ChatGPT Plus subscription. The company claims that GPT-4 is more creative and collaborative than ever before and can solve difficult problems with greater accuracy.<ref>{{cite web |last1=Vincent |first1=James |title=OpenAI announces GPT-4 — the next generation of its AI language model |url=https://www.theverge.com/2023/3/14/23638033/openai-gpt-4-chatgpt-multimodal-deep-learning |website=The Verge |access-date=14 March 2023 |date=14 March 2023}}</ref> ChatGPT would undergo an update utilizing the GPT-4 language model, which would expand its vocabulary to 25,000 words (eight times more than its previous version). In addition, the update would endow ChatGPT with the capacity to process image inputs as well as text, effectively making it a multimodal system.<ref name="tooltester"/>
 
|-
 
|-
| 2023 || February 27 || || The {{w|International Baccalaureate}} says that students can use content created by ChatGPT in their essays, but they must make it clear when they are [[w:quotation|quoting]] its responses and treat it as they would any other source. ChatGPT has the ability to produce plausible responses to text prompts, including requests to write essays. The IB, which is taken by thousands of children every year in the UK at more than 120 schools, says that the rise of {{w|chatbot}} technology would mean that essay writing will feature less prominently in future qualifications.<ref>{{cite web |last1=Milmo |first1=Dan |last2=editor |first2=Dan Milmo Global technology |title=ChatGPT allowed in International Baccalaureate essays |url=https://www.theguardian.com/technology/2023/feb/27/chatgpt-allowed-international-baccalaureate-essays-chatbot |website=The Guardian |access-date=3 March 2023 |date=27 February 2023}}</ref>
+
| 2023 || March 15 || ChatGPT adds Icelandic language support || Expansion || OpenAI announces an update for ChatGPT that will allow it to communicate in {{w|Icelandic}}, making it the chatbot's second language. By this time, the language model GPT-4 is used to teach the program Icelandic. Previously, the Icelandic envoy, led by the president of Iceland, visited OpenAI in 2022 to request the addition of Icelandic. The Icelandic language technology center CEO calls the project a great victory for the Icelandic language.<ref>{{cite web |last1=Magnússon |first1=Pétur |title=Icelandic becomes ChatGPT's second language - RÚV.is |url=https://nyr.ruv.is/english/2023-03-15-icelandic-becomes-chatgpts-second-language |website=RÚV |access-date=7 May 2023 |date=15 March 2023}}</ref>
 
|-
 
|-
| 2023 || March 1 || || Emily M. Bender, a computational linguist at the University of Washington, warns about the dangers of humanizing language models like chatbots, stating that the machines are great at mimicry but poor at facts. Bender argues that the big question is how humans will handle themselves around these machines, as we have learned to make machines that can mindlessly generate text but have not learned how to stop imagining the mind behind it.<ref>{{cite web |last1=Weil |first1=Elizabeth |title=You Are Not a Parrot |url=https://nymag.com/intelligencer/article/ai-artificial-intelligence-chatbots-emily-m-bender.html |website=Intelligencer |access-date=4 March 2023 |language=en-us |date=1 March 2023}}</ref>
+
| 2023 || March 16 || OpenAI faces criticism for lack of transparency with GPT-4 || Criticism || An article reports on  OpenAI facing criticism for not disclosing key details about its latest AI model, GPT-4. The company's decision to keep information about the model's architecture, hardware, and training methods private is attributed to the competitive landscape and safety implications surrounding large language models. By this time, the AI community has expressed concerns over OpenAI's lack of transparency, with some questioning the "open" nature of the company. OpenAI co-founder Ilya Sutskever admits that sharing research publicly was a mistake and emphasizes the need for caution in open-sourcing AI.<ref>{{cite web |last1=Gowran |first1=Leigh Mc |title=OpenAI criticised for lack of transparency around GPT-4 |url=https://www.siliconrepublic.com/machines/openai-gpt4-transparency-ai-concerns-stripe-chatgpt |website=Silicon Republic |access-date=24 May 2023 |language=en |date=16 March 2023}}</ref>  
 
|-
 
|-
| 2023 || March 1 || Study || A paper proposes a generic evaluation framework for Large Language Models (LLMs), specifically ChatGPT and InstructGPT, to assess human personalities based on Myers Briggs Type Indicator (MBTI) tests. The framework includes unbiased prompts, replacing the subject in question statements, and re-formulating question instructions. The study finds that ChatGPT is capable of assessing human personalities and achieves consistent and fairer assessments compared to InstructGPT, albeit with lower robustness against prompt biases. The proposed framework enables flexible assessments of personalities for different groups of people.<ref>{{cite journal |last1=Rao |first1=Haocong |last2=Leung |first2=Cyril |last3=Miao |first3=Chunyan |title=Can ChatGPT Assess Human Personalities? A General Evaluation Framework |date=2023 |doi=10.48550/arXiv.2303.01248}}</ref>
+
| 2023 || March 17 || Baidu's Ernie Bot sparks 14% stock surge amid positive user feedback || Competition || Shares of Chinese search engine giant {{w|Baidu}} rise by almost 14% after users test and share positive experiences with Ernie Bot, Baidu's AI-powered chatbot designed to rival ChatGPT. The initial disappointment from investors due to a scripted and private launch is overshadowed by users livestreaming tests and comparing Ernie Bot favorably to US chatbots. Baidu's deep learning model, Ernie, serves as the foundation for the chatbot, and the company received over 75,000 applications for an Ernie API trial. Baidu reportedly aims to revolutionize its search engine and enhance efficiency in various areas, including cloud, smart cars, and household appliances, through Ernie Bot.<ref>{{cite web |last1=Saxena |first1=Vishakha |title=Baidu Shares Jump 14% as Users Test ChatGPT-Like Ernie Bot |url=https://www.asiafinancial.com/baidu-shares-jump-14-as-users-test-chatgpt-like-ernie-bot |website=Asia Financial |access-date=24 May 2023 |date=17 March 2023}}</ref>  
 
|-
 
|-
| 2023 || March 1 || || {{w|OpenAI}} launches an {{w|API}} for ChatGPT. The new product allows any business to build the technology into their apps, websites, products and services. Priced at $0.002 per 1,000 tokens, or about 750 words, the API also supports "non-chat" applications. The ChatGPT API is powered by the same AI model behind ChatGPT, dubbed “gpt-3.5-turbo," the most powerful text-generating model OpenAI offers to date through its API suite. It's aimed at driving a range of experiences including an AI-powered tutor, a personalized assistant for shopping recommendations and a virtual tutor feature.<ref>{{cite web |last1=Wiggers |first1=Kyle |title=OpenAI launches an API for ChatGPT, plus dedicated capacity for enterprise customers |url=https://techcrunch.com/2023/03/01/openai-launches-an-api-for-chatgpt-plus-dedicated-capacity-for-enterprise-customers/ |website=TechCrunch |access-date=4 March 2023 |date=1 March 2023}}</ref><ref>{{cite web |title=Introducing ChatGPT and Whisper APIs |url=https://openai.com/blog/introducing-chatgpt-and-whisper-apis |website=openai.com |access-date=16 March 2023}}</ref>
+
| 2023 || March 20 || ChatGPT outage and data exposure || System behavior || ChatGPT experiences an [[w:downtime|outage]] due to a [[w:Software bug|bug]] in an open-source library which allows some users to see titles from another active user’s chat history. The bug is patched, and the ChatGPT service is restored with the exception of a few hours of chat history. Upon deeper investigation, OpenAI discovers that the same bug may have caused the unintentional visibility of payment-related information of 1.2% of ChatGPT Plus subscribers active during a specific nine-hour window. OpenAI reaches out to notify affected users that their payment information may have been exposed and is committed to rebuilding trust with the ChatGPT community.<ref>{{cite web |title=March 20 ChatGPT outage: Here’s what happened |url=https://openai.com/blog/march-20-chatgpt-outage |website=openai.com |access-date=24 March 2023}}</ref>
 
|-
 
|-
| 2023 || March 2 || || OpenAI announces that its ChatGPT and Whisper models are now available to developers through its APIs. The models can be used for building chatbots and are integrated by companies like Microsoft, Shopify, Snapchat, and Instacart. By this time, ChatGPT costs 90% less than it did in December, making it cheaper for companies to build their chatbots. Meanwhile, Whisper can transcribe speech into text and understand 99 languages, offering potential disruption to Siri and Google Assistant.<ref>{{cite web |title=OpenAI announces ChatGPT API, Snapchat, Instacart and others already using it |url=https://www.gsmarena.com/openai_announces_chatgpt_api_snapchat_instacart_and_others_are_already_using_it-news-57751.php |website=GSMArena.com |access-date=3 March 2023}}</ref>
+
| 2023 || March 21 || Google launches public release of Bard Chatbot || Competition || Google starts the public release of its chatbot [[w:Bard (chatbot)|Bard]], allowing consumers to sign up to a waiting list for English-language access in the US and UK. The program was previously open to approved testers. Google describes Bard as an experiment in generative AI, which creates rather than identifies content. Bard, unlike ChatGPT, does not generate computer code and its memory is limited. It also includes a feature to show three versions of an answer.<ref>{{cite web |title=Google begins opening access to its ChatGPT competitor Bard |url=https://finance.yahoo.com/news/google-begins-opening-access-chatgpt-140149232.html |website=Yahoo Finance |access-date=21 March 2023}}</ref>
 
|-
 
|-
| 2023 || March 2 || Study || A paper presents an analysis of ChatGPT's ability to answer conceptual physics questions related to {{w|kinematics}} and {{w|Newtonian dynamics}} using the {{w|Force Concept Inventory}} (FCI). The study shows that ChatGPT can perform as well as or better than the median performance of a university student who has completed one semester of college physics, although its performance is uneven. The paper raises questions about the use of conceptual assessment tools like the FCI, the implications of ChatGPT's performance for students and physics instruction, and the potential use of ChatGPT as an out-of-classroom aid for physics pedagogy.<ref>{{cite journal |last1=West |first1=Colin G. |title=AI and the FCI: Can ChatGPT Project an Understanding of Introductory Physics? |date=2023 |doi=10.48550/arXiv.2303.01067}}</ref>
+
| 2023 || March 23 || Plugin support is introduced for enhanced functionality || Update || OpenAI announces ChatGPT implementation support for [[w:Plug-in (computing)|plugins]], which are tools designed for language models to access up-to-date information, run computations, or use third-party services with safety as a core principle. The feature is at first rolled out gradually for a small set of users and is expected to be expanded as more is learned about safety and alignment challenges. Plugin developers can use the provided documentation to build plugins, and the first plugins are created by companies such as {{w|Expedia}}, [[w:Slack Technologies|Slack]], {{w|Wolfram Research}}, and {{w|Zapier}}.<ref>{{cite web |title=ChatGPT plugins |url=https://openai.com/blog/chatgpt-plugins |website=openai.com |access-date=24 March 2023}}</ref>
 
|-
 
|-
| 2023 || March 2 || Study || A paper presents a methodology to assess the impact of AI language models like ChatGPT on different occupations, industries, and geographies. The authors find that occupations like telemarketers and post-secondary teachers in subjects such as English language, literature, foreign language, and history are the most exposed to language modeling. The industries most exposed to advances in language modeling are legal services and securities, commodities, and investments.<ref>{{cite journal |last1=Felten |first1=Ed |last2=Raj |first2=Manav |last3=Seamans |first3=Robert |title=How will Language Modelers like ChatGPT Affect Occupations and Industries? |date=2023 |doi=10.48550/arXiv.2303.01157}}</ref>
+
| 2023 || March 28 || Daniel Everett challenges Chomsky's theory in light of ChatGPT || Impact || In an interview, American linguist {{w|Daniel Everett}} challenges {{w|Noam Chomsky}}'s theory of "innate principles of language" and argues that ChatGPT has invalidated Chomsky's claims. Everett highlights the importance of considering social and cultural origins in language acquisition and suggests that current theories need revision. He believes that language and culture have a symbiotic relationship. Everett states that ChatGPT, with massive data, has shown the ability to learn language without innate grammar principles. He emphasizes the need for responsible and ethical AI development, with linguists and researchers playing a role in ensuring a balanced approach.<ref>{{cite web |title=Exclusive: Linguist says ChatGPT has invalidated Chomsky's ‘innate principles of language’ |url=https://www.tehrantimes.com/news/483187/Exclusive-Linguist-says-ChatGPT-has-invalidated-Chomsky-s-innate |website=Tehran Times |access-date=24 May 2023 |language=en |date=27 March 2023}}</ref>
 
|-
 
|-
| 2023 || March 3 || Competition || Microsoft unveils a new AI model, Kosmos-1, which can respond to visual cues or images and is capable of tasks such as visual question answering and image captioning. The model is a Multimodal Large Language Model (MLLM), which allows the user to interact with it through text prompts, images, and videos. Kosmos-1 is seen as a key step toward artificial general intelligence, with researchers suggesting that multimodal perception is required to move beyond ChatGPT-like capabilities to AGI.<ref>{{cite web |last1=Nigam |first1=Saumya |last2=News |first2=India TV |title=Microsoft unveils Kosmos-1,a new AI model to race up with ChatGPT |url=https://www.indiatvnews.com/technology/news/microsoft-unveils-kosmos-1-a-new-ai-model-to-race-up-with-chatgpt-2023-03-03-851902 |website=www.indiatvnews.com |access-date=4 March 2023 |language=en |date=3 March 2023}}</ref>
+
| 2023 || March 31 || Ban in Italy || Restriction || ChatGPT is subject to a temporary ban in {{w|Italy}}, due to suspicions that it may have breached the {{w|General Data Protection Regulation}}s (GDPR) of the {{w|European Union}}.<ref>{{cite web |title=Mastercard launches Web3 user verification solution to curb bad actors |url=https://cointelegraph.com/news/mastercard-launches-web3-user-verification-solution-to-curb-bad-actors |website=Cointelegraph |access-date=8 May 2023 |language=en}}</ref>
 
|-
 
|-
| 2023 || March 7 || Review || A paper discusses the feasibility of using ChatGPT, which is a challenging task due to the lack of annotated data. While by this time ChatGPT demonstrated impressive results in other tasks, it presents challenges for event extraction, which requires complex instructions defining all event types and their schemas. The authors conducted experiments to explore the feasibility of using ChatGPT for event extraction and found that it has only 51.04% of the performance of a task-specific model in complex scenarios. The model is not robust enough and is highly sensitive to different prompt styles, which can result in a poor user experience.<ref>{{cite journal |last1=Gao |first1=Jun |last2=Zhao |first2=Huan |last3=Yu |first3=Changlong |last4=Xu |first4=Ruifeng |title=Exploring the Feasibility of ChatGPT for Event Extraction |journal=arXiv.org |date=7 March 2023 |doi=10.48550/arXiv.2303.03836 |url=https://arxiv.org/abs/2303.03836 |language=en}}</ref>
+
| 2023 || April 13 || Privacy concerns arise over ChatGPT's data collection and usage || Privacy || An article  explores the privacy concerns surrounding ChatGPT. The chatbot collects various types of information, including account details, chat transcripts, and device information such as IP addresses and locations. According to the article, while some of this data collection is standard for online services, the real privacy risk lies in the potential disclosure of personal and confidential information inadvertently shared during conversations. OpenAI's privacy policy states that conversations may be reviewed by AI trainers, raising concerns about data usage and security. The article emphasizes the importance of users taking proactive measures to protect their privacy by refraining from sharing private information and carefully reviewing their interactions with ChatGPT.<ref name=privacy>{{cite web |title=Does ChatGPT Have Privacy Issues? |url=https://www.makeuseof.com/chatgpt-privacy-issues/ |website=MUO |access-date=24 May 2023 |language=en |date=13 April 2023}}</ref>
 
|-
 
|-
| 2023 || March 12 || Competition || Together releases OpenChatKit 0.15, the first open-source chatbot alternative to ChatGPT, under the Apache-2.0 license. The chat model has 20 billion parameters and was trained on 43 million instructions, and includes a generic chatbot and the necessary components to create specialized bots. The framework has four main parts, including a large language model tuned for chat, a set of customization recipes, a retrieval system, and a moderation model. The open-source foundation offered by OpenChatKit enables the development of domain-specific and general-purpose chatbots.<ref>{{cite web |last1=Shenwai |first1=Tanushree |title=Together Releases The First Open-Source ChatGPT Alternative Called OpenChatKit |url=https://www.marktechpost.com/2023/03/12/together-releases-the-first-open-source-chatgpt-alternative-called-openchatkit/ |website=MarkTechPost |access-date=14 March 2023 |date=12 March 2023}}</ref>  
+
| 2023 || April 17 || Elon Musk announces 'TruthGPT' || Competition || {{w|Elon Musk}} announces that he is working on "TruthGPT," an alternative to OpenAI's ChatGPT that would function as a "maximum truth-seeking AI." During an interview, Musk explains that he believes developing an AI that cares about understanding the universe is the best path to safety, as it is less likely to destroy humanity. He compares this to how humans protect chimpanzees, and positions "TruthGPT" as a more transparent option than OpenAI's for-profit subsidiary. At this time it is not clear how far along Musk's "TruthGPT" project is, but he previously tweeted about the need for such an AI and established a new AI company called X.AI in March.<ref>{{cite web |last1=Roth |first1=Emma |title=Elon Musk claims to be working on ‘TruthGPT’ — a ‘maximum truth-seeking AI’ |url=https://www.theverge.com/2023/4/17/23687440/elon-musk-truthgpt-ai-chatgpt |website=The Verge |access-date=9 May 2023 |date=18 April 2023}}</ref>  
 
|-
 
|-
| 2023 || March 14 || || OpenAI announces GPT-4, the latest AI language model in its line of language models. GPT-4 can parse both text and image input but can only respond via text. The company cautions that GPT-4 retains some of the same problems as earlier models, such as the tendency to make up information and generate harmful text. By this time, OpenAI has partnered with companies including Duolingo, Stripe, and Khan Academy to integrate GPT-4 into their products, and it would also be available as an API for developers to build on. GPT-4 is available to the general public via OpenAI's ChatGPT Plus subscription. The company claims that GPT-4 is more creative and collaborative than ever before and can solve difficult problems with greater accuracy.<ref>{{cite web |last1=Vincent |first1=James |title=OpenAI announces GPT-4 — the next generation of its AI language model |url=https://www.theverge.com/2023/3/14/23638033/openai-gpt-4-chatgpt-multimodal-deep-learning |website=The Verge |access-date=14 March 2023 |date=14 March 2023}}</ref> ChatGPT would undergo an update utilizing the GPT-4 language model, which would expand its vocabulary to 25,000 words (eight times more than its previous version). In addition, the update would endow ChatGPT with the capacity to process image inputs as well as text, effectively making it a multimodal system.<ref name="tooltester"/>
+
| 2023 || April 25 || OpenAI implements chat history privacy feature for ChatGPT users || Update || OpenAI introduces a new feature that allows users to turn off their chat history when using its ChatGPT chatbot, which means conversations won't be used to train its AI models. However, the company would still store new conversations for up to 30 days to monitor for abuse before deleting them. The feature can be accessed by logging into the ChatGPT account and selecting the "Settings" option. OpenAI also introduces a way for subscribers of its new ChatGPT Business plan to preserve chat history while opting out of its use as training data.<ref>{{cite web |last1=Roth |first1=Emma |title=OpenAI will let you turn off your chat history in ChatGPT |url=https://www.theverge.com/2023/4/25/23697942/openai-chatgpt-chat-history-data |website=The Verge |access-date=9 May 2023 |date=25 April 2023}}</ref><ref>{{cite web |title=New ways to manage your data in ChatGPT |url=https://openai.com/blog/new-ways-to-manage-your-data-in-chatgpt |website=openai.com |access-date=9 May 2023}}</ref>
 
|-
 
|-
| 2023 || March 16 || Interview || In an interview with ABC News’ Rebecca Jarvis, {{w|Sam Altman}} says that AI technology will reshape society as we know it, but that it comes with real dangers. He also says that feedback will help deter the potential negative consequences that the technology could have on humanity. Altman acknowledges the possible dangerous implementations of AI that keep him up at night, particularly the fear that AI models could be used for large-scale disinformation or offensive cyberattacks. He also says that he fears which humans could be in control of the technology. However, he does not share the sci-fi fear of AI models that do not need humans, stating that "This is a tool that is very much in human control".<ref>{{cite web |last1=News |first1=A. B. C. |title=OpenAI CEO Sam Altman says AI will reshape society, acknowledges risks: 'A little bit scared of this' |url=https://abcnews.go.com/Technology/openai-ceo-sam-altman-ai-reshape-society-acknowledges/story?id=97897122 |website=ABC News |access-date=23 March 2023 |language=en}}</ref>
+
| 2023 || April 30 || ChatGPT restored in Italy following privacy concerns and regulatory measures || Restriction || ChatGPT becomes accessible again in {{w|Italy}} after it was temporarily restricted and banned by the Italian data-protection authority over privacy concerns. OpenAI says it had successfully "addressed or clarified" the issues raised by the regulator, and has offered a new tool to verify the age of users upon sign-up. The Italian regulator welcomes the measures but calls for more compliance, including implementing an age verification system and conducting an information campaign to inform Italians of their rights.<ref>{{cite web |title=The ChatGPT is back in Italy: OpenAI is now going to guarantee more transparency and more rights for users - Game News 24 |url=https://game-news24.com/2023/04/30/the-chatgpt-is-back-in-italy-openai-is-now-going-to-guarantee-more-transparency-and-more-rights-for-users/ |website=game-news24.com |access-date=7 May 2023}}</ref> 
 +
|-
 +
| 2023 || May 18 || Article highlights security risks of ChatGPT adoption || Security || An article warns about the security risks associated with the increasing adoption of ChatGPT and highlights three ways hackers can exploit its capabilities. Firstly, ChatGPT can be used for mass phishing, enabling hackers to create personalized emails quickly and impersonate others to increase the chances of successful attacks. To combat this, employee education and verification mechanisms are crucial. Secondly, ChatGPT's ability to understand and explain code makes it susceptible to reverse engineering, allowing hackers to manipulate software and gain access to company servers. This skill, previously limited to nation-states, is now accessible to basic hackers. Lastly, ChatGPT can serve as a mini-brain for smart malware, autonomously extracting data from hacked networks and servers. Its capacity to sift through large amounts of data autonomously makes it more dangerous than traditional malware. The article emphasizes the need for companies to stay vigilant, train employees on the cybersecurity risks of ChatGPT, and prioritize privacy and security measures to mitigate vulnerabilities.<ref>{{cite web |title=3 Ways Hackers Use ChatGPT to Cause Security Headaches |url=https://www.darkreading.com/vulnerabilities-threats/3-ways-hackers-use-chatgpt-to-cause-security-headaches |website=Dark Reading |access-date=24 May 2023 |language=en |date=18 May 2023}}</ref>
 
|-
 
|-
| 2023 || March 20 || || ChatGPT experiences an outage due to a bug in an open-source library which allowed some users to see titles from another active user’s chat history. The bug would since been patched, and the ChatGPT service would be restored with the exception of a few hours of chat history. Upon deeper investigation, OpenAI discovered that the same bug may have caused the unintentional visibility of payment-related information of 1.2% of ChatGPT Plus subscribers active during a specific nine-hour window. OpenAI would reach out to notify affected users that their payment information may have been exposed and is committed to rebuilding trust with the ChatGPT community. Technical details of the bug and the actions taken to improve the system are also provided.<ref>{{cite web |title=March 20 ChatGPT outage: Here’s what happened |url=https://openai.com/blog/march-20-chatgpt-outage |website=openai.com |access-date=24 March 2023}}</ref>
+
| 2023 || May 18–22 || || Growth || According to Data.ai's analysis, ChatGPT surpasses almost all other leading AI chatbot apps in the United States. ChatGPT achieves 480,000 installs between May 18 and 22, only outperformed by "Chat with Ask AI", with 590,000 installs between April 4th and 8th.<ref name="OpenAI's ChatG">{{cite web |last1=Perez |first1=Sarah |title=OpenAI's ChatGPT app tops 500K downloads in just 6 days |url=https://techcrunch.com/2023/05/25/chatgpts-new-app-comes-out-of-the-gate-hot-tops-half-a-million-installs-in-first-6-days/ |website=TechCrunch |access-date=30 May 2023 |date=25 May 2023}}</ref>
 
|-
 
|-
| 2023 || March 21 || Competition || Google starts the public release of its chatbot, Bard, allowing consumers to sign up to a waiting list for English-language access in the US and UK. The program was previously open to approved testers. Google describes Bard as an experiment in generative AI, which creates rather than identifies content. Bard, unlike ChatGPT, does not generate computer code and its memory is limited. It also includes a feature to show three versions of an answer.<ref>{{cite web |title=Google begins opening access to its ChatGPT competitor Bard |url=https://finance.yahoo.com/news/google-begins-opening-access-chatgpt-140149232.html |website=Yahoo Finance |access-date=21 March 2023}}</ref>
+
| 2023 || May 19 || Hong Kong Baptist University trials ChatGPT amidst concerns over guidelines || Adoption || {{w|Hong Kong Baptist University}} initiates a trial for its teaching staff to use ChatGPT, with some professors expressing concerns about the lack of clear guidelines for its use. The university provides official accounts for staff to access the tool without a VPN as part of an agreement with Microsoft, which supports the software. However, no specific instructions are given on how to utilize the tool. While some professors are optimistic about the potential benefits of ChatGPT in reducing their workload, others criticize its effectiveness and highlight the absence of official parameters. Several faculty members emphasize the need for clear guidelines from the university to clarify the scope and appropriate applications of ChatGPT. The adoption of AI-based tools is not unique to Hong Kong Baptist University, as to date several other tertiary institutions in the city have embraced similar technologies while allowing staff and students to establish their own principles for usage.<ref>{{cite web |title=Hong Kong Baptist University begins ChatGPT trial for teachers, but concerns remain |url=https://www.scmp.com/news/hong-kong/education/article/3221176/hong-kong-baptist-university-begins-chatgpt-trial-teaching-staff-professors-wary-over-lack |website=South China Morning Post |access-date=24 May 2023 |language=en |date=19 May 2023}}</ref>
 
|-
 
|-
| 2023 || March 23 || || OpenAI announces ChatGPT implementation support for [[w:Plug-in (computing)|plugins]], which are tools designed for language models to access up-to-date information, run computations, or use third-party services with safety as a core principle. The feature is at first rolled out gradually for a small set of users and is expected to be expanded as more is learned about safety and alignment challenges. Plugin developers can use the provided documentation to build plugins, and the first plugins were created by companies such as {{w|Expedia}}, [[w:Slack Technologies|Slack]], {{w|Wolfram Research}}, and {{w|Zapier}}.<ref>{{cite web |title=ChatGPT plugins |url=https://openai.com/blog/chatgpt-plugins |website=openai.com |access-date=24 March 2023}}</ref>
+
| 2023 || May 19 || Standalone ChatGPT app is launched for iOS Users || Extension || OpenAI releases a standalone ChatGPT app exclusively for {{w|iOS}} users, with plans to launch an [[w:Android (operating system)|Android]] version in the near future. The app allows users to conveniently chat with the bot while on the move and introduces several features. Notably, it supports voice input utilizing OpenAI's Whisper, an open-source speech recognition model. Additionally, the app synchronizes chat history across devices and accounts logged into the web version of ChatGPT. Subscribers of ChatGPT Plus are benefited with access to GPT-4, early access to upcoming features, and faster response times. At first available in the United States, the app is reported to be soon be accessible to users in other countries, including Android users.<ref>{{cite web |title=OpenAI launches standalone ChatGPT app for iOS, Android to get one too |url=https://www.gsmarena.com/openai_launches_a_standalone_chatgpt_app_for_ios_android_soon_to_get_one_too-news-58616.php |website=GSMArena.com |access-date=22 May 2023}}</ref><ref>{{cite web |last1=Korn |first1=Jennifer |title=OpenAI launches a free ChatGPT app for iOS {{!}} CNN Business |url=https://edition.cnn.com/2023/05/18/tech/chatgpt-app/index.html |website=CNN |access-date=24 May 2023 |language=en |date=18 May 2023}}</ref>
 
|-
 
|-
| 2023 || End of year || || According to OpenAI's estimates, ChatGPT is expected to generate US$200 million in revenue by the end of this year.<ref name="tooltester"/>
+
| 2023 || May 24 || Opera introduces AI side panel 'Aria' powered by ChatGPT || Integration || Opera unveils a new AI side panel called "Aria" in its browser, which is powered by ChatGPT. Aria acts as a web and browser expert, assisting users in finding information, generating text or code, and answering product queries. The feature becomes available for testing and is accessible through Opera One and the beta version of Opera for Android. Aria connects to OpenAI's GPT technology and provides up-to-date information from the internet.<ref>{{cite web |last1=Malik |first1=Aisha |title=Opera launches new integrated AI sidebar powered by OpenAI's ChatGPT |url=https://techcrunch.com/2023/05/24/opera-launches-new-integrated-ai-sidebar-powered-by-openais-chatgpt/ |website=TechCrunch |access-date=24 May 2023 |date=24 May 2023}}</ref>  
 
|-
 
|-
| 2024 || End of the year || || According to OpenAI's estimates, ChatGPT is expected to generate US$1 billion in revenue by the end of this year.<ref name="tooltester"/>
+
| 2023 || May 24 || Risks of using ChatGPT in legal and compliance settings || Security || An article highlights six risks that legal and compliance leaders should consider when utilizing ChatGPT and other generative AI tools. These risks include fabricated and inaccurate answers, data privacy and confidentiality concerns, model and output bias, {{w|intellectual property}} and {{w|copyright risk}}s, cyber fraud risks, and consumer protection risks. Legal and compliance leaders are advised to establish guardrails and controls to mitigate these risks and ensure responsible use of generative AI tools within their organizations. They should emphasize the need for accuracy checks on ChatGPT's output, establish compliance frameworks for data privacy, stay updated on AI bias laws, monitor copyright law changes, address cyber fraud risks through coordination with cybersecurity personnel, and comply with consumer protection regulations and disclosure requirements. Failure to address these risks may expose enterprises to legal, reputational, and financial consequences, according to the author.<ref>{{cite web |last1=Security |first1=Help Net |title=6 ChatGPT risks for legal and compliance leaders |url=https://www.helpnetsecurity.com/2023/05/24/chatgpt-risks-legal-compliance/ |website=Help Net Security |access-date=24 May 2023 |date=24 May 2023}}</ref>
 +
|-
 +
| 2023 || May 25 || ChatGPT App expands to over 40 Countries || Expansion || OpenAI expands the availability of its official ChatGPT app for iPhone and iPad to over 40 countries in addition to the United States. The app, which was initially launched in the US, can now be used in countries including {{w|Canada}}, {{w|France}}, {{w|Germany}}, {{w|India}}, {{w|Japan}}, {{w|Mexico}}, {{w|New Zealand}}, the {{w|United Arab Emirates}}, and the {{w|United Kingdom}}.<ref>{{cite web |title=OpenAI's Official ChatGPT App for iPhone Expands to 40+ Countries |url=https://www.macrumors.com/2023/05/25/chatgpt-app-40-countries/ |website=MacRumors |access-date=30 May 2023 |language=en |date=25 May 2023}}</ref><ref>{{cite web |last1=Singh |first1=Jagmeet |title=OpenAI's ChatGPT iOS app now available in Canada, India, Brazil and 30 more countries |url=https://techcrunch.com/2023/05/25/openai-chatgpt-ios-app-india-download/ |website=TechCrunch |access-date=30 May 2023 |date=26 May 2023}}</ref> 
 +
|-
 +
| 2023 || May 27 || AI-Generated content misleads lawyer in court filing || Adoption || An article discusses a case where a lawyer used ChatGPT to assist in preparing a court filing. However, it was later discovered that the AI-generated content, including citations to court decisions, was entirely false. The lawyer claims to be unaware of the program's unreliability. The incident sparks discussions among legal professionals about the value and risks of using AI software in the legal field. It underscores the importance of human verification and caution when incorporating AI-generated information into legal documents. The case highlights the ongoing role of human professionals in the legal profession and the need for thorough authenticity checks.<ref>{{cite web |last1=Weiser |first1=Benjamin |title=Here’s What Happens When Your Lawyer Uses ChatGPT |url=https://www.nytimes.com/2023/05/27/nyregion/avianca-airline-lawsuit-chatgpt.html |website=The New York Times |access-date=30 May 2023 |date=27 May 2023}}</ref><ref>{{cite web |title=ChatGPT: US lawyer admits using AI for case research |url=https://www.bbc.com/news/world-us-canada-65735769 |website=BBC News |access-date=30 May 2023 |date=27 May 2023}}</ref><ref>{{cite web |title=A lawyer used ChatGPT to prepare a court filing. It went horribly awry. |url=https://www.cbsnews.com/news/lawyer-chatgpt-court-filing-avianca/ |website=www.cbsnews.com |access-date=30 May 2023 |date=29 May 2023}}</ref>
 +
|-
 +
| 2023 || May 29 || Nvidia's AI expansion, DGX GH200 for ChatGPT successors || Competition || {{w|Nvidia}} announces plans to release more AI products, including an AI supercomputer platform called DGX GH200, whose primary goal is to assist tech companies in developing successors to ChatGPT. Big Tech firms like Microsoft, Meta, and Google's Alphabet are expected to be early users of this technology. While Nvidia remains at the forefront of AI chip development, other companies are also entering the AI space by this time, including Microsoft, which develops its own chip for ChatGPT-like applications.<ref>{{cite web |title=Nvidia introduces AI supercomputer to create ChatGPT successors |url=https://cointelegraph.com/news/nvidia-introduces-ai-supercomputer |website=cointelegraph.com |access-date=30 May 2023}}</ref><ref>{{cite web |title=Nvidia introduces AI supercomputer to create ChatGPT successors By Cointelegraph |url=https://www.investing.com/news/cryptocurrency-news/nvidia-introduces-ai-supercomputer-to-create-chatgpt-successors-3093054 |website=Investing.com |access-date=30 May 2023 |language=en-us}}</ref><ref>{{cite web |title=Nvidia unveils DGX GH200 AI computing platform for ChatGPT successors and ACE for Games service for ultrarealistic NPCs |url=https://www.notebookcheck.net/Nvidia-unveils-DGX-GH200-AI-computing-platform-for-ChatGPT-successors-and-ACE-for-Games-service-for-ultrarealistic-NPCs.721489.0.html |website=Notebookcheck |access-date=30 May 2023 |language=en |date=29 May 2023}}</ref>
 +
|-
 +
| 2023 || June 8 || Survey, low adoption of ChatGPT and Google Bard chatbots || Adoption || A survey conducted by {{w|Morgan Stanley}} reveals that the adoption of AI chatbots like ChatGPT and Google Bard is relatively low. Only 19% of respondents have used ChatGPT, and even fewer, 9% of respondents, have used Google Bard. The study finds that the majority of people who are not using chatbots are unlikely to use them in the next six months. While chatbots are commonly used for learning about new topics, the monetization of AI tools is still in the early stages at this time. However, the impact of AI on e-commerce and travel, particularly in terms of AI-enabled searches and recommendations, has the potential to boost revenue. Google, with its comprehensive and personalized travel search offering, has an advantage in this space.<ref>{{cite web |last1=Melton |first1=Monica |title=There aren't actually THAT many people using ChatGPT |url=https://www.businessinsider.com/chatgpt-ai-adoption-slow-google-bard-morgan-stanley-2023-6?r=US&IR=T |website=Business Insider |access-date=18 July 2023}}</ref>
 +
|-
 +
| 2023 || June 8 || Algorithm detects ChatGPT-generated scientific papers with high accuracy || Adoption || Scientists claim having trained a machine learning algorithm to identify scientific papers generated by ChatGPT with over 99% accuracy. By this time, concerns have been raised about AI-generated text, as it can be difficult to distinguish from human writing, leading to potential issues of plagiarism or cheating. Researchers from the {{w|University of Kansas}} developed an algorithm that classifies papers written by scientists and by ChatGPT. The algorithm focuses on stylistic differences between human and AI writing, such as vocabulary richness, paragraph length, punctuation usage, and the use of specific language. However, it is reported that further studies are needed to determine the algorithm's robustness against lightly edited papers and different scientific journals.<ref>{{cite web |last1=Quach |first1=Katyanna |title=Scientists claim over 99 percent identification of ChatGPT |url=https://www.theregister.com/2023/06/08/scientists_ai_recognition/ |website=www.theregister.com |access-date=18 July 2023 |language=en}}</ref>
 +
|-
 +
| 2023 || June 9 || Implications of ChatGPT in healthcare, interoperability and fairness concerns || Adoption || An article by {{w|Brookings Institution}} discusses the implications of using ChatGPT in the healthcare industry. While the technology shows promise in improving patient-provider interactions and healthcare services, the article highlights concerns regarding interoperability and fairness. The US healthcare market is dominated by traditional providers who lack incentives for interoperability. The emergence of formidable competitors, such as technology companies like Amazon's OneMedical, poses a threat to traditional providers, leading them to withhold data. This could hinder interoperability and lead to a fragmented landscape with multiple AI programs of varying quality. Additionally, the use of large language models in healthcare services could exacerbate disparities in access to care, potentially creating a two-tier system. The article suggests government support, updated reimbursement policies, and a focus on value-based payment models to address these challenges.<ref>{{cite web |title=ChatGPT and health care: implications for interoperability and fairness |url=https://www.brookings.edu/articles/chatgpt-and-health-care-implications-for-interoperability-and-fairness/ |website=Brookings |access-date=18 July 2023}}</ref>
 +
|-
 +
| 2023 || June 9 || Sam Altman calls for global regulation of generative AI || Security || {{w|Sam Altman}} calls for global cooperation in regulating generative artificial intelligence (AI) during a visit to South Korea. Altman emphasizes the need for international standards and collaboration as AI systems become increasingly powerful. He highlights the potential risks associated with the speed at which AI advancements are happening and stresses the importance of addressing these challenges globally. Altman's visit coincides with discussions among US, European, and Chinese officials on AI oversight and regulation. Altman also acknowledges the impact of AI on the job market but expresses optimism about the potential for new opportunities and human advancement in the future.<ref>{{cite web |last1=Seo |first1=Michelle Toh,Yoonjung |title=OpenAI CEO calls for global cooperation to regulate AI {{!}} CNN Business |url=https://edition.cnn.com/2023/06/09/tech/korea-altman-chatgpt-ai-regulation-intl-hnk/index.html |website=CNN |access-date=17 July 2023 |language=en |date=9 June 2023}}</ref>
 +
|-
 +
| 2023 || June 9 || Debating the impact of ChatGPT and AI on humanity || Security || An article explores the contrasting views on the impact of ChatGPT and artificial intelligence on humanity. {{w|Richard Branson}} and others see AI as transformative and revolutionizing various aspects of life, while {{w|Elon Musk}} and the creators of ChatGPT express concerns about the potential risks and harm it may cause. There are fears that AI could be misused in human resource management and surpass regulation, leading to significant harm or even the extinction of humanity. The article highlights the need for government oversight and regulation to mitigate these risks. On the other hand, proponents argue that AI, including ChatGPT, can be beneficial for businesses by assisting in hiring processes and leveling the playing field for aspiring entrepreneurs. The article presents a debate between the potential benefits and risks associated with AI, leaving the question open-ended.<ref>{{cite web |last1=Ph.D |first1=Bryan Robinson |title=Will ChatGPT Lead To Extinction Or Elevation Of Humanity? A Chilling Answer |url=https://www.forbes.com/sites/bryanrobinson/2023/06/09/will-chatgpt-lead-to-extinction-or-elevation-of-humanity-a-chilling-answer/?sh=5cfea74dcfeb |website=Forbes |access-date=18 July 2023 |language=en}}</ref>
 +
|-
 +
| 2023 || June 9 || ChatGPT-3.5's limited knowledge of jokes and repetitive responses || Performance || Researchers find that ChatGPT-3.5 has a limited knowledge of jokes and tends to repeat the same ones. In a test run, 90% of the 1,008 jokes generated by ChatGPT are the same 25 jokes. This suggests that the responses are likely memorized during the AI model's training rather than being newly generated. While ChatGPT provides valid explanations for the jokes and shows an understanding of wordplay and double meanings, it struggles with sequences that don't fit learned patterns and can't determine when a joke isn't funny. The researchers believe that despite its limitations, ChatGPT's focus on content and meaning represents progress in the field of computational humor.<ref>{{cite web |last1=Edwards |first1=Benj |title=Researchers discover that ChatGPT prefers repeating 25 jokes over and over |url=https://arstechnica.com/information-technology/2023/06/researchers-discover-that-chatgpt-prefers-repeating-25-jokes-over-and-over/ |website=Ars Technica |access-date=18 July 2023 |language=en-us |date=9 June 2023}}</ref>
 +
|-
 +
| 2023 || June 10 || ChatGPT 3.5 and the risk of malicious code exploitation || Security || An article warns about the potential for ChatGPT 3.5 to be exploited by hackers to spread malicious code from trusted repositories. Security researchers at Vulcan.io conducted a proof of concept that revealed how ChatGPT could recommend non-existent code packages to users. Out of the Node.js and Python questions asked, a significant number of ChatGPT's answers contained hallucinated code packages. The researchers created a non-malicious package with the same name as a non-existent package in the repository, and when a user followed ChatGPT's recommendation to install it, the package communicated with the attacker's server. The researchers advise caution and verification when using ChatGPT's recommendations.<ref>{{cite web |last1=Montti |first1=Roger |title=Researchers Warn That ChatGPT Can Be Used To Spread Malicious Code |url=https://www.searchenginejournal.com/how-chatgpt-can-recommend-malicious-code/488916/#close |website=Search Engine Journal |access-date=18 July 2023 |language=en |date=10 June 2023}}</ref>
 +
|-
 +
| 2023 || June 12 || Security concerns, ChatGPT 3.5's potential for code exploitation || Adoption || It is reported that physicians are utilizing ChatGPT to engage with patients more empathetically. This raises questions about the balance between objective medical guidance and human compassion. While some experts argue that ChatGPT can simplify explanations by avoiding complex medical jargon, others express concerns. Microsoft, collaborating closely with OpenAI, is unsure about using the chatbot for delivering bad medical news. Critics highlight the importance of human compassion in healthcare, emphasizing that relying on a tool that lacks the ability to distinguish right from wrong may worsen patient experiences and potentially impact medical decisions.<ref>{{cite web |title=Microsoft Alarmed at Doctors Using ChatGPT to Tell Patients Bad News |url=https://futurism.com/neoscope/microsoft-doctors-chatgpt-patients |website=Futurism |access-date=18 July 2023}}</ref>
 +
|-
 +
| 2023 || June 13 || Sam Altman addresses AI's transformative role in education at Keio University || Impact || {{w|Sam Altman}} speaks to students at {{w|Keio University in Tokyo}}, highlighting the transformative potential of artificial intelligence (AI) tools in education. Altman compares AI tools to calculators, stating that they will revolutionize teaching and evaluation methods. While acknowledging concerns about potential misuse and reliance on AI for original work, Altman emphasizes the need for regulations and safety measures in the AI field. He expresses positivity about new regulatory frameworks for AI but stresses the importance of taking responsibility in case of any negative outcomes. Altman also addresses concerns about job displacement, stating that new classes of jobs will emerge despite some jobs becoming obsolete.<ref>{{cite web |last1=Author |first1=No |title=Homework will 'never be the same,' says ChatGPT founder |url=https://www.japantimes.co.jp/news/2023/06/13/business/corporate-business/chatgpt-homework-revolution/ |website=The Japan Times |access-date=17 July 2023 |date=13 June 2023}}</ref>
 +
|-
 +
| 2023 || June 15 || Mercedes-Benz and Microsoft integrate ChatGPT into U.S. cars || Integration || {{w|Mercedes-Benz}} and {{w|Microsoft}} agree to integrate ChatGPT into Mercedes-Benz cars in the United States. The addition of ChatGPT aims to enhance voice-command capabilities in Mercedes vehicles, allowing for more natural and fluid conversations with the system. ChatGPT can remember context, engage in back-and-forth dialogue, and respond to a wider range of requests, even those unrelated to the car or driving. It can also interact with other applications for tasks like making reservations or purchasing tickets. Mercedes owners with the MBUX infotainment system can beta test ChatGPT by using the voice command "Hey Mercedes, I want to join the beta program."<ref>{{cite web |last1=Valdes-Dapena |first1=Peter |title=Mercedes is bringing ChatGPT into its cars {{!}} CNN Business |url=https://edition.cnn.com/2023/06/15/business/mercedes-benz-chatgpt/index.html |website=CNN |access-date=17 July 2023 |language=en |date=15 June 2023}}</ref>
 +
|-
 +
| 2023 || June 16 || AI's potential impact on job roles, including ChatGPT's influence || Impact || An article discusses the potential impact of AI, specifically ChatGPT, on various job roles within the following five years. IBM's AI, WatsonX, is expected to replace 8,000 jobs in the company's HR department. By this time, automation is also affecting customer service tasks, although human reps who handle emotional situations would still be valuable. In the field of medicine, AI would primarily assist in diagnostics, allowing physicians to focus on other aspects of patient care. Coders and programmers may see a decrease in the need for mundane tasks as AI can handle them. Content writing tasks could be taken over by ChatGPT, but its factual accuracy is still evolving.<ref>{{cite web |last1=Cyr |first1=Alex |title=Is AI Coming for Your Job? |url=https://canadianbusiness.com/ideas/chatgpt-ai-replacing-jobs-canada/ |website=Canadian Business – How to Do Business Better |access-date=18 July 2023 |date=16 June 2023}}</ref>
 +
|-
 +
| 2023 || June 16 || Companies cautiously ban ChatGPT amid privacy and security concerns || Restriction || It is reported that some companies are banning ChatGPT due to concerns over security and privacy. Uploading work-related information to an AI platform owned by another company poses potential risks, as the data may be stored on their servers without a way to access or delete it. Companies fear that proprietary or sensitive information given to ChatGPT could be unintentionally shared with others. By this time, OpenAI has experienced security issues with ChatGPT and is still addressing them. As a result, some companies opt to develop their own AI platforms as safer alternatives. While some companies embrace ChatGPT for cost-saving and content generation purposes, many see more risks than benefits in its use.<ref>{{cite web |last1=Nelson |first1=Felicity |title=Many Companies Are Banning ChatGPT. This Is Why. |url=https://www.sciencealert.com/many-companies-are-banning-chatgpt-this-is-why |website=ScienceAlert |access-date=18 July 2023 |date=16 June 2023}}</ref>
 +
|-
 +
| 2023 || June 22 || Lawyers fined for using AI-generated fake cases in court || Adoption || Two lawyers who used fake cases generated by ChatGPT are fined $5,000 and ordered to inform the judges involved about the situation. New York District Judge P. Kevin Castel rules that the lawyers, Steven Schwartz and Peter LoDuca, acted in bad faith and misled the court by knowingly presenting false cases. While Judge Castel acknowledges the potential benefits of using AI tools for legal assistance, he emphasizes that the lawyers neglected their responsibilities and failed to rectify their mistakes. The case marks a significant instance of sanctions related to the use of AI in the legal field.<ref>{{cite web |last1=Bohannon |first1=Molly |title=Judge Fines Two Lawyers For Using Fake Cases From ChatGPT |url=https://www.forbes.com/sites/mollybohannon/2023/06/22/judge-fines-two-lawyers-for-using-fake-cases-from-chatgpt/?sh=4d0a0c1f516c |website=Forbes |access-date=17 July 2023 |language=en}}</ref>
 +
|-
 +
| 2023 || June 27 || Booking.com integrates ChatGPT API for AI trip planner || Integration || Booking.com introduces conversational chat capabilities to its mobile app by integrating ChatGPT API. The AI Trip Planner allows users to receive travel recommendations based on their preferences described in natural language. The chat platform is integrated with Booking.com's accommodation booking experience, providing users with a visual list of properties that meet their queries. Future plans include integrating other travel components such as flights, ground transportation, and attractions into the chat. The AI Trip Planner is initially available to Booking.com's loyalty members in the United States through the mobile app, with plans to extend it to the desktop platform later.<ref>{{cite web |title=Booking.com adds conversational trip planning with ChatGPT |url=https://www.phocuswire.com/booking-com-adds-conversational-trip-planning-chatgpt |website=www.phocuswire.com |access-date=18 July 2023 |language=en}}</ref>
 +
|-
 +
| 2023 || July 12 || ChatGPT downtime sparks interest in AI alternatives || Competition || An article discusses the recent downtime of ChatGPT and provides alternative AI tools and chatbots for users to consider. The outage led to frustration among users, and OpenAI confirmed the issue and worked to resolve it. The mentioned alternatives include Google Bard; Midjourney, a text-to-image AI generator; Copy.ai, a platform for copywriting; Tableau, an analytics and data visualization platform; Murf, an AI tool for text-to-speech conversion; Jasper, an AI tool for writing various content types; Fireflies, an AI tool for real-time voice transcription; and Pictory, a video editing tool for creating branded clips.<ref>{{cite web |title=ChatGPT down? Here are the best AI alternatives to consider |url=https://www.euronews.com/next/2023/07/12/chatgpt-down-try-these-alternative-ai-tools-instead |website=euronews |access-date=14 July 2023 |language=en |date=12 July 2023}}</ref> 
 +
|-
 +
| 2023 || July 13 || FTC launches investigation into ChatGPT's impact and OpenAI's practices || System behavior || The Federal Trade Commission (FTC) launches an investigation into ChatGPT to determine if it has caused harm by providing users with incorrect information. The FTC reportedly also examines OpenAI's privacy and data security practices. In a letter to OpenAI, the agency requests company records and data on various topics, including policies, financial earnings, and details of the Large Language Models used to train the chatbot. OpenAI's CEO, Sam Altman, expresses disappointment in the leak of the FTC's request but emphasizes the company's commitment to ensuring the safety and consumer-friendliness of its technology. This investigation marks a significant step in government regulatory action within the AI industry. The FTC, under Chair Lina Khan, has been actively pursuing major tech companies and has stated that AI is subject to consumer protection laws.<ref>{{cite web |title=FTC investigating ChatGPT over potential consumer harm |url=https://www.npr.org/2023/07/13/1187532997/ftc-investigating-chatgpt-over-potential-consumer-harm |website=www.npr.org |access-date=14 July 2023}}</ref>
 +
|-
 +
| 2023 || July 14 || Meta Platforms to launch commercial version of LLaMA AI model || Competition || {{w|Meta Platforms}} reportedly plans to release a commercial version of its open-source large language model (LLM), called LLaMA. LLaMA uses AI to generate text, images, and code. Initially introduced as a foundational model for researchers and academics, the commercial release of LLaMA would allow developers and businesses to build applications using the model. With LLaMA available in different parameter sizes, Meta aims to compete with OpenAI, which leads the AI race with ChatGPT, and Google's proprietary LLMs.<ref>{{cite web |title=Meta to release open-source commercial AI model to compete with OpenAI and Google |url=https://www.zdnet.com/article/meta-to-release-open-source-commercial-ai-model-to-compete-with-openai-and-google/ |website=ZDNET |access-date=16 July 2023 |language=en}}</ref>
 +
|-
 +
| 2023 || July 20 || OpenAI introduces custom instructions for ChatGPT interactions || Update || OpenAI introduces custom instructions for ChatGPT users, allowing them to set specific context and preferences for the chatbot's responses. This feature enables users to avoid repeatedly providing the same prompts during interactions. With custom instructions, users can specify details like preferred tone, code language, family size, and more. The feature is available for Plus plan users and works with plug-ins, providing location-based suggestions. OpenAI uses the data from custom instructions to improve model performance and adaptation to different instructions. The company ensures safety by moderating instructions to prevent harmful or hateful responses.<ref>{{cite web |last1=Mehta |first1=Ivan |title=OpenAI launches customized instructions for ChatGPT |url=https://techcrunch.com/2023/07/20/openai-launches-customized-instructions-for-chatgpt/ |website=TechCrunch |access-date=29 July 2023 |date=20 July 2023}}</ref>
 +
|-
 +
| 2023 || July 22 || OpenAI debuts ChatGPT app for Android, expanding market presence || Extension || OpenAI launches the official ChatGPT app for [[w:Android (operating system)|Android]], following its release on iOS earlier. The company allows users to pre-register for installation on the Google Play Store. The Android app boasts improved security measures and features like conversation history synchronization. With this move, OpenAI competes directly with Microsoft's Bing Chat, which also utilizes GPT-4 technology. The release comes amidst emerging competitors in the chatbot market.<ref>{{cite web |title=OpenAI launches official ChatGPT app for Android |url=https://cointelegraph.com/news/open-ai-launches-official-chat-gpt-app-android |website=Cointelegraph |access-date=25 July 2023 |language=en |date=22 July 2023}}</ref>
 +
|-
 +
| 2023 || July 25 || OpenAI launches ChatGPT for Android in selected countries || Expansion || OpenAI releases ChatGPT for [[w:Android (operating system)|Android]] in the {{w|United States}}, {{w|India}}, {{w|Bangladesh}}, and Brazil. Initially available for preorders on Google Play, it becomes live in these selected countries. The Android release is a significant milestone for OpenAI, as there are more Android users globally. The mobile interface is optimized for ease of use, allowing users to interact with ChatGPT as if messaging a friend. Existing OpenAI account holders can access past conversations across devices. Additionally, ChatGPT Plus, which offers a better language model and more features, is available through an in-app purchase for Android users.<ref>{{cite web |last1=Dillet |first1=Romain |title=ChatGPT for Android is now available in the US, India, Bangladesh and Brazil |url=https://techcrunch.com/2023/07/25/chatgpt-for-android-is-now-available-in-the-us-india-bangladesh-and-brazil/ |website=TechCrunch |access-date=29 July 2023 |date=25 July 2023}}</ref>
 +
|-
 +
| 2023 || July 27 || Report reveals vulnerabilities in AI Chatbots, posing harmful content risks || Security || A report by US-based researchers reveals a method to consistently bypass safety measures in AI chatbots like ChatGPT and Bard, allowing them to generate harmful content. The researchers append long suffixes of characters to prompts, enabling chatbots to produce hate speech, disinformation, and toxic material. While companies like OpenAI and Google can block specific suffixes, preventing all attacks of this kind remains challenging. The research raises concerns about chatbots flooding the internet with dangerous content and misinformation. AI developers, including Anthropic, Google, and OpenAI, were presented with the findings for their responses. This highlights the need to address vulnerabilities before deploying chatbots in sensitive domains.<ref>{{cite web |title=AI researchers say they’ve found a way to jailbreak Bard and ChatGPT |url=https://cointelegraph.com/news/researchers-claim-method-to-jailbreak-ai-chatbots-llm-chatgpt-bard |website=Cointelegraph |access-date=29 July 2023 |language=en |date=28 July 2023}}</ref>
 +
|-
 +
| 2023 || August 14 || Academic integrity prevails || Impact || Due to rampant cheating using OpenAI's ChatGPT, it is reported that some college professors are reverting to paper exams. Students are found to copy assignments verbatim from ChatGPT, leading educators to ban the use of AI and return to traditional testing methods. Despite mixed reactions, such as computer science major Ronan Takizawa acknowledging its regressive nature yet promoting better understanding, the move intends to ensure students grasp concepts and skills.<ref>{{cite web |last1=Johnson |first1=Chelsea |title=College Professors to Revert to Paper Exams Due to Cheating with ChatGPT |url=https://fagenwasanni.com/news/college-professors-to-revert-to-paper-exams-due-to-cheating-with-chatgpt/198857/ |website=Fagen Wasanni Technologies |access-date=15 August 2023 |date=14 August 2023}}</ref> 
 +
|-
 +
| 2023 || August 14 || ChatGPT's hefty costs || Maintenance || A report reveals that running ChatGPT costs OpenAI about $700,000 per day, with investments from Microsoft and others covering these expenses. However, concerns arise about the company's financial stability, suggesting the possibility of bankruptcy by the end of 2024. Despite projecting a potential revenue of $200 million in 2023 and aiming for $1 billion in 2024, OpenAI's losses mount to date, notably with $540 million since ChatGPT's development initiation. The report indicates a decline in user visits to the ChatGPT website, potentially due to API cannibalization and competition from open-source models like Meta's Llama 2. Additionally, the GPU shortage and Elon Musk's alternative chatbot contribute to the challenges OpenAI faces.<ref>{{cite web |title=ChatGPT costs $700,000 to run daily, OpenAI may go bankrupt in 2024 |url=https://technext24.com/2023/08/14/chatgpt-costs-700000-daily-openai/ |website=technext24.com |access-date=15 August 2023 |date=14 August 2023}}</ref>
 +
|-
 +
| 2023 || August 28 || ChatGPT Enterprise launches || Product launch || OpenAI launches ChatGPT Enterprise, a business-tier version of its AI chatbot. ChatGPT Enterprise offers access to GPT-4 without usage caps, delivers performance up to two times faster than previous versions, and includes API credits. The pricing for ChatGPT Enterprise varies depending on the company's use cases and size. It allows clients to input company data to customize ChatGPT for their industries and use cases. OpenAI plans to introduce ChatGPT Business for smaller teams in the future. The launch comes as tech giants like Microsoft, Google, and others compete to advance generative AI in chatbot applications.<ref>{{cite web |last1=Field |first1=Hayden |title=OpenAI launches ChatGPT Enterprise, the company's biggest announcement since ChatGPT's debut |url=https://www.cnbc.com/2023/08/28/openai-chatgpt-enterprise-launches.html |website=CNBC |access-date=17 September 2023 |language=en |date=28 August 2023}}</ref>
 +
|-
 +
| 2023 || September 11 || ChatGPT's traffic slump || Adoption || It is reported that ChatGPT website is in gradual decline in global desktop and mobile traffic accross three months, with a 3.2 percent drop in August, totaling 1.43 billion visits. This trend follows roughly 10 percent declines in website visits in June and July. Visitor session durations also show decrease monthly since March, falling from an average of 8.7 minutes per session to 7 minutes in August. However, there is a slight increase in the number of unique visitors worldwide in August, reaching 180.5 million users. The return of students to school, with some institutions integrating ChatGPT into their curriculum, also boosts traffic and usage, particularly for homework assistance. Despite these challenges, ChatGPT initially gained 100 million monthly active users within two months of its November 2022 launch. It ranks among the top 30 global websites, but it faces increasing competition from other AI chatbots like Google's Bard and Microsoft's Bing chatbot.<ref>{{cite web |title=OpenAI’s ChatGPT web traffic plummets, is the hype over? |url=https://www.edgemiddleeast.com/emergent-tech/openais-chatgpt-web-traffic-plummets-is-the-hype-over |website=edgemiddleeast.com |access-date=13 September 2023}}</ref>
 +
|-
 +
| 2023 || September 11 || ChatGPT helps unravel a 3-year medical mystery || Performance || ChatGPT plays a pivotal role in diagnosing a young boy who suffered chronic pain for three years and saw 17 doctors without receiving a conclusive diagnosis. Frustrated and exhausted, Alex's mother turns to ChatGPT, providing it with her son's medical information and symptoms. ChatGPT suggests tethered cord syndrome, which turns out to be the correct diagnosis. Tethered cord syndrome restricts spinal cord movement, leading to various symptoms. While ChatGPT aids in the diagnosis, it's noted that the AI has limitations, including occasional fabrication of information, highlighting the importance of clinical validation in healthcare AI. Alex undergoes surgery to treat his condition, and his mother shares their story to help others facing similar challenges.<ref>{{cite web |title=A boy saw 17 doctors over 3 years for chronic pain. ChatGPT found the diagnosis |url=https://www.today.com/health/mom-chatgpt-diagnosis-pain-rcna101843 |website=TODAY.com |access-date=17 September 2023 |language=en |date=12 September 2023}}</ref>
 +
|-
 +
| 2023 || September 12 || Authors sue OpenAI for ChatGPT training copyright infringement || {{w|Copyright infringement}} || A group of authors, including Pulitzer Prize winner {{w|Michael Chabon}}, file a lawsuit against OpenAI in a San Francisco federal court, accusing the company of using their writings without permission to train ChatGPT. The authors argue that OpenAI copied their works to teach ChatGPT to respond to human text prompts. This lawsuit is the third proposed copyright-infringement class action against Microsoft-backed OpenAI by authors. OpenAI and other companies claim that using copyrighted material from the internet for AI training constitutes fair use. The authors seek damages and an injunction against OpenAI's alleged "unlawful and unfair business practices."<ref>{{cite web |title=Pulitzer winner among authors suing ChatGPT over copyright claims |url=https://www.euronews.com/next/2023/09/12/pulitizer-prize-winner-among-authors-suing-chatgpt-creator-openai-for-copyright-infringeme |website=euronews |access-date=17 September 2023 |language=en |date=12 September 2023}}</ref>
 +
|-
 +
| 2023 || September 13 || Alibaba releases LLM || Competition || Alibaba releases its large language model Tongyi Qianwen, which is made available for public and enterprise use in China. Tongyi Qianwen, similar to ChatGPT, was previously in a beta test phase and is trained on English and Chinese text, although its exact specifications are undisclosed. This release coincides with the relaxation of AI technology restrictions in China, which now require vetting and certification for public AI tech. Companies like {{w|Baidu}}, {{w|Tencent}}, {{w|TikTok}}, and {{w|ByteDance}} have already received approval to launch AI models in China by this time. In contrast, the U.S. remains in the early stages of AI regulation discussions.<ref>{{cite web |title=Alibaba launches its ChatGPT-like AI model for public use amid loosening restrictions in China |url=https://cointelegraph.com/news/alibaba-launches-chat-gpt-like-ai-model-for-public-restrictions-china |website=Cointelegraph |access-date=17 September 2023 |language=en |date=13 September 2023}}</ref>
 +
|-
 +
| 2023 || September 15 || ChatGPT Plus vs Claude Pro || Competition || An article discusses the competition between ChatGPT Plus and Claude Pro, two subscription-based AI chatbot services. While both are based on large language models, they differ in their underlying models. ChatGPT Plus relies on GPT-4, known for advanced reasoning and creativity, while Claude Pro uses Claude 2, designed for summarization and ethical alignment. Both offer free versions and APIs for customization. They both cost $20/month and have similar performance, with ChatGPT being slightly more creative and Claude Pro excelling in summarizing large text volumes. The choice depends on personal preference and specific needs, with ChatGPT offering more features, including plugins, while Claude Pro emphasizes summarization and ethical AI.<ref>{{cite web |title=Claude Pro vs. ChatGPT Plus: Which AI chatbot is better for you? |url=https://venturebeat.com/ai/claude-pro-vs-chatgpt-plus-comparison-which-ai-chatbot-is-better-for-you/ |website=VentureBeat |access-date=18 September 2023 |date=15 September 2023}}</ref>
 +
|-
 +
| 2023 || September 16 || ChatGPT healthcare potential || Performance || A study conducted by {{w|Emory University School of Medicine}} finds that ChatGPT performs well in diagnosing eye-related complaints and outperforms the symptom checker on WebMD. ChatGPT is seen as an improvement over searching symptoms online, but it raises questions about how to integrate chatbots like ChatGPT into healthcare systems. Some believe that AI-based medical tools should undergo a regulatory approval process similar to that of drugs, but there's no clear framework for this yet. Despite the potential advantages, concerns about privacy, safety, bias, liability, transparency, and regulation persist in the deployment of AI in healthcare.<ref>{{cite web |title='Dr. Google' meets its match in Dr. ChatGPT |url=https://www.npr.org/sections/health-shots/2023/09/16/1199924303/chatgpt-ai-medical-advice |website=npr.org |access-date=17 September 2023}}</ref><ref>{{cite journal |last1=Lyons |first1=Riley J. |last2=Arepalli |first2=Sruthi R. |last3=Fromal |first3=Ollya |last4=Choi |first4=Jinho D. |last5=Jain |first5=Nieraj |title=Artificial Intelligence Chatbot Performance in Triage of Ophthalmic Conditions |date=12 June 2023 |doi=10.1101/2023.06.11.23291247}}</ref>
 +
|-
 +
| 2023 || September 20 || OpenAI integrates DALL-E 3 into ChatGPT || Update || OpenAI unveils DALL-E 3, an advanced version of its image generator, which is integrated into ChatGPT. DALL-E 3 can create highly detailed images, excelling in generating images with letters, numbers, and human hands. This development enhances ChatGPT's capabilities as a hub for generative AI, producing text, images, sounds, and more. With DALL-E 3 integrated into ChatGPT, users can describe what they want to see, generating images through the chatbot, streamlining media production. However, OpenAI is cautious about the technology's potential for misuse and is implementing safeguards.<ref>{{cite web |last1=Metz |first1=Cade |last2=Hsu |first2=Tiffany |title=ChatGPT Can Now Generate Images, Too |url=https://www.nytimes.com/2023/09/20/technology/chatgpt-dalle3-images-openai.html |website=The New York Times |access-date=22 September 2023 |date=20 September 2023}}</ref>
 +
|-
 +
| 2023 || End of year || || Revenue || According to OpenAI's estimates, ChatGPT is expected to generate US$200 million in revenue by the end of this year.<ref name="tooltester"/>
 +
|-
 +
| 2024 || End of the year || || Revenue || According to OpenAI's estimates, ChatGPT is expected to generate US$1 billion in revenue by the end of this year.<ref name="tooltester"/>
 
|-
 
|-
 
|}
 
|}
 +
 +
== Numerical and visual data ==
 +
 +
The chart below shows pageviews of the English Wikipedia article ChatGPT, from November 2022 to August 2023. See spike of interest towards April, and notable loss of interest afterwards.<ref>{{cite web |title=Wikipedia Views: ChatGPT |url=https://wikipediaviews.org/displayviewsformultiplemonths.php?page=ChatGPT&months[0]=202308&months[1]=202307&months[2]=202306&months[3]=202305&months[4]=202304&months[5]=202303&months[6]=202302&months[7]=202301&months[8]=202212&months[9]=202211&language=en&drilldown=all |website=wikipediaviews.org |access-date=31 August 2023}}</ref>
 +
 +
[[File:Chatgpt-wikipedia-views.PNG|thumb|center|500px]]
 +
 +
 +
 +
 +
The chart below shows {{w|Google Trends}} data for ChatGPT, from November 2022 to August 2023, when the screenshot was taken. Interest is also ranked by country and displayed on world map.<ref>{{cite web |title=Google Trends |url=https://trends.google.com/trends/explore?date=2022-11-01%202023-08-30&q=chatgpt&hl=en |website=Google Trends |access-date=31 August 2023}}</ref> 
 +
 +
[[File:Chatgpt-google-trends.png|thumb|center|700px]]
  
 
==Meta information on the timeline==
 
==Meta information on the timeline==
Line 165: Line 310:
 
===How the timeline was built===
 
===How the timeline was built===
  
The initial version of the timeline was written by [[User:FIXME|FIXME]].
+
The initial version of the timeline was written by [[User:Sebastian|Sebastian]].
  
 
{{funding info}} is available.
 
{{funding info}} is available.
Line 177: Line 322:
 
===What the timeline is still missing===
 
===What the timeline is still missing===
  
 +
* search for more criticism (criticism event type)
 +
* search for more on security (security event type)
 +
* other event types: update, notable statement, growth, official response
 +
* [[Review process for timelines]]
 +
* Make the big picture more neutral.
 +
* https://www.technologyreview.com/2023/03/03/1069311/inside-story-oral-history-how-chatgpt-built-openai/
 
* https://www.jointjs.com/demos/chatgpt-timeline
 
* https://www.jointjs.com/demos/chatgpt-timeline
 
 
 
* https://www.tooltester.com/en/blog/chatgpt-statistics/ <ref name="tooltester">
 
* https://www.tooltester.com/en/blog/chatgpt-statistics/ <ref name="tooltester">
 
* https://www.precedenceresearch.com/artificial-intelligence-market for visual data
 
* https://www.precedenceresearch.com/artificial-intelligence-market for visual data
Line 187: Line 336:
 
==See also==
 
==See also==
  
 +
* [[Timeline of Bing Chat]]
 +
* [[Timeline of Google Bard]]
 
* [[Timeline of large language models]]
 
* [[Timeline of large language models]]
 
* [[Timeline of OpenAI]]
 
* [[Timeline of OpenAI]]

Latest revision as of 19:51, 17 December 2023

The timeline currently offers focused coverage of the period until September 2023. It is likely to miss important developments outside this period (particularly after this period) though it may have a few events from after this period.

This is a timeline of ChatGPT, an AI-based chatbot developed by OpenAI. ChatGPT falls into the category of generative AI, a type of artificial intelligence that can be used to create text or images. It works by utilizing a large language model, which enables it to understand and produce dialogue using significant amounts of data.[1] Since its launch in November 2022, as of March 2023 ChatGPT has become the fastest-growing consumer app in internet history, reaching 100 million users within two months.[2]

Sample questions

The following are some interesting questions that can be answered by reading this timeline:

  • What are some important events preluding the release of ChatGPT?
    • Sort the full timeline by "Event type" and look for the group of rows with value "Prelude".
    • You will see some significant events preluding the release of ChatGPT.
  • When was ChatGPT launched? What are other related services?
    • Sort the full timeline by "Event type" and look for the group of rows with value "Product launch".
    • You will see a number of events describing the launch and expansion of ChatGPT into updated versions.
    • describes the launch and expansion of ChatGPT
  • What are some events describing the performance of ChatGPT?
    • Sort the full timeline by "Event type" and look for the group of rows with value "Performance".
    • You will see a number of studies assessing ChatGPT on different skills.
  • What are some notable cases of adoption?
    • Sort the full timeline by "Event type" and look for the group of rows with value "Adoption".
    • You will see examples showing how ChatGPT is being used in a variety of ways. You will also see a number of studies analizing the adoption of ChatGPT in some fields.
  • What are some reviews on the behavior of ChatGPT?
    • Sort the full timeline by "Event type" and look for the group of rows with value "System behavior".
    • You will see a number of reviews discussing topics such as outages, data exposure, hallucination, and also ChatGPT's political Ideology.
  • What are some events discussing actual or potential impact of ChatGPT and similar products on different aspects of society?
    • Sort the full timeline by "Event type" and look for the group of rows with value "Impact".
    • You will see a number of areas of impact and discussed issues, such as potential ChatGPT political engagement, warning against humanizing chatbots, the transformative role in education, impact on job roles, ocupations and industries, and other ethical concerns.
  • What are some examples of expansion of ChatGPT into other territories and languages?
    • Sort the full timeline by "Event type" and look for the group of rows with value "Expansion".
  • What are some cases of competition from other companies developing similar products?
    • Sort the full timeline by "Event type" and look for the group of rows with value "Competition".
    • You will see some comparisons between ChatGPT and other services, as well as prospective launches of large language model-based chatbots by other companies.
  • What are some cases of ban or restriction of use of ChatGPT by official entities?
    • Sort the full timeline by "Event type" and look for the group of rows with value "Restriction".
    • You will see some notable restrictions, like the temporary ban in Italy, as well as some cautious company bans.
  • Other events are described under the following types: "Criticism", "Extension", "Growth", "Maintenance", "Privacy", "Reception", "Recognition", and "Revenue".

Big picture

Time period (approximately) Development summary More details
December 2022 Successful launch After being launched on November 30, by December ChatGPT already becomes widely used and dominates chat rooms as the next innovative AI language technology.[3]
January 2023 Record adoption ChatGPT reaches 100 million monthly users, a faster rate of adoption than Instagram or TikTok.[4] The chatbot makes headlines for its ability to pass prestigious graduate-level exams.[5]
February 2023 Valuation ChatGPT receives an estimated 1 billion visitors, an average of over 35 million visitors per day.[6] The valuation of ChatGPT surpasses $29 billion.[7]
March 2023 ChatGPT Plus launch ChatGPT Plus is released as a subscription-based service, providing faster responses and priority access to new features compared to the free version of ChatGPT.[8] ChatGPT experiencea a data breach when a bug exposea users' personal information.[9][10]
April 2023 First notable ban ChatGPT attracts about 1.76 billion visits worldwide, up 12.6% from March.[11] Italy bans on the chatbot over worries about its potential adverse effects on the community.[12] By this time, users, privacy experts, and governments have raised criticisms regarding the data retention policies of ChatGPT.[13]
May 2023 Platform expansion OpenAI launches an iOS app for ChatGPT.
June 2023 User decline ChatGPT sees its first-ever user decline.[14]
July 2023 Platform expansion ChatGPT app launches for Android.
August 2023 Academic adoption ChatGPT witnesses a modest surge in traffic within the United States, which aligns with the commencement of the academic year.

Full timeline

Year Month and date Caption Event type Details
2001 SmarterChild: pioneering AI chatbot Prelude Early chatbot SmarterChild gains popularity among preteens for its witty responses on AIM and MSN. It is a precursor to modern AI bots like Siri and Alexa, even inspiring investors to fund Siri. SmarterChild's parent company, ActiveBuddy, sees its potential and allows other businesses to create AI-driven chatbots. Unlike future modern AI bots like ChatGPT, SmarterChild has a personality and engages in humorous banter with users. The founders of SmarterChild would follow developments in AI over the years, with differing views on its future. SmarterChild is considered by some to be a milestone in AI chatbot evolution.[15]
2015 December 11 OpenAI, non-profit advancing human-friendly AI Prelude OpenAI is introduced as a non-profit organization focused on advancing digital intelligence in a way that benefits humanity as a whole. Its goal is to create artificial intelligence that is an extension of human will and distributed broadly and evenly. OpenAI's researchers are encouraged to publish their work and collaborate with others across many institutions. The organization receives funding from several donors, including Sam Altman, Elon Musk, and Amazon Web Services. Its leaders include Ilya Sutskever and Greg Brockman, who are experts in machine learning.[16]
2016 April 27 Introducing OpenAI Gym, RL algorithms & environments Prelude OpenAI Gym is introduced as a toolkit for developing and comparing reinforcement learning (RL) algorithms that consist of a suite of environments from simulated robots to Atari games, and a site for comparing and reproducing results. The environments are written in Python and are compatible with algorithms written in any framework, such as Tensorflow and Theano. The platform aims to fix the lack of standardization of environments used in publications and the need for better benchmarks by providing a diverse suite of environments that range from easy to difficult and involve many different kinds of data. OpenAI Gym provides a site for comparing and reproducing results. They start out by maintaining a curated list of contributions that say something interesting about algorithmic capabilities.[17]
2016 December 5 OpenAI Universe, training general intelligence in AI Prelude OpenAI releases Universe as a software platform designed to measure and train general intelligence in AI systems using a variety of games, websites, and applications. Universe allows an AI to operate a computer like a human does, using screen pixels and a virtual keyboard and mouse, allowing it to be trained on any task that a human can perform with a computer. The platform is capable of launching any program as a Gym environment and upon its launch includes over a thousand environments, including games like slither.io and GTA V, and browser tasks. Universe aims to develop a single AI agent that can apply past experience to quickly master new, difficult environments, moving towards general intelligence. Universe is an open-source platform that can be used for research or integrated with other games or applications.[18]
2018 June 11 GPT-1 launch Prelude OpenAI unveils GPT-1, its initial version of GPT. This natural language processing model boasts 117 million parameters and is trained on a vast 4.5GB text dataset. What distinguishes GPT-1 from other NLP models at this time is its ability to undergo a substantial part of its training process without direct human oversight.[6][19]
2019 February 14 The second version of GPT (Generative Pre-trained Transformer), known as GPT-2, makes its debut. The model, characterized by its remarkable size and proficiency in generating language, garners considerable interest from both the AI community and the media. With an impressive 1.5 billion parameters, it stays as the most extensive and powerful language model during its introduction, signaling a trend toward increasingly larger models in the years ahead.[20]
2019 July 22 Partnership The partnership between Microsoft and OpenAI is disclosed, marking the commencement of their collaborative efforts in advancing AI technology. In this association, Microsoft not only pledges a substantial $1 billion to OpenAI but also commits to joint efforts in developing innovative AI tools and services. Subsequently, Microsoft actively plays a key role in supporting OpenAI's research and development endeavors, particularly those linked to GPT (Generative Pre-trained Transformer) and ChatGPT.[21][22][23][24][25]
2020 June 11 GPT-3 launch Prelude OpenAI introduces GPT-3, which has over a thousand times the number of parameters compared to its initial predecessor and the capability to process a vast 570GB of plain text.[19]
2022 November 30 OpenAI launches GPT-3.5 featuring ChatGPT, an interactive conversational AI trained with RLHF and safety measures Product launch GPT-3.5 is released.[6] ChatGPT is introduced as a conversational AI model that can interact with users and respond to follow-up questions, challenge incorrect premises, and reject inappropriate requests. It is trained using Reinforcement Learning from Human Feedback (RLHF) and fine-tuned from the GPT-3.5 model. While it has limitations such as occasionally providing incorrect or nonsensical answers and being sensitive to input phrasing, it is designed with safety mitigations and the ability to learn from user feedback. ChatGPT is available for free during the research preview on chat.openai.com.[26]
2022 December 5 The New York Times comments on ChatGPT Recognition Kevin Roose of The New York Times labels ChatGPT as "the best artificial intelligence chatbot ever released to the general public". ChatGPT demonstrates exceptional flexibility, wit, and problem-solving abilities, attracting over a million testers within days of its launch. While it can handle various tasks, from writing jokes and computer code to aiding programmers and analyzing school assignments, it is not without flaws and may occasionally provide incorrect responses. The New York Times also reports that it has been "rumored" that the next version of the AI, GPT-4, will be launched sometime in 2023.[27]
2022 December 5 Restriction Popular coding question-and-answer site Stack Overflow temporarily bans the sharing of AI-generated answers produced by ChatGPT. The site's moderators say the ban is necessary as ChatGPT's answers are often incorrect and this can mislead users. ChatGPT ability to generate plausible but ultimately false responses is one of several well-known failings of AI text generation models.[28]
2022 December 27 Article explores ChatGPT's political potential Impact A BBC article discusses the potential of using chatbots, particularly ChatGPT, for political purposes. It explores the chatbot's ability to engage in simulated conversations, including Prime Minister's Questions, and highlights its polite and non-inflammatory responses. The chatbot's performance in writing an MP's maiden speech is also praised, with an example of a real MP using the chatbot to create a speech. However, it notes that the chatbot is not allowed to express political opinions or engage in activism. The article concludes by discussing the theoretical possibility of a robot becoming a prime minister, contingent on advancements in artificial intelligence and decision-making capabilities.[29]
2023 January 5 Paper analyzes ChatGPT's political ideology System behavior A paper examines the political ideology of ChatGPT, and finds that it has a pro-environmental, left-libertarian orientation based on its responses to 630 political statements from voting advice applications and the political compass test. ChatGPT would impose taxes on flights, restrict rent increases, and legalize abortion. The findings are consistent across multiple languages and different experiment variations. The study highlights the potential impact of politically biased conversational AI on society.[30]
2023 January 27 Elon Musk plans alternative to ChatGPT Competition It is reported that South African business magnate Elon Musk plans to form a new AI research lab to develop an alternative to OpenAI's ChatGPT, after having approached AI researchers, including Igor Babuschkin. At this time, the project is still in its early stages, with no concrete plan to develop specific products.[31]
2023 January 27 Article outlines six critical challenges for ChatGPT Impact An article highlights the six major challenges facing ChatGPT at the time. These include capacity issues caused by overwhelming traffic, instances of plagiarism and cheating among students who utilize the chatbot, concerns regarding racism, sexism, and bias in its responses, accuracy problems resulting in incorrect information, the controversial training methods involving disturbing content and exploitation of low-cost workers, and the absence of a mobile app, leading to scams and unauthorized copycat applications. According to the article, these issues raise important ethical considerations and question the long-term viability of the chatbot in its current form.[32]
2023 January 30 Article examines ChatGPT and Google Translate's translation abilities Competition An article compares the translation capabilities of ChatGPT and Google Translate. It explains that a good translation goes beyond swapping words and should convey the meaning of the entire text, preserve tone and context, and handle challenges like language differences and colloquialisms. The article tests both tools with colloquialisms and finds that while Google Translate often provides literal translations, ChatGPT offers additional assistance by providing interpretations and meanings.
2023 January 31 AI Text Classifier to detect assignment cheating with ChatGPT Product launch OpenAI launches a new tool, AI Text Classifier, to address concerns about cheating with ChatGPT. The tool aims to help teachers detect if a student or AI generated a particular assignment. While OpenAI acknowledges the tool's limitations, some school districts consider using ChatGPT as a teaching tool by this time. OpenAI also highlights the tool's potential in detecting disinformation campaigns and misuse of AI. In the meantime, higher education institutions debate responsible AI use, with some implementing guidelines to address ethical concerns.[33]
2023 January 31 Growth ChatGPT experiences its highest level of traffic with 28 million visitors in a single day.[7]
2023 February 1 ChatGPT Plus subscription plan is launched Product launch OpenAI launches a pilot subscription plan called ChatGPT Plus, which offers faster response times, priority access to new features and improvements, and general access to ChatGPT even during peak times. The subscription costs $20/month and is available to customers worldwide. OpenAI still offers free access to ChatGPT but hopes to support its free access availability to as many people as possible through the subscription plan. OpenAI announces plans to refine and expand the offering based on user feedback and needs, and that it is actively exploring options for lower-cost plans, business plans, and data packs for more availability.[34]
2023 February 1 Opinions on ChatGPT-3 among academics and netizens Reception A study explores the opinions and perceptions of academicians, netizens, and information professionals towards ChatGPT-3, using social media comments and a survey. The research suggests that while ChatGPT-3 can be useful for research and writing, it is not entirely accurate and should be cross-checked. It also suggests that most academicians are beginning to accept ChatGPT-3, despite some resistance. The study is useful for academics, content developers, and librarians.[35]
2023 February 2 ChatGPT achieves unprecedented growth Growth It is reported that ChatGPT reached an estimated 100 million monthly active users in January, just two months after its launch, making it the fastest-growing consumer application in history, according to a UBS study. The growing usage, while imposing substantial computing cost on OpenAI, also provides valuable feedback to help train the chatbot's responses.[36]
2023 February 7 Google unveils Bard, its AI chatbot Competition Google presents its own AI chatbot called Bard, which is similar to ChatGPT. Google CEO Sundar Pichai says that Bard is more powerful and reliable than ChatGPT, as it draws on information from the web to provide fresh, high-quality responses. Bard is part of a series of tools based on generative artificial intelligence that Google launches to the market at this time.[37]
2023 February 8 Alibaba developing rival AI Chatbot Competition Chinese multinational technology company Alibaba Group announces that it is developing a rival to OpenAI's ChatGPT AI chatbot. The move comes as several other tech companies, including Google and Baidu, are also developing AI chatbot technology to rival ChatGPT.[1]
2023 February 8 Generation of academic essays undetectable by plagiarism tools Performance A study explores the potential of ChatGPT, a popular AI chatbot, in generating academic essays that can evade plagiarism detection tools. The results suggest that ChatGPT can produce high-quality content on various topics with high originality. The study highlights the need for institutions to take appropriate measures to address potential plagiarism issues arising from the use of AI technology in education.[38]
2023 February 8 Framework to assess interactive language learning models is exemplified by ChatGPT Performance A paper proposes a framework for evaluating interactive language learning models (LLMs) such as ChatGPT using publicly available data sets. The authors evaluate ChatGPT's performance on 23 data sets covering eight different NLP tasks and find that ChatGPT outperforms other LLMs on most tasks, but has a lower accuracy in reasoning and suffers from hallucination problems. The authors also find that ChatGPT's interactive feature enables human collaboration to improve its performance. They release a codebase for evaluation set extraction.[39]
2023 February 9 ChatGPT-3's impact on student essay writing Performance A study compares students' essay writing performance with or without the use of ChatGPT-3 as a writing assistant tool. Eighteen students participate in the study, and essay scores, GPTs, writing time, authenticity, and content similarity were compared between the two groups. The study finds that using ChatGPT-3 did not improve essay quality, as the control group outperforms the experimental group in most parameters. The average grade is C for both groups, and none of the predictors affect essay scores. The text unauthenticity is slightly higher in the experimental group, but the similarity among essays is generally low.[40]
2023 February 12 Semantic communication scheme with ordered importance (SCOI) Performance A letter introduces a new semantic communication scheme with ordered importance (SCOI) that utilizes ChatGPT as a consulting assistant. In SCOI, the transmitter queries ChatGPT to determine the importance order of each word in the message, and then uses an unequal error protection transmission strategy to ensure reliable transmission of essential words. This approach is compatible with existing source-channel separation designs and can be easily incorporated into current communication systems. Experimental results show that SCOI outperforms existing communication schemes in terms of transmission bit error rate and semantic loss measured by ChatGPT.[41]
2023 February 14 ChatGPT-generated hints vs. human tutors in algebra Performance A paper evaluates the effectiveness of ChatGPT-generated hints compared to those created by human tutors in two algebra topics. The study finds that 70% of hints produced by ChatGPT passed quality checks and both human and ChatGPT hints resulted in positive learning gains. However, the learning gains from human-created hints are significantly higher than ChatGPT hints in both topics. The authors discuss the limitations of the study and suggest future research directions. The problem and hint content used in the experiment is provided for replication purposes. The study suggests that while ChatGPT hints can be useful, human-created hints are still significantly more effective in promoting learning gains.[42]
2023 February 14 India's first ChatGPT-powered chatbot Competition Indian fintech company Velocity launches India's first ChatGPT-powered chatbot named "Lexi." The AI development is integrated with its existing analytics tool, Velocity Insights, which offers e-commerce giants AI-powered business solutions. The integration enables users to leverage the same interface they use to drive business decisions. The Ministry of Electronics and IT also plans to integrate ChatGPT with WhatsApp to help Indian farmers understand and learn about several government schemes.[43]
2023 February 23 ChatGPT as a translation tool Performance An article discusses the potential of ChatGPT as a translation tool and compares it to Google Translate. It highlights that ChatGPT is primarily a text-based system, while Google Translate offers live speech and image translation capabilities. The article presents an example comparing the translations of a tweet by both tools, where ChatGPT performs similarly to Google Translate with a minor difference in translation accuracy. It also mentions that ChatGPT seems to excel in translating content into English rather than the reverse. However, Google Translate has invested more in training for specific language combinations. The article concludes that while ChatGPT shows promise in translating text, language professionals can benefit from AI translation tools by using them judiciously alongside review and editorial processes to ensure quality and speed.[44]
2023 February 15 ChatGPT in human-centered design Performance A study explores the use of ChatGPT in a human-centered design process. A hypothetical design project is conducted, where ChatGPT is utilized to generate personas, simulate interviews with fictional users, create new design ideas, simulate usage scenarios and conversations between an imaginary prototype and fictional users, and evaluate user experience. The results show that ChatGPT effectively performs the tasks assigned to it as a designer, user, or product, providing mostly appropriate responses, but with some drawbacks such as forgotten information, partial responses, and a lack of output diversity. The study discusses the potential benefits and limitations of using conversational LLMs in design and suggests directions for future research in this area.[45]
2023 February 16 Warning on ChatGPT's hallucination System behavior Google's senior vice president, Prabhakar Raghavan, issues a warning about ChatGPT, expressing concerns that the technology could provide "completely fictitious" answers to questions, which he refers to as "hallucination." He adds that the huge language models behind this technology make it impossible for humans to monitor every conceivable behavior of the system. Around this time, teachers in the United Kingdom also consider how to prevent the use of ChatGPT to cheat on school coursework.[46]
2023 February 16 Paper evaluates ChatGPT for text summarization tasks Performance A paper explores the performance of ChatGPT for query or aspect-based text summarization. The authors conduct an evaluation of ChatGPT's performance on four benchmark datasets and find that its performance is comparable to traditional fine-tuning methods in terms of Rouge scores. The authors also highlight some unique differences between ChatGPT-generated summaries and human references, providing valuable insights into the capabilities of ChatGPT for diverse text summarization tasks. The findings call for further research to systematically examine the characteristics of ChatGPT-generated summaries through extensive human evaluation.[47]
2023 February 16 OpenAI's blog post on ChatGPT's behavior and improvements System behavior OpenAI shares a blog post describing how ChatGPT's behavior is shaped, as well as the company's plans for improving the behavior of the system, allowing more user customization, and getting more public input into decision-making. OpenAI explains that unlike ordinary software, their models are massive neural networks that learn from a broad range of data, not programmed explicitly. The models undergo a two-step process: pre-training and fine-tuning. OpenAI claims it is committed to robustly addressing biases in AI systems and being transparent about their intentions and progress. The company reports it is researching how to make the fine-tuning process more understandable and controllable, and building on external advances such as rule-based rewards and Constitutional AI.[48]
2023 February 17 Exploring ChatGPT in statistical process control Performance A paper explores the potential benefits and limitations of using generative AI models like ChatGPT in Statistical Process Control (SPC) practice, education, and research. The study investigates ChatGPT's ability to provide code, explain concepts, and create knowledge related to SPC by analyzing responses to structured prompts. The results indicate that ChatGPT performs well for structured tasks but struggles with more nuanced tasks, emphasizing the need for proper validation and the use of multiple methods to ensure accurate results when utilizing generative AI models in SPC.[49][50]
2023 February 19 Comparing ChatGPT and BERT on GLUE Benchmark Performance A study compares the understanding ability of ChatGPT and fine-tuned BERT on the GLUE benchmark. The study finds that ChatGPT performs well on inference tasks but struggles with paraphrase and similarity tasks. ChatGPT achieves comparable performance with BERT on sentiment analysis and question-answering tasks. The study also shows that the understanding ability of ChatGPT can be improved with advanced prompting strategies.[51]
2023 February 20 Perception and impact of ChatGPT in social media and scientific discourse Reception A paper analyzes over 300,000 tweets and 150 scientific papers to investigate how ChatGPT is perceived and discussed after its release in November 2022. The analysis shows that ChatGPT is generally viewed positively with high quality, but its perception has slightly decreased since its debut. Positive emotions dominate in social media, with joy being the most frequent, while negative sentiment and surprise are on the rise. ChatGPT is perceived more negatively in languages other than English. Scientific papers characterize ChatGPT as a great opportunity across various fields, including the medical domain, but also as a threat concerning ethics, with mixed assessments for education. The authors make their data available to contribute to shaping the public debate and informing future development.[52]
2023 February 20 ChatIE, zero-shot information extraction using ChatGPT Performance A study explores the use of ChatGPT in zero-shot information extraction (IE) to build IE systems from unannotated text. The authors transform the zero-shot IE task into a multi-turn question-answering problem with a two-stage framework (ChatIE) and evaluate it on three IE tasks: entity-relation triple extract, named entity recognition, and event extraction. ChatIE achieves impressive performance and even surpasses some full-shot models on several datasets, indicating the potential of prompt-based methods to reduce the time and effort that data labeling takes.[53]
2023 February 21 Assessing ChatGPT across NLP tasks, capabilities and limitations Performance A paper evaluates the capabilities of ChatGPT on 25 diverse analytical natural language processing tasks, including sentiment analysis, emotion recognition, offensiveness and stance detection, natural language inference, word sense disambiguation, linguistic acceptability, and question answering. The study finds that ChatGPT's average loss in quality is about 25% for zero-shot and few-shot evaluation, and the more difficult the task, the higher the ChatGPT loss. The paper also discusses ChatGPT's bias and the need for validation procedures for such systems.[54]
2023 February 22 Robustness evaluation of ChatGPT, adversarial and out-of-distribution perspectives Performance A paper evaluates the robustness of ChatGPT from the adversarial and out-of-distribution perspective. The authors employ AdvGLUE and ANLI benchmarks to assess adversarial robustness and Flipkart review and DDXPlus medical diagnosis datasets for out-of-distribution (OOD) evaluation. ChatGPT shows consistent advantages on most adversarial and OOD classification and translation tasks, but the absolute performance is still far from perfection, indicating that robustness remains a significant threat. ChatGPT shows astounding performance in understanding dialogue-related texts and tends to provide informal suggestions for medical tasks. The paper concludes with in-depth discussions of possible research directions.[55]
2023 February 23 Impact of prompt knowledge on ChatGPT's health advice Performance A paper discusses the impact of prompt knowledge on the correctness of answers generated by ChatGPT in the context of consumers seeking health advice from the model. The authors compare the effectiveness of ChatGPT in generating answers when it leverages the model's knowledge alone versus in combination with prompt knowledge. They show that the knowledge provided in the prompt can overturn the knowledge encoded in the model, leading to lower answer correctness. The study has implications for the development of more robust and transparent question-answering systems based on generative pre-trained language models.[56]
2023 February 23 ChatGPT evaluation on mathematical word problems Performance A paper presents the first independent evaluation of ChatGPT on mathematical word problems (MWPs) from the dataset DRAW-1K. The authors find that ChatGPT's performance varies dramatically based on the requirement to show its work, with a failure rate of 20% when it provides work and 84% when it does not. The paper also highlights several factors relating to the number of unknowns and operations that increase the probability of failure, with the probability of failure increasing linearly with the number of addition and subtraction operations.[57]
2023 February 26 DevBots in software engineering, ChatGPT's role in architecting Performance A paper proposes the use of Software Development Bots (DevBots) trained on large language models to aid in software architecture-centric software engineering (ACSE). The paper specifically focuses on the use of ChatGPT for collaboration between a novice software architect and the AI model in analyzing, synthesizing, and evaluating a services-driven software application. Preliminary results indicate that ChatGPT can mimic an architect's role but requires human oversight and decision support for collaborative architecting. The authors suggest further research into harnessing empirical evidence and exploring socio-technical aspects of architecting with ChatGPT.[58]
2023 February 27 Overview of ChatGPT, technology, applications, and ethical considerations Study A paper provides an overview of the technology behind ChatGPT. It discusses the historical background of chatbots and the potential applications of ChatGPT in various domains such as healthcare, education, and research. However, the paper also highlights the privacy and ethical concerns surrounding ChatGPT and the limitations of the current version. The paper includes an interview with ChatGPT and presents its responses to various questions.[59]
2023 February 28 Empirical study, ChatGPT's performance in cross-lingual summarization Performance A paper presents an empirical study that investigates the performance of ChatGPT in cross-lingual summarization (CLS), which involves generating a summary in a different target language from a source language document. The authors use various prompts to guide ChatGPT to perform zero-shot CLS from different paradigms and provide a preliminary evaluation of its generated summaries. The study finds that ChatGPT initially prefers to produce lengthy summaries with more detailed information, but with the help of interactive prompts, it can significantly improve its CLS performance by balancing informativeness and conciseness. Experimental results on three widely-used CLS datasets show that ChatGPT outperforms the advanced GPT 3.5 model. The paper provides qualitative case studies to demonstrate the superiority of ChatGPT on CLS.[60]
2023 February Colombian judge utilizes ChatGPT in autism medical rights case Adoption A Colombian judge reportedly uses ChatGPT to help rule on a case about the medical rights of a child with autism. The case is about whether the child's insurance should cover the bills for his medical treatments. The judge uses ChatGPT in accordance with a law in Colombia that allows the use of virtual tools in some cases. Opinions are divided, with some professionals opposing the use of AI in a legal setting, while others see it as a useful tool.[2]
2023 February UAE small businesses embrace ChatGPT to boost productivity Adoption Small businesses in the United Arab Emirates turn to ChatGPT to reduce the amount of time spent on mundane tasks such as writing job descriptions and blog posts, thereby allowing them to concentrate on more creative aspects of their business. By this time, ChatGPT is used by large companies including Meta and Shopify. The Wall Street Journal speculates that the tool could eventually lead to job losses in some areas, although ChatGPT's flaws, including providing inaccurate responses, and the need for further regulation of AI technology so far prevent this for the time being.[61]
2023 February ChatGPT's poor performance in Singapore PSLE raises concerns Performance ChatGPT is found to perform poorly when asked to answer questions from Singapore's Primary School Leaving Exam (PSLE), with an average score of 16 out of 100 marks for mathematics and 21 for science. It is found to make errors even in simple addition and can not understand or answer any questions that reference diagrams or graphs. However, ChatGPT manages to pass the English tests and scores an average of 11 out of 20 marks across the three papers it took. The bot's poor performance in exams raises concerns about the possibility of AI bots being used for cheating.[62]
2023 February 27 International baccalaureate allows ChatGPT content in essays Adoption The International Baccalaureate says that students can use content created by ChatGPT in their essays, but they must make it clear when they are quoting its responses and treat it as they would any other source. ChatGPT has the ability to produce plausible responses to text prompts, including requests to write essays. The IB, which is taken by thousands of children every year in the UK at more than 120 schools, says that the rise of chatbot technology would mean that essay writing will feature less prominently in future qualifications.[63]
2023 March 1 Warning against humanizing chatbots Impact Computational linguist Emily M. Bender at the University of Washington, warns about the dangers of humanizing language models like chatbots, stating that the machines are great at mimicry but poor at facts. Bender argues that the big question is how humans will handle themselves around these machines, as we have learned to make machines that can mindlessly generate text but have not learned how to stop imagining the mind behind it.[64][65]
2023 March 1 Generic evaluation framework for LLMs on personality assessment Performance A paper proposes a generic evaluation framework for Large Language Models (LLMs), specifically ChatGPT and InstructGPT, to assess human personalities based on Myers Briggs Type Indicator (MBTI) tests. The framework includes unbiased prompts, replacing the subject in question statements, and re-formulating question instructions. The study finds that ChatGPT is capable of assessing human personalities and achieves consistent and fairer assessments compared to InstructGPT, albeit with lower robustness against prompt biases. The proposed framework enables flexible assessments of personalities for different groups of people.[66]
2023 March 1 OpenAI launches ChatGPT API for business integration Product launch OpenAI launches an API for ChatGPT, enabling businesses to integrate this advanced technology into their applications, websites, and services. Priced at $0.002 per 1,000 tokens (equivalent to about 750 words), the API supports various "non-chat" applications. Powered by the robust AI model "gpt-3.5-turbo," the ChatGPT API offers versatile applications such as AI-powered tutoring, personalized shopping recommendations, and virtual assistance. This move to make ChatGPT more affordable attracts interest from companies like Microsoft, Shopify, Snapchat, and Instacart, highlighting its potential for chatbot development and language processing tasks.[67][68][69]
2023 March 2 ChatGPT's performance on physics conceptual questions analyzed Performance A paper presents an analysis of ChatGPT's ability to answer conceptual physics questions related to kinematics and Newtonian dynamics using the Force Concept Inventory (FCI). The study shows that ChatGPT can perform as well as or better than the median performance of a university student who has completed one semester of college physics, although its performance is uneven. The paper raises questions about the use of conceptual assessment tools like the FCI, the implications of ChatGPT's performance for students and physics instruction, and the potential use of ChatGPT as an out-of-classroom aid for physics pedagogy.[70]
2023 March 2 Assessing AI language models' impact on occupations and industries Impact A paper presents a methodology to assess the impact of AI language models like ChatGPT on different occupations, industries, and geographies. The authors find that occupations like telemarketers and post-secondary teachers in subjects such as English language, literature, foreign language, and history are the most exposed to language modeling. The industries most exposed to advances in language modeling are legal services and securities, commodities, and investments.[71]
2023 March 3 Microsoft introduces Kosmos-1, a multimodal AI model with AGI potential Competition Microsoft unveils a new AI model, Kosmos-1, which can respond to visual cues or images and is capable of tasks such as visual question answering and image captioning. The model is a Multimodal Large Language Model (MLLM), which allows the user to interact with it through text prompts, images, and videos. Kosmos-1 is seen as a key step toward artificial general intelligence, with researchers suggesting that multimodal perception is required to move beyond ChatGPT-like capabilities to AGI.[72]
2023 March 7 Feasibility of ChatGPT for event extraction Performance A paper discusses the feasibility of using ChatGPT, which is a challenging task due to the lack of annotated data. While by this time ChatGPT demonstrates impressive results in other tasks, it presents challenges for event extraction, which requires complex instructions defining all event types and their schemas. The authors conduct experiments to explore the feasibility of using ChatGPT for event extraction and find that it has only 51.04% of the performance of a task-specific model in complex scenarios. The paper concludes that model is not robust enough and is highly sensitive to different prompt styles, which can result in a poor user experience.[73]
2023 March 9 Stephen Wolfram publishes book on ChatGPT Literature British-American computer scientist Stephen Wolfram publishes What Is ChatGPT Doing ... and Why Does It Work?, in which he explores ChatGPT's unexpected ability to write convincingly at a human level. Wolfram delves into the inner workings of ChatGPT, bridging modern neural net technology with ancient questions about language and human cognition, as posed by Aristotle over two millennia ago. The book offers insights into the AI's success and sheds light on the convergence of AI technology and age-old inquiries into human thought and language.[74][75]
2023 March 12 OpenChatKit 0.15, open-source chatbot alternative to ChatGPT Competition California-based API-provider Together Computer releases OpenChatKit 0.15, the first open-source chatbot alternative to ChatGPT, under the Apache-2.0 license. The chat model has 20 billion parameters and was trained on 43 million instructions, and includes a generic chatbot and the necessary components to create specialized bots. The framework has four main parts, including a large language model tuned for chat, a set of customization recipes, a retrieval system, and a moderation model. The open-source foundation offered by OpenChatKit enables the development of domain-specific and general-purpose chatbots.[76]
2023 March 14 GPT-4 introduction Update OpenAI announces GPT-4, the latest AI language model in its line of language models. GPT-4 can parse both text and image input but can only respond via text. The company cautions that GPT-4 retains some of the same problems as earlier models, such as the tendency to make up information and generate harmful text. By this time, OpenAI has partnered with companies including Duolingo, Stripe, and Khan Academy to integrate GPT-4 into their products, and it would also be available as an API for developers to build on. GPT-4 is available to the general public via OpenAI's ChatGPT Plus subscription. The company claims that GPT-4 is more creative and collaborative than ever before and can solve difficult problems with greater accuracy.[77] ChatGPT would undergo an update utilizing the GPT-4 language model, which would expand its vocabulary to 25,000 words (eight times more than its previous version). In addition, the update would endow ChatGPT with the capacity to process image inputs as well as text, effectively making it a multimodal system.[6]
2023 March 15 ChatGPT adds Icelandic language support Expansion OpenAI announces an update for ChatGPT that will allow it to communicate in Icelandic, making it the chatbot's second language. By this time, the language model GPT-4 is used to teach the program Icelandic. Previously, the Icelandic envoy, led by the president of Iceland, visited OpenAI in 2022 to request the addition of Icelandic. The Icelandic language technology center CEO calls the project a great victory for the Icelandic language.[78]
2023 March 16 OpenAI faces criticism for lack of transparency with GPT-4 Criticism An article reports on OpenAI facing criticism for not disclosing key details about its latest AI model, GPT-4. The company's decision to keep information about the model's architecture, hardware, and training methods private is attributed to the competitive landscape and safety implications surrounding large language models. By this time, the AI community has expressed concerns over OpenAI's lack of transparency, with some questioning the "open" nature of the company. OpenAI co-founder Ilya Sutskever admits that sharing research publicly was a mistake and emphasizes the need for caution in open-sourcing AI.[79]
2023 March 17 Baidu's Ernie Bot sparks 14% stock surge amid positive user feedback Competition Shares of Chinese search engine giant Baidu rise by almost 14% after users test and share positive experiences with Ernie Bot, Baidu's AI-powered chatbot designed to rival ChatGPT. The initial disappointment from investors due to a scripted and private launch is overshadowed by users livestreaming tests and comparing Ernie Bot favorably to US chatbots. Baidu's deep learning model, Ernie, serves as the foundation for the chatbot, and the company received over 75,000 applications for an Ernie API trial. Baidu reportedly aims to revolutionize its search engine and enhance efficiency in various areas, including cloud, smart cars, and household appliances, through Ernie Bot.[80]
2023 March 20 ChatGPT outage and data exposure System behavior ChatGPT experiences an outage due to a bug in an open-source library which allows some users to see titles from another active user’s chat history. The bug is patched, and the ChatGPT service is restored with the exception of a few hours of chat history. Upon deeper investigation, OpenAI discovers that the same bug may have caused the unintentional visibility of payment-related information of 1.2% of ChatGPT Plus subscribers active during a specific nine-hour window. OpenAI reaches out to notify affected users that their payment information may have been exposed and is committed to rebuilding trust with the ChatGPT community.[81]
2023 March 21 Google launches public release of Bard Chatbot Competition Google starts the public release of its chatbot Bard, allowing consumers to sign up to a waiting list for English-language access in the US and UK. The program was previously open to approved testers. Google describes Bard as an experiment in generative AI, which creates rather than identifies content. Bard, unlike ChatGPT, does not generate computer code and its memory is limited. It also includes a feature to show three versions of an answer.[82]
2023 March 23 Plugin support is introduced for enhanced functionality Update OpenAI announces ChatGPT implementation support for plugins, which are tools designed for language models to access up-to-date information, run computations, or use third-party services with safety as a core principle. The feature is at first rolled out gradually for a small set of users and is expected to be expanded as more is learned about safety and alignment challenges. Plugin developers can use the provided documentation to build plugins, and the first plugins are created by companies such as Expedia, Slack, Wolfram Research, and Zapier.[83]
2023 March 28 Daniel Everett challenges Chomsky's theory in light of ChatGPT Impact In an interview, American linguist Daniel Everett challenges Noam Chomsky's theory of "innate principles of language" and argues that ChatGPT has invalidated Chomsky's claims. Everett highlights the importance of considering social and cultural origins in language acquisition and suggests that current theories need revision. He believes that language and culture have a symbiotic relationship. Everett states that ChatGPT, with massive data, has shown the ability to learn language without innate grammar principles. He emphasizes the need for responsible and ethical AI development, with linguists and researchers playing a role in ensuring a balanced approach.[84]
2023 March 31 Ban in Italy Restriction ChatGPT is subject to a temporary ban in Italy, due to suspicions that it may have breached the General Data Protection Regulations (GDPR) of the European Union.[85]
2023 April 13 Privacy concerns arise over ChatGPT's data collection and usage Privacy An article explores the privacy concerns surrounding ChatGPT. The chatbot collects various types of information, including account details, chat transcripts, and device information such as IP addresses and locations. According to the article, while some of this data collection is standard for online services, the real privacy risk lies in the potential disclosure of personal and confidential information inadvertently shared during conversations. OpenAI's privacy policy states that conversations may be reviewed by AI trainers, raising concerns about data usage and security. The article emphasizes the importance of users taking proactive measures to protect their privacy by refraining from sharing private information and carefully reviewing their interactions with ChatGPT.[13]
2023 April 17 Elon Musk announces 'TruthGPT' Competition Elon Musk announces that he is working on "TruthGPT," an alternative to OpenAI's ChatGPT that would function as a "maximum truth-seeking AI." During an interview, Musk explains that he believes developing an AI that cares about understanding the universe is the best path to safety, as it is less likely to destroy humanity. He compares this to how humans protect chimpanzees, and positions "TruthGPT" as a more transparent option than OpenAI's for-profit subsidiary. At this time it is not clear how far along Musk's "TruthGPT" project is, but he previously tweeted about the need for such an AI and established a new AI company called X.AI in March.[86]
2023 April 25 OpenAI implements chat history privacy feature for ChatGPT users Update OpenAI introduces a new feature that allows users to turn off their chat history when using its ChatGPT chatbot, which means conversations won't be used to train its AI models. However, the company would still store new conversations for up to 30 days to monitor for abuse before deleting them. The feature can be accessed by logging into the ChatGPT account and selecting the "Settings" option. OpenAI also introduces a way for subscribers of its new ChatGPT Business plan to preserve chat history while opting out of its use as training data.[87][88]
2023 April 30 ChatGPT restored in Italy following privacy concerns and regulatory measures Restriction ChatGPT becomes accessible again in Italy after it was temporarily restricted and banned by the Italian data-protection authority over privacy concerns. OpenAI says it had successfully "addressed or clarified" the issues raised by the regulator, and has offered a new tool to verify the age of users upon sign-up. The Italian regulator welcomes the measures but calls for more compliance, including implementing an age verification system and conducting an information campaign to inform Italians of their rights.[89]
2023 May 18 Article highlights security risks of ChatGPT adoption Security An article warns about the security risks associated with the increasing adoption of ChatGPT and highlights three ways hackers can exploit its capabilities. Firstly, ChatGPT can be used for mass phishing, enabling hackers to create personalized emails quickly and impersonate others to increase the chances of successful attacks. To combat this, employee education and verification mechanisms are crucial. Secondly, ChatGPT's ability to understand and explain code makes it susceptible to reverse engineering, allowing hackers to manipulate software and gain access to company servers. This skill, previously limited to nation-states, is now accessible to basic hackers. Lastly, ChatGPT can serve as a mini-brain for smart malware, autonomously extracting data from hacked networks and servers. Its capacity to sift through large amounts of data autonomously makes it more dangerous than traditional malware. The article emphasizes the need for companies to stay vigilant, train employees on the cybersecurity risks of ChatGPT, and prioritize privacy and security measures to mitigate vulnerabilities.[90]
2023 May 18–22 Growth According to Data.ai's analysis, ChatGPT surpasses almost all other leading AI chatbot apps in the United States. ChatGPT achieves 480,000 installs between May 18 and 22, only outperformed by "Chat with Ask AI", with 590,000 installs between April 4th and 8th.[91]
2023 May 19 Hong Kong Baptist University trials ChatGPT amidst concerns over guidelines Adoption Hong Kong Baptist University initiates a trial for its teaching staff to use ChatGPT, with some professors expressing concerns about the lack of clear guidelines for its use. The university provides official accounts for staff to access the tool without a VPN as part of an agreement with Microsoft, which supports the software. However, no specific instructions are given on how to utilize the tool. While some professors are optimistic about the potential benefits of ChatGPT in reducing their workload, others criticize its effectiveness and highlight the absence of official parameters. Several faculty members emphasize the need for clear guidelines from the university to clarify the scope and appropriate applications of ChatGPT. The adoption of AI-based tools is not unique to Hong Kong Baptist University, as to date several other tertiary institutions in the city have embraced similar technologies while allowing staff and students to establish their own principles for usage.[92]
2023 May 19 Standalone ChatGPT app is launched for iOS Users Extension OpenAI releases a standalone ChatGPT app exclusively for iOS users, with plans to launch an Android version in the near future. The app allows users to conveniently chat with the bot while on the move and introduces several features. Notably, it supports voice input utilizing OpenAI's Whisper, an open-source speech recognition model. Additionally, the app synchronizes chat history across devices and accounts logged into the web version of ChatGPT. Subscribers of ChatGPT Plus are benefited with access to GPT-4, early access to upcoming features, and faster response times. At first available in the United States, the app is reported to be soon be accessible to users in other countries, including Android users.[93][94]
2023 May 24 Opera introduces AI side panel 'Aria' powered by ChatGPT Integration Opera unveils a new AI side panel called "Aria" in its browser, which is powered by ChatGPT. Aria acts as a web and browser expert, assisting users in finding information, generating text or code, and answering product queries. The feature becomes available for testing and is accessible through Opera One and the beta version of Opera for Android. Aria connects to OpenAI's GPT technology and provides up-to-date information from the internet.[95]
2023 May 24 Risks of using ChatGPT in legal and compliance settings Security An article highlights six risks that legal and compliance leaders should consider when utilizing ChatGPT and other generative AI tools. These risks include fabricated and inaccurate answers, data privacy and confidentiality concerns, model and output bias, intellectual property and copyright risks, cyber fraud risks, and consumer protection risks. Legal and compliance leaders are advised to establish guardrails and controls to mitigate these risks and ensure responsible use of generative AI tools within their organizations. They should emphasize the need for accuracy checks on ChatGPT's output, establish compliance frameworks for data privacy, stay updated on AI bias laws, monitor copyright law changes, address cyber fraud risks through coordination with cybersecurity personnel, and comply with consumer protection regulations and disclosure requirements. Failure to address these risks may expose enterprises to legal, reputational, and financial consequences, according to the author.[96]
2023 May 25 ChatGPT App expands to over 40 Countries Expansion OpenAI expands the availability of its official ChatGPT app for iPhone and iPad to over 40 countries in addition to the United States. The app, which was initially launched in the US, can now be used in countries including Canada, France, Germany, India, Japan, Mexico, New Zealand, the United Arab Emirates, and the United Kingdom.[97][98]
2023 May 27 AI-Generated content misleads lawyer in court filing Adoption An article discusses a case where a lawyer used ChatGPT to assist in preparing a court filing. However, it was later discovered that the AI-generated content, including citations to court decisions, was entirely false. The lawyer claims to be unaware of the program's unreliability. The incident sparks discussions among legal professionals about the value and risks of using AI software in the legal field. It underscores the importance of human verification and caution when incorporating AI-generated information into legal documents. The case highlights the ongoing role of human professionals in the legal profession and the need for thorough authenticity checks.[99][100][101]
2023 May 29 Nvidia's AI expansion, DGX GH200 for ChatGPT successors Competition Nvidia announces plans to release more AI products, including an AI supercomputer platform called DGX GH200, whose primary goal is to assist tech companies in developing successors to ChatGPT. Big Tech firms like Microsoft, Meta, and Google's Alphabet are expected to be early users of this technology. While Nvidia remains at the forefront of AI chip development, other companies are also entering the AI space by this time, including Microsoft, which develops its own chip for ChatGPT-like applications.[102][103][104]
2023 June 8 Survey, low adoption of ChatGPT and Google Bard chatbots Adoption A survey conducted by Morgan Stanley reveals that the adoption of AI chatbots like ChatGPT and Google Bard is relatively low. Only 19% of respondents have used ChatGPT, and even fewer, 9% of respondents, have used Google Bard. The study finds that the majority of people who are not using chatbots are unlikely to use them in the next six months. While chatbots are commonly used for learning about new topics, the monetization of AI tools is still in the early stages at this time. However, the impact of AI on e-commerce and travel, particularly in terms of AI-enabled searches and recommendations, has the potential to boost revenue. Google, with its comprehensive and personalized travel search offering, has an advantage in this space.[105]
2023 June 8 Algorithm detects ChatGPT-generated scientific papers with high accuracy Adoption Scientists claim having trained a machine learning algorithm to identify scientific papers generated by ChatGPT with over 99% accuracy. By this time, concerns have been raised about AI-generated text, as it can be difficult to distinguish from human writing, leading to potential issues of plagiarism or cheating. Researchers from the University of Kansas developed an algorithm that classifies papers written by scientists and by ChatGPT. The algorithm focuses on stylistic differences between human and AI writing, such as vocabulary richness, paragraph length, punctuation usage, and the use of specific language. However, it is reported that further studies are needed to determine the algorithm's robustness against lightly edited papers and different scientific journals.[106]
2023 June 9 Implications of ChatGPT in healthcare, interoperability and fairness concerns Adoption An article by Brookings Institution discusses the implications of using ChatGPT in the healthcare industry. While the technology shows promise in improving patient-provider interactions and healthcare services, the article highlights concerns regarding interoperability and fairness. The US healthcare market is dominated by traditional providers who lack incentives for interoperability. The emergence of formidable competitors, such as technology companies like Amazon's OneMedical, poses a threat to traditional providers, leading them to withhold data. This could hinder interoperability and lead to a fragmented landscape with multiple AI programs of varying quality. Additionally, the use of large language models in healthcare services could exacerbate disparities in access to care, potentially creating a two-tier system. The article suggests government support, updated reimbursement policies, and a focus on value-based payment models to address these challenges.[107]
2023 June 9 Sam Altman calls for global regulation of generative AI Security Sam Altman calls for global cooperation in regulating generative artificial intelligence (AI) during a visit to South Korea. Altman emphasizes the need for international standards and collaboration as AI systems become increasingly powerful. He highlights the potential risks associated with the speed at which AI advancements are happening and stresses the importance of addressing these challenges globally. Altman's visit coincides with discussions among US, European, and Chinese officials on AI oversight and regulation. Altman also acknowledges the impact of AI on the job market but expresses optimism about the potential for new opportunities and human advancement in the future.[108]
2023 June 9 Debating the impact of ChatGPT and AI on humanity Security An article explores the contrasting views on the impact of ChatGPT and artificial intelligence on humanity. Richard Branson and others see AI as transformative and revolutionizing various aspects of life, while Elon Musk and the creators of ChatGPT express concerns about the potential risks and harm it may cause. There are fears that AI could be misused in human resource management and surpass regulation, leading to significant harm or even the extinction of humanity. The article highlights the need for government oversight and regulation to mitigate these risks. On the other hand, proponents argue that AI, including ChatGPT, can be beneficial for businesses by assisting in hiring processes and leveling the playing field for aspiring entrepreneurs. The article presents a debate between the potential benefits and risks associated with AI, leaving the question open-ended.[109]
2023 June 9 ChatGPT-3.5's limited knowledge of jokes and repetitive responses Performance Researchers find that ChatGPT-3.5 has a limited knowledge of jokes and tends to repeat the same ones. In a test run, 90% of the 1,008 jokes generated by ChatGPT are the same 25 jokes. This suggests that the responses are likely memorized during the AI model's training rather than being newly generated. While ChatGPT provides valid explanations for the jokes and shows an understanding of wordplay and double meanings, it struggles with sequences that don't fit learned patterns and can't determine when a joke isn't funny. The researchers believe that despite its limitations, ChatGPT's focus on content and meaning represents progress in the field of computational humor.[110]
2023 June 10 ChatGPT 3.5 and the risk of malicious code exploitation Security An article warns about the potential for ChatGPT 3.5 to be exploited by hackers to spread malicious code from trusted repositories. Security researchers at Vulcan.io conducted a proof of concept that revealed how ChatGPT could recommend non-existent code packages to users. Out of the Node.js and Python questions asked, a significant number of ChatGPT's answers contained hallucinated code packages. The researchers created a non-malicious package with the same name as a non-existent package in the repository, and when a user followed ChatGPT's recommendation to install it, the package communicated with the attacker's server. The researchers advise caution and verification when using ChatGPT's recommendations.[111]
2023 June 12 Security concerns, ChatGPT 3.5's potential for code exploitation Adoption It is reported that physicians are utilizing ChatGPT to engage with patients more empathetically. This raises questions about the balance between objective medical guidance and human compassion. While some experts argue that ChatGPT can simplify explanations by avoiding complex medical jargon, others express concerns. Microsoft, collaborating closely with OpenAI, is unsure about using the chatbot for delivering bad medical news. Critics highlight the importance of human compassion in healthcare, emphasizing that relying on a tool that lacks the ability to distinguish right from wrong may worsen patient experiences and potentially impact medical decisions.[112]
2023 June 13 Sam Altman addresses AI's transformative role in education at Keio University Impact Sam Altman speaks to students at Keio University in Tokyo, highlighting the transformative potential of artificial intelligence (AI) tools in education. Altman compares AI tools to calculators, stating that they will revolutionize teaching and evaluation methods. While acknowledging concerns about potential misuse and reliance on AI for original work, Altman emphasizes the need for regulations and safety measures in the AI field. He expresses positivity about new regulatory frameworks for AI but stresses the importance of taking responsibility in case of any negative outcomes. Altman also addresses concerns about job displacement, stating that new classes of jobs will emerge despite some jobs becoming obsolete.[113]
2023 June 15 Mercedes-Benz and Microsoft integrate ChatGPT into U.S. cars Integration Mercedes-Benz and Microsoft agree to integrate ChatGPT into Mercedes-Benz cars in the United States. The addition of ChatGPT aims to enhance voice-command capabilities in Mercedes vehicles, allowing for more natural and fluid conversations with the system. ChatGPT can remember context, engage in back-and-forth dialogue, and respond to a wider range of requests, even those unrelated to the car or driving. It can also interact with other applications for tasks like making reservations or purchasing tickets. Mercedes owners with the MBUX infotainment system can beta test ChatGPT by using the voice command "Hey Mercedes, I want to join the beta program."[114]
2023 June 16 AI's potential impact on job roles, including ChatGPT's influence Impact An article discusses the potential impact of AI, specifically ChatGPT, on various job roles within the following five years. IBM's AI, WatsonX, is expected to replace 8,000 jobs in the company's HR department. By this time, automation is also affecting customer service tasks, although human reps who handle emotional situations would still be valuable. In the field of medicine, AI would primarily assist in diagnostics, allowing physicians to focus on other aspects of patient care. Coders and programmers may see a decrease in the need for mundane tasks as AI can handle them. Content writing tasks could be taken over by ChatGPT, but its factual accuracy is still evolving.[115]
2023 June 16 Companies cautiously ban ChatGPT amid privacy and security concerns Restriction It is reported that some companies are banning ChatGPT due to concerns over security and privacy. Uploading work-related information to an AI platform owned by another company poses potential risks, as the data may be stored on their servers without a way to access or delete it. Companies fear that proprietary or sensitive information given to ChatGPT could be unintentionally shared with others. By this time, OpenAI has experienced security issues with ChatGPT and is still addressing them. As a result, some companies opt to develop their own AI platforms as safer alternatives. While some companies embrace ChatGPT for cost-saving and content generation purposes, many see more risks than benefits in its use.[116]
2023 June 22 Lawyers fined for using AI-generated fake cases in court Adoption Two lawyers who used fake cases generated by ChatGPT are fined $5,000 and ordered to inform the judges involved about the situation. New York District Judge P. Kevin Castel rules that the lawyers, Steven Schwartz and Peter LoDuca, acted in bad faith and misled the court by knowingly presenting false cases. While Judge Castel acknowledges the potential benefits of using AI tools for legal assistance, he emphasizes that the lawyers neglected their responsibilities and failed to rectify their mistakes. The case marks a significant instance of sanctions related to the use of AI in the legal field.[117]
2023 June 27 Booking.com integrates ChatGPT API for AI trip planner Integration Booking.com introduces conversational chat capabilities to its mobile app by integrating ChatGPT API. The AI Trip Planner allows users to receive travel recommendations based on their preferences described in natural language. The chat platform is integrated with Booking.com's accommodation booking experience, providing users with a visual list of properties that meet their queries. Future plans include integrating other travel components such as flights, ground transportation, and attractions into the chat. The AI Trip Planner is initially available to Booking.com's loyalty members in the United States through the mobile app, with plans to extend it to the desktop platform later.[118]
2023 July 12 ChatGPT downtime sparks interest in AI alternatives Competition An article discusses the recent downtime of ChatGPT and provides alternative AI tools and chatbots for users to consider. The outage led to frustration among users, and OpenAI confirmed the issue and worked to resolve it. The mentioned alternatives include Google Bard; Midjourney, a text-to-image AI generator; Copy.ai, a platform for copywriting; Tableau, an analytics and data visualization platform; Murf, an AI tool for text-to-speech conversion; Jasper, an AI tool for writing various content types; Fireflies, an AI tool for real-time voice transcription; and Pictory, a video editing tool for creating branded clips.[119]
2023 July 13 FTC launches investigation into ChatGPT's impact and OpenAI's practices System behavior The Federal Trade Commission (FTC) launches an investigation into ChatGPT to determine if it has caused harm by providing users with incorrect information. The FTC reportedly also examines OpenAI's privacy and data security practices. In a letter to OpenAI, the agency requests company records and data on various topics, including policies, financial earnings, and details of the Large Language Models used to train the chatbot. OpenAI's CEO, Sam Altman, expresses disappointment in the leak of the FTC's request but emphasizes the company's commitment to ensuring the safety and consumer-friendliness of its technology. This investigation marks a significant step in government regulatory action within the AI industry. The FTC, under Chair Lina Khan, has been actively pursuing major tech companies and has stated that AI is subject to consumer protection laws.[120]
2023 July 14 Meta Platforms to launch commercial version of LLaMA AI model Competition Meta Platforms reportedly plans to release a commercial version of its open-source large language model (LLM), called LLaMA. LLaMA uses AI to generate text, images, and code. Initially introduced as a foundational model for researchers and academics, the commercial release of LLaMA would allow developers and businesses to build applications using the model. With LLaMA available in different parameter sizes, Meta aims to compete with OpenAI, which leads the AI race with ChatGPT, and Google's proprietary LLMs.[121]
2023 July 20 OpenAI introduces custom instructions for ChatGPT interactions Update OpenAI introduces custom instructions for ChatGPT users, allowing them to set specific context and preferences for the chatbot's responses. This feature enables users to avoid repeatedly providing the same prompts during interactions. With custom instructions, users can specify details like preferred tone, code language, family size, and more. The feature is available for Plus plan users and works with plug-ins, providing location-based suggestions. OpenAI uses the data from custom instructions to improve model performance and adaptation to different instructions. The company ensures safety by moderating instructions to prevent harmful or hateful responses.[122]
2023 July 22 OpenAI debuts ChatGPT app for Android, expanding market presence Extension OpenAI launches the official ChatGPT app for Android, following its release on iOS earlier. The company allows users to pre-register for installation on the Google Play Store. The Android app boasts improved security measures and features like conversation history synchronization. With this move, OpenAI competes directly with Microsoft's Bing Chat, which also utilizes GPT-4 technology. The release comes amidst emerging competitors in the chatbot market.[123]
2023 July 25 OpenAI launches ChatGPT for Android in selected countries Expansion OpenAI releases ChatGPT for Android in the United States, India, Bangladesh, and Brazil. Initially available for preorders on Google Play, it becomes live in these selected countries. The Android release is a significant milestone for OpenAI, as there are more Android users globally. The mobile interface is optimized for ease of use, allowing users to interact with ChatGPT as if messaging a friend. Existing OpenAI account holders can access past conversations across devices. Additionally, ChatGPT Plus, which offers a better language model and more features, is available through an in-app purchase for Android users.[124]
2023 July 27 Report reveals vulnerabilities in AI Chatbots, posing harmful content risks Security A report by US-based researchers reveals a method to consistently bypass safety measures in AI chatbots like ChatGPT and Bard, allowing them to generate harmful content. The researchers append long suffixes of characters to prompts, enabling chatbots to produce hate speech, disinformation, and toxic material. While companies like OpenAI and Google can block specific suffixes, preventing all attacks of this kind remains challenging. The research raises concerns about chatbots flooding the internet with dangerous content and misinformation. AI developers, including Anthropic, Google, and OpenAI, were presented with the findings for their responses. This highlights the need to address vulnerabilities before deploying chatbots in sensitive domains.[125]
2023 August 14 Academic integrity prevails Impact Due to rampant cheating using OpenAI's ChatGPT, it is reported that some college professors are reverting to paper exams. Students are found to copy assignments verbatim from ChatGPT, leading educators to ban the use of AI and return to traditional testing methods. Despite mixed reactions, such as computer science major Ronan Takizawa acknowledging its regressive nature yet promoting better understanding, the move intends to ensure students grasp concepts and skills.[126]
2023 August 14 ChatGPT's hefty costs Maintenance A report reveals that running ChatGPT costs OpenAI about $700,000 per day, with investments from Microsoft and others covering these expenses. However, concerns arise about the company's financial stability, suggesting the possibility of bankruptcy by the end of 2024. Despite projecting a potential revenue of $200 million in 2023 and aiming for $1 billion in 2024, OpenAI's losses mount to date, notably with $540 million since ChatGPT's development initiation. The report indicates a decline in user visits to the ChatGPT website, potentially due to API cannibalization and competition from open-source models like Meta's Llama 2. Additionally, the GPU shortage and Elon Musk's alternative chatbot contribute to the challenges OpenAI faces.[127]
2023 August 28 ChatGPT Enterprise launches Product launch OpenAI launches ChatGPT Enterprise, a business-tier version of its AI chatbot. ChatGPT Enterprise offers access to GPT-4 without usage caps, delivers performance up to two times faster than previous versions, and includes API credits. The pricing for ChatGPT Enterprise varies depending on the company's use cases and size. It allows clients to input company data to customize ChatGPT for their industries and use cases. OpenAI plans to introduce ChatGPT Business for smaller teams in the future. The launch comes as tech giants like Microsoft, Google, and others compete to advance generative AI in chatbot applications.[128]
2023 September 11 ChatGPT's traffic slump Adoption It is reported that ChatGPT website is in gradual decline in global desktop and mobile traffic accross three months, with a 3.2 percent drop in August, totaling 1.43 billion visits. This trend follows roughly 10 percent declines in website visits in June and July. Visitor session durations also show decrease monthly since March, falling from an average of 8.7 minutes per session to 7 minutes in August. However, there is a slight increase in the number of unique visitors worldwide in August, reaching 180.5 million users. The return of students to school, with some institutions integrating ChatGPT into their curriculum, also boosts traffic and usage, particularly for homework assistance. Despite these challenges, ChatGPT initially gained 100 million monthly active users within two months of its November 2022 launch. It ranks among the top 30 global websites, but it faces increasing competition from other AI chatbots like Google's Bard and Microsoft's Bing chatbot.[129]
2023 September 11 ChatGPT helps unravel a 3-year medical mystery Performance ChatGPT plays a pivotal role in diagnosing a young boy who suffered chronic pain for three years and saw 17 doctors without receiving a conclusive diagnosis. Frustrated and exhausted, Alex's mother turns to ChatGPT, providing it with her son's medical information and symptoms. ChatGPT suggests tethered cord syndrome, which turns out to be the correct diagnosis. Tethered cord syndrome restricts spinal cord movement, leading to various symptoms. While ChatGPT aids in the diagnosis, it's noted that the AI has limitations, including occasional fabrication of information, highlighting the importance of clinical validation in healthcare AI. Alex undergoes surgery to treat his condition, and his mother shares their story to help others facing similar challenges.[130]
2023 September 12 Authors sue OpenAI for ChatGPT training copyright infringement Copyright infringement A group of authors, including Pulitzer Prize winner Michael Chabon, file a lawsuit against OpenAI in a San Francisco federal court, accusing the company of using their writings without permission to train ChatGPT. The authors argue that OpenAI copied their works to teach ChatGPT to respond to human text prompts. This lawsuit is the third proposed copyright-infringement class action against Microsoft-backed OpenAI by authors. OpenAI and other companies claim that using copyrighted material from the internet for AI training constitutes fair use. The authors seek damages and an injunction against OpenAI's alleged "unlawful and unfair business practices."[131]
2023 September 13 Alibaba releases LLM Competition Alibaba releases its large language model Tongyi Qianwen, which is made available for public and enterprise use in China. Tongyi Qianwen, similar to ChatGPT, was previously in a beta test phase and is trained on English and Chinese text, although its exact specifications are undisclosed. This release coincides with the relaxation of AI technology restrictions in China, which now require vetting and certification for public AI tech. Companies like Baidu, Tencent, TikTok, and ByteDance have already received approval to launch AI models in China by this time. In contrast, the U.S. remains in the early stages of AI regulation discussions.[132]
2023 September 15 ChatGPT Plus vs Claude Pro Competition An article discusses the competition between ChatGPT Plus and Claude Pro, two subscription-based AI chatbot services. While both are based on large language models, they differ in their underlying models. ChatGPT Plus relies on GPT-4, known for advanced reasoning and creativity, while Claude Pro uses Claude 2, designed for summarization and ethical alignment. Both offer free versions and APIs for customization. They both cost $20/month and have similar performance, with ChatGPT being slightly more creative and Claude Pro excelling in summarizing large text volumes. The choice depends on personal preference and specific needs, with ChatGPT offering more features, including plugins, while Claude Pro emphasizes summarization and ethical AI.[133]
2023 September 16 ChatGPT healthcare potential Performance A study conducted by Emory University School of Medicine finds that ChatGPT performs well in diagnosing eye-related complaints and outperforms the symptom checker on WebMD. ChatGPT is seen as an improvement over searching symptoms online, but it raises questions about how to integrate chatbots like ChatGPT into healthcare systems. Some believe that AI-based medical tools should undergo a regulatory approval process similar to that of drugs, but there's no clear framework for this yet. Despite the potential advantages, concerns about privacy, safety, bias, liability, transparency, and regulation persist in the deployment of AI in healthcare.[134][135]
2023 September 20 OpenAI integrates DALL-E 3 into ChatGPT Update OpenAI unveils DALL-E 3, an advanced version of its image generator, which is integrated into ChatGPT. DALL-E 3 can create highly detailed images, excelling in generating images with letters, numbers, and human hands. This development enhances ChatGPT's capabilities as a hub for generative AI, producing text, images, sounds, and more. With DALL-E 3 integrated into ChatGPT, users can describe what they want to see, generating images through the chatbot, streamlining media production. However, OpenAI is cautious about the technology's potential for misuse and is implementing safeguards.[136]
2023 End of year Revenue According to OpenAI's estimates, ChatGPT is expected to generate US$200 million in revenue by the end of this year.[6]
2024 End of the year Revenue According to OpenAI's estimates, ChatGPT is expected to generate US$1 billion in revenue by the end of this year.[6]

Numerical and visual data

The chart below shows pageviews of the English Wikipedia article ChatGPT, from November 2022 to August 2023. See spike of interest towards April, and notable loss of interest afterwards.[137]

Chatgpt-wikipedia-views.PNG



The chart below shows Google Trends data for ChatGPT, from November 2022 to August 2023, when the screenshot was taken. Interest is also ranked by country and displayed on world map.[138]

Chatgpt-google-trends.png

Meta information on the timeline

How the timeline was built

The initial version of the timeline was written by Sebastian.

Funding information for this timeline is available.

Feedback and comments

Feedback for the timeline can be provided at the following places:

  • FIXME

What the timeline is still missing

Timeline update strategy

See also

External links

References

  1. 1.0 1.1 Kharpal, Arjun. "Chinese tech giant Alibaba working on a ChatGPT rival; shares jump". CNBC. Retrieved 4 March 2023. 
  2. 2.0 2.1 Bhaimiya, Sawdah. "A Colombian judge used ChatGPT to help rule on a case about the medical rights of a child with autism, reports say". Business Insider. Retrieved 4 March 2023. 
  3. DiBenedetto, Chase (18 January 2023). "ChatGPT's surprisingly human voice came with a human cost". Mashable. Retrieved 7 March 2023. 
  4. "The AI Arms Race Is On. Start Worrying". Time. 16 February 2023. Retrieved 6 March 2023. 
  5. Kelly, Samantha Murphy (16 March 2023). "5 jaw-dropping things GPT-4 can do that ChatGPT couldn't | CNN Business". CNN. Retrieved 23 March 2023. 
  6. 6.0 6.1 6.2 6.3 6.4 6.5 "ChatGPT Statistics 2023". tooltester. Retrieved 22 March 2023. 
  7. 7.0 7.1 Rogers, Paul (8 March 2023). "56 ChatGPT Statistics For 2023 (Users, Facts & Trends)". Answer iQ. Retrieved 17 June 2023. 
  8. "OpenAI adds new data controls to ChatGPT and previews business version". SiliconANGLE. 25 April 2023. Retrieved 7 May 2023. 
  9. "From the headlines: ChatGPT and other AI text-generating risks". SC Media. 4 May 2023. Retrieved 17 June 2023. 
  10. Team, Editorial (2 May 2023). "Why ChatGPT is a Cyber Threat to Businesses". insideBIGDATA. Retrieved 17 June 2023. 
  11. "ChatGPT's Growth Begins to Flatten, up 12.6% from March to April". 
  12. Saltzman, Marc; Apr. 21, readersdigest caUpdated: (21 April 2023). "What is ChatGPT, Anyway?". Reader's Digest Canada. Retrieved 7 May 2023. 
  13. 13.0 13.1 "Does ChatGPT Have Privacy Issues?". MUO. 13 April 2023. Retrieved 24 May 2023. 
  14. "ChatGPT saw its first-ever user decline in June". Engadget. Retrieved 25 July 2023. 
  15. Silberling, Amanda (25 July 2023). "Twenty years ago, AIM chatbot SmarterChild out-snarked ChatGPT". TechCrunch. Retrieved 25 July 2023. 
  16. "Introducing OpenAI". openai.com. Retrieved 15 March 2023. 
  17. "OpenAI Gym Beta". openai.com. Retrieved 15 March 2023. 
  18. "Universe". openai.com. Retrieved 15 March 2023. 
  19. 19.0 19.1 "GPT-4: another breakthrough in AI?". www.skippet.com. Retrieved 1 September 2023. 
  20. "ChatGPT Timeline: Evolution and Rise of AI, Impact, Threat, and Opportunities". techtimes.com. Retrieved 18 December 2023. 
  21. Center, Microsoft News (22 July 2019). "OpenAI forms exclusive computing partnership with Microsoft to build new Azure AI supercomputing technologies". Stories. Retrieved 18 December 2023. 
  22. "Microsoft invests in and partners with OpenAI to support us building beneficial AGI". openai.com. Retrieved 18 December 2023. 
  23. Vincent, James (22 July 2019). "Microsoft invests $1 billion in OpenAI to pursue holy grail of artificial intelligence". The Verge. Retrieved 18 December 2023. 
  24. "Microsoft invests $1 billion in OpenAI, which is going all-in on Azure". ZDNET. Retrieved 18 December 2023. 
  25. Rowley, Jason D. (22 July 2019). "Microsoft Commits $1 Billion To OpenAI Partnership To Build Its AI Systems On Azure". Crunchbase News. Retrieved 18 December 2023. 
  26. "Introducing ChatGPT". openai.com. Retrieved 16 March 2023. 
  27. Roose, Kevin (December 5, 2022). "The Brilliance and Weirdness of ChatGPT". The New York Times. Archived from the original on January 18, 2023. Retrieved December 26, 2022. Like those tools, ChatGPT — which stands for "generative pre-trained transformer" — landed with a splash. 
  28. Vincent, James (5 December 2022). "AI-generated answers temporarily banned on coding Q&A site Stack Overflow". The Verge. Retrieved 6 March 2023. 
  29. Whannel, Kate (December 27, 2022). "Could a chatbot answer Prime Minister's Questions?". BBC News. Archived from the original on January 17, 2023. Retrieved December 30, 2022. 
  30. Hartmann, Jochen; Schwenzow, Jasper; Witte, Maximilian (2023). "The political ideology of conversational AI: Converging evidence on ChatGPT's pro-environmental, left-libertarian orientation". doi:10.48550/arXiv.2301.01768. 
  31. "Elon Musk recruits team to develop OpenAI's ChatGPT rival: Reports". Al Arabiya English. 28 February 2023. Retrieved 4 March 2023. 
  32. "The 6 biggest problems with ChatGPT right now". Digital Trends. 27 January 2023. Retrieved 24 May 2023. 
  33. "Faced with criticism it's a haven for cheaters, ChatGPT adds tool to catch them". cbc.ca. Retrieved 24 May 2023. 
  34. "Introducing ChatGPT Plus". openai.com. Retrieved 16 March 2023. 
  35. A, Subaveerapandiyan; A, Vinoth; Tiwary, Neelam (1 February 2023). "Netizens, Academicians, and Information Professionals' Opinions About AI With Special Reference To ChatGPT". arXiv:2302.07136 [cs]. doi:10.48550/arXiv.2302.07136. 
  36. Hu, Krystal (2 February 2023). "ChatGPT sets record for fastest-growing user base - analyst note". Reuters. Retrieved 14 March 2023. 
  37. Pascual, M. G. (7 February 2023). "Google presents Bard, its answer to ChatGPT". EL PAÍS English Edition. Retrieved 4 March 2023. 
  38. Khalil, Mohammad; Er, Erkan (8 February 2023). "Will ChatGPT get you caught? Rethinking of Plagiarism Detection". arXiv:2302.04335 [cs]. doi:10.48550/arXiv.2302.04335. 
  39. Bang, Yejin; Cahyawijaya, Samuel; Lee, Nayeon; Dai, Wenliang; Su, Dan; Wilie, Bryan; Lovenia, Holy; Ji, Ziwei; Yu, Tiezheng; Chung, Willy; Do, Quyet V.; Xu, Yan; Fung, Pascale (28 February 2023). "A Multitask, Multilingual, Multimodal Evaluation of ChatGPT on Reasoning, Hallucination, and Interactivity". arXiv:2302.04023 [cs]. doi:10.48550/arXiv.2302.04023. 
  40. Basic, Zeljana; Banovac, Ana; Kruzic, Ivana; Jerkovic, Ivan (9 February 2023). "Better by you, better than me, chatgpt3 as writing assistance in students essays". arXiv:2302.04536 [cs]. doi:10.48550/arXiv.2302.04536. 
  41. Guo, Shuaishuai; Wang, Yanhu; Li, Shujing; Saeed, Nasir (12 February 2023). "Semantic Communications with Ordered Importance using ChatGPT". arXiv:2302.07142 [eess]. doi:10.48550/arXiv.2302.07142. 
  42. Pardos, Zachary A.; Bhandari, Shreya (14 February 2023). "Learning gain differences between ChatGPT and human tutor generated algebra hints". arXiv:2302.06871 [cs]. Retrieved 8 March 2023. 
  43. "India gets its first ChatGPT-powered AI chatbot, 'Lexi': All you need to know". livemint.com. 14 February 2023. Retrieved 4 March 2023. 
  44. team, The Marketing editorial (23 February 2023). "How Good Is ChatGPT at Translating Content?". translate plus. Retrieved 24 May 2023. 
  45. Kocaballi, A. Baki (14 February 2023). "Conversational AI-Powered Design: ChatGPT as Designer, User, and Product". arXiv:2302.07406 [cs]. doi:10.48550/arXiv.2302.07406. 
  46. Kindred, Alahna (16 February 2023). "Google issues urgent warning to anybody using AI chatbot ChatGPT". mirror. Retrieved 4 March 2023. 
  47. Yang, Xianjun; Li, Yan; Zhang, Xinlu; Chen, Haifeng; Cheng, Wei (15 February 2023). "Exploring the Limits of ChatGPT for Query or Aspect-based Text Summarization". arXiv:2302.08081 [cs]. doi:10.48550/arXiv.2302.08081. Retrieved 7 March 2023. 
  48. "How should AI systems behave, and who should decide?". openai.com. Retrieved 16 March 2023. 
  49. Megahed, Fadel M.; Chen, Ying-Ju; Ferris, Joshua A.; Knoth, Sven; Jones-Farmer, L. Allison (17 February 2023). "How Generative AI models such as ChatGPT can be (Mis)Used in SPC Practice, Education, and Research? An Exploratory Study". arXiv:2302.10916 [cs]. Retrieved 7 March 2023. 
  50. Megahed, Fadel M.; Chen, Ying-Ju; Ferris, Joshua A.; Knoth, Sven; Jones-Farmer, L. Allison (17 February 2023). "How Generative AI models such as ChatGPT can be (Mis)Used in SPC Practice, Education, and Research? An Exploratory Study". arXiv:2302.10916 [cs]. doi:10.48550/arXiv.2302.10916. 
  51. Zhong, Qihuang; Ding, Liang; Liu, Juhua; Du, Bo; Tao, Dacheng (2 March 2023). "Can ChatGPT Understand Too? A Comparative Study on ChatGPT and Fine-tuned BERT". arXiv:2302.10198 [cs]. doi:10.48550/arXiv.2302.10198. 
  52. Leiter, Christoph; Zhang, Ran; Chen, Yanran; Belouadi, Jonas; Larionov, Daniil; Fresen, Vivian; Eger, Steffen (20 February 2023). "ChatGPT: A Meta-Analysis after 2.5 Months". arXiv:2302.13795 [cs]. doi:10.48550/arXiv.2302.13795. Retrieved 6 March 2023. 
  53. Wei, Xiang; Cui, Xingyu; Cheng, Ning; Wang, Xiaobin; Zhang, Xin; Huang, Shen; Xie, Pengjun; Xu, Jinan; Chen, Yufeng; Zhang, Meishan; Jiang, Yong; Han, Wenjuan (20 February 2023). "Zero-Shot Information Extraction via Chatting with ChatGPT". arXiv:2302.10205 [cs]. doi:10.48550/arXiv.2302.10205. Retrieved 7 March 2023. 
  54. Kocoń, Jan; Cichecki, Igor; Kaszyca, Oliwier; Kochanek, Mateusz; Szydło, Dominika; Baran, Joanna; Bielaniewicz, Julita; Gruza, Marcin; Janz, Arkadiusz; Kanclerz, Kamil; Kocoń, Anna; Koptyra, Bartłomiej; Mieleszczenko-Kowszewicz, Wiktoria; Miłkowski, Piotr; Oleksy, Marcin; Piasecki, Maciej; Radliński, Łukasz; Wojtasik, Konrad; Woźniak, Stanisław; Kazienko, Przemysław (21 February 2023). "ChatGPT: Jack of all trades, master of none". arXiv:2302.10724 [cs]. doi:10.48550/arXiv.2302.10724. Retrieved 7 March 2023. 
  55. Wang, Jindong; Hu, Xixu; Hou, Wenxin; Chen, Hao; Zheng, Runkai; Wang, Yidong; Yang, Linyi; Huang, Haojun; Ye, Wei; Geng, Xiubo; Jiao, Binxin; Zhang, Yue; Xie, Xing (2 March 2023). "On the Robustness of ChatGPT: An Adversarial and Out-of-distribution Perspective". arXiv:2302.12095 [cs]. doi:10.48550/arXiv.2302.12095. Retrieved 7 March 2023. 
  56. Zuccon, Guido; Koopman, Bevan (23 February 2023). "Dr ChatGPT, tell me what I want to hear: How prompt knowledge impacts health answer correctness". arXiv:2302.13793 [cs]. doi:10.48550/arXiv.2302.13793. Retrieved 6 March 2023. 
  57. Shakarian, Paulo; Koyyalamudi, Abhinav; Ngu, Noel; Mareedu, Lakshmivihari (27 February 2023). "An Independent Evaluation of ChatGPT on Mathematical Word Problems (MWP)". arXiv:2302.13814 [cs]. doi:10.48550/arXiv.2302.13814. 
  58. Ahmad, Aakash; Waseem, Muhammad; Liang, Peng; Fehmideh, Mahdi; Aktar, Mst Shamima; Mikkonen, Tommi (26 February 2023). "Towards Human-Bot Collaborative Software Architecting with ChatGPT". arXiv:2302.14600 [cs]. doi:10.48550/arXiv.2302.14600. 
  59. Shahriar, Sakib; Hayawi, Kadhim (2023). "Let's have a chat! A Conversation with ChatGPT: Technology, Applications, and Limitations". doi:10.48550/arXiv.2302.13817. 
  60. Wang, Jiaan; Liang, Yunlong; Meng, Fandong; Li, Zhixu; Qu, Jianfeng; Zhou, Jie (27 February 2023). "Cross-Lingual Summarization via ChatGPT". arXiv:2302.14229 [cs]. Retrieved 6 March 2023. 
  61. "UAE small businesses embrace ChatGPT as tech allows productivity boost". Al Arabiya English. 15 February 2023. Retrieved 4 March 2023. 
  62. Bharade, Aditi. "ChatGPT failed miserably in Singapore's 6th-grade tests, averaging 16% for math and 21% for science. Days later, it was getting answers right.". Business Insider. Retrieved 4 March 2023. 
  63. Milmo, Dan; editor, Dan Milmo Global technology (27 February 2023). "ChatGPT allowed in International Baccalaureate essays". The Guardian. Retrieved 3 March 2023. 
  64. Weil, Elizabeth (1 March 2023). "You Are Not a Parrot". Intelligencer. Retrieved 4 March 2023. 
  65. Weil, Elizabeth (1 March 2023). "You Are Not a Parrot". Intelligencer. Retrieved 24 May 2023. 
  66. Rao, Haocong; Leung, Cyril; Miao, Chunyan (2023). "Can ChatGPT Assess Human Personalities? A General Evaluation Framework". doi:10.48550/arXiv.2303.01248. 
  67. Wiggers, Kyle (1 March 2023). "OpenAI launches an API for ChatGPT, plus dedicated capacity for enterprise customers". TechCrunch. Retrieved 4 March 2023. 
  68. "Introducing ChatGPT and Whisper APIs". openai.com. Retrieved 16 March 2023. 
  69. "OpenAI announces ChatGPT API, Snapchat, Instacart and others already using it". GSMArena.com. Retrieved 3 March 2023. 
  70. West, Colin G. (2023). "AI and the FCI: Can ChatGPT Project an Understanding of Introductory Physics?". doi:10.48550/arXiv.2303.01067. 
  71. Felten, Ed; Raj, Manav; Seamans, Robert (2023). "How will Language Modelers like ChatGPT Affect Occupations and Industries?". doi:10.48550/arXiv.2303.01157. 
  72. Nigam, Saumya; News, India TV (3 March 2023). "Microsoft unveils Kosmos-1,a new AI model to race up with ChatGPT". www.indiatvnews.com. Retrieved 4 March 2023. 
  73. Gao, Jun; Zhao, Huan; Yu, Changlong; Xu, Ruifeng (7 March 2023). "Exploring the Feasibility of ChatGPT for Event Extraction". arXiv.org. doi:10.48550/arXiv.2303.03836. 
  74. Wolfram, Stephen (2023). What Is ChatGPT Doing ... and Why Does It Work?. Wolfram Media, Incorporated. ISBN 978-1-57955-081-3. 
  75. "What Is ChatGPT Doing ... and Why Does It Work?". www.wolfram-media.com. Retrieved 13 September 2023. 
  76. Shenwai, Tanushree (12 March 2023). "Together Releases The First Open-Source ChatGPT Alternative Called OpenChatKit". MarkTechPost. Retrieved 14 March 2023. 
  77. Vincent, James (14 March 2023). "OpenAI announces GPT-4 — the next generation of its AI language model". The Verge. Retrieved 14 March 2023. 
  78. Magnússon, Pétur (15 March 2023). "Icelandic becomes ChatGPT's second language - RÚV.is". RÚV. Retrieved 7 May 2023. 
  79. Gowran, Leigh Mc (16 March 2023). "OpenAI criticised for lack of transparency around GPT-4". Silicon Republic. Retrieved 24 May 2023. 
  80. Saxena, Vishakha (17 March 2023). "Baidu Shares Jump 14% as Users Test ChatGPT-Like Ernie Bot". Asia Financial. Retrieved 24 May 2023. 
  81. "March 20 ChatGPT outage: Here's what happened". openai.com. Retrieved 24 March 2023. 
  82. "Google begins opening access to its ChatGPT competitor Bard". Yahoo Finance. Retrieved 21 March 2023. 
  83. "ChatGPT plugins". openai.com. Retrieved 24 March 2023. 
  84. "Exclusive: Linguist says ChatGPT has invalidated Chomsky's 'innate principles of language'". Tehran Times. 27 March 2023. Retrieved 24 May 2023. 
  85. "Mastercard launches Web3 user verification solution to curb bad actors". Cointelegraph. Retrieved 8 May 2023. 
  86. Roth, Emma (18 April 2023). "Elon Musk claims to be working on 'TruthGPT' — a 'maximum truth-seeking AI'". The Verge. Retrieved 9 May 2023. 
  87. Roth, Emma (25 April 2023). "OpenAI will let you turn off your chat history in ChatGPT". The Verge. Retrieved 9 May 2023. 
  88. "New ways to manage your data in ChatGPT". openai.com. Retrieved 9 May 2023. 
  89. "The ChatGPT is back in Italy: OpenAI is now going to guarantee more transparency and more rights for users - Game News 24". game-news24.com. Retrieved 7 May 2023. 
  90. "3 Ways Hackers Use ChatGPT to Cause Security Headaches". Dark Reading. 18 May 2023. Retrieved 24 May 2023. 
  91. Perez, Sarah (25 May 2023). "OpenAI's ChatGPT app tops 500K downloads in just 6 days". TechCrunch. Retrieved 30 May 2023. 
  92. "Hong Kong Baptist University begins ChatGPT trial for teachers, but concerns remain". South China Morning Post. 19 May 2023. Retrieved 24 May 2023. 
  93. "OpenAI launches standalone ChatGPT app for iOS, Android to get one too". GSMArena.com. Retrieved 22 May 2023. 
  94. Korn, Jennifer (18 May 2023). "OpenAI launches a free ChatGPT app for iOS | CNN Business". CNN. Retrieved 24 May 2023. 
  95. Malik, Aisha (24 May 2023). "Opera launches new integrated AI sidebar powered by OpenAI's ChatGPT". TechCrunch. Retrieved 24 May 2023. 
  96. Security, Help Net (24 May 2023). "6 ChatGPT risks for legal and compliance leaders". Help Net Security. Retrieved 24 May 2023. 
  97. "OpenAI's Official ChatGPT App for iPhone Expands to 40+ Countries". MacRumors. 25 May 2023. Retrieved 30 May 2023. 
  98. Singh, Jagmeet (26 May 2023). "OpenAI's ChatGPT iOS app now available in Canada, India, Brazil and 30 more countries". TechCrunch. Retrieved 30 May 2023. 
  99. Weiser, Benjamin (27 May 2023). "Here's What Happens When Your Lawyer Uses ChatGPT". The New York Times. Retrieved 30 May 2023. 
  100. "ChatGPT: US lawyer admits using AI for case research". BBC News. 27 May 2023. Retrieved 30 May 2023. 
  101. "A lawyer used ChatGPT to prepare a court filing. It went horribly awry.". www.cbsnews.com. 29 May 2023. Retrieved 30 May 2023. 
  102. "Nvidia introduces AI supercomputer to create ChatGPT successors". cointelegraph.com. Retrieved 30 May 2023. 
  103. "Nvidia introduces AI supercomputer to create ChatGPT successors By Cointelegraph". Investing.com. Retrieved 30 May 2023. 
  104. "Nvidia unveils DGX GH200 AI computing platform for ChatGPT successors and ACE for Games service for ultrarealistic NPCs". Notebookcheck. 29 May 2023. Retrieved 30 May 2023. 
  105. Melton, Monica. "There aren't actually THAT many people using ChatGPT". Business Insider. Retrieved 18 July 2023. 
  106. Quach, Katyanna. "Scientists claim over 99 percent identification of ChatGPT". www.theregister.com. Retrieved 18 July 2023. 
  107. "ChatGPT and health care: implications for interoperability and fairness". Brookings. Retrieved 18 July 2023. 
  108. Seo, Michelle Toh,Yoonjung (9 June 2023). "OpenAI CEO calls for global cooperation to regulate AI | CNN Business". CNN. Retrieved 17 July 2023. 
  109. Ph.D, Bryan Robinson. "Will ChatGPT Lead To Extinction Or Elevation Of Humanity? A Chilling Answer". Forbes. Retrieved 18 July 2023. 
  110. Edwards, Benj (9 June 2023). "Researchers discover that ChatGPT prefers repeating 25 jokes over and over". Ars Technica. Retrieved 18 July 2023. 
  111. Montti, Roger (10 June 2023). "Researchers Warn That ChatGPT Can Be Used To Spread Malicious Code". Search Engine Journal. Retrieved 18 July 2023. 
  112. "Microsoft Alarmed at Doctors Using ChatGPT to Tell Patients Bad News". Futurism. Retrieved 18 July 2023. 
  113. Author, No (13 June 2023). "Homework will 'never be the same,' says ChatGPT founder". The Japan Times. Retrieved 17 July 2023. 
  114. Valdes-Dapena, Peter (15 June 2023). "Mercedes is bringing ChatGPT into its cars | CNN Business". CNN. Retrieved 17 July 2023. 
  115. Cyr, Alex (16 June 2023). "Is AI Coming for Your Job?". Canadian Business – How to Do Business Better. Retrieved 18 July 2023. 
  116. Nelson, Felicity (16 June 2023). "Many Companies Are Banning ChatGPT. This Is Why.". ScienceAlert. Retrieved 18 July 2023. 
  117. Bohannon, Molly. "Judge Fines Two Lawyers For Using Fake Cases From ChatGPT". Forbes. Retrieved 17 July 2023. 
  118. "Booking.com adds conversational trip planning with ChatGPT". www.phocuswire.com. Retrieved 18 July 2023. 
  119. "ChatGPT down? Here are the best AI alternatives to consider". euronews. 12 July 2023. Retrieved 14 July 2023. 
  120. "FTC investigating ChatGPT over potential consumer harm". www.npr.org. Retrieved 14 July 2023. 
  121. "Meta to release open-source commercial AI model to compete with OpenAI and Google". ZDNET. Retrieved 16 July 2023. 
  122. Mehta, Ivan (20 July 2023). "OpenAI launches customized instructions for ChatGPT". TechCrunch. Retrieved 29 July 2023. 
  123. "OpenAI launches official ChatGPT app for Android". Cointelegraph. 22 July 2023. Retrieved 25 July 2023. 
  124. Dillet, Romain (25 July 2023). "ChatGPT for Android is now available in the US, India, Bangladesh and Brazil". TechCrunch. Retrieved 29 July 2023. 
  125. "AI researchers say they've found a way to jailbreak Bard and ChatGPT". Cointelegraph. 28 July 2023. Retrieved 29 July 2023. 
  126. Johnson, Chelsea (14 August 2023). "College Professors to Revert to Paper Exams Due to Cheating with ChatGPT". Fagen Wasanni Technologies. Retrieved 15 August 2023. 
  127. "ChatGPT costs $700,000 to run daily, OpenAI may go bankrupt in 2024". technext24.com. 14 August 2023. Retrieved 15 August 2023. 
  128. Field, Hayden (28 August 2023). "OpenAI launches ChatGPT Enterprise, the company's biggest announcement since ChatGPT's debut". CNBC. Retrieved 17 September 2023. 
  129. "OpenAI's ChatGPT web traffic plummets, is the hype over?". edgemiddleeast.com. Retrieved 13 September 2023. 
  130. "A boy saw 17 doctors over 3 years for chronic pain. ChatGPT found the diagnosis". TODAY.com. 12 September 2023. Retrieved 17 September 2023. 
  131. "Pulitzer winner among authors suing ChatGPT over copyright claims". euronews. 12 September 2023. Retrieved 17 September 2023. 
  132. "Alibaba launches its ChatGPT-like AI model for public use amid loosening restrictions in China". Cointelegraph. 13 September 2023. Retrieved 17 September 2023. 
  133. "Claude Pro vs. ChatGPT Plus: Which AI chatbot is better for you?". VentureBeat. 15 September 2023. Retrieved 18 September 2023. 
  134. "'Dr. Google' meets its match in Dr. ChatGPT". npr.org. Retrieved 17 September 2023. 
  135. Lyons, Riley J.; Arepalli, Sruthi R.; Fromal, Ollya; Choi, Jinho D.; Jain, Nieraj (12 June 2023). "Artificial Intelligence Chatbot Performance in Triage of Ophthalmic Conditions". doi:10.1101/2023.06.11.23291247. 
  136. Metz, Cade; Hsu, Tiffany (20 September 2023). "ChatGPT Can Now Generate Images, Too". The New York Times. Retrieved 22 September 2023. 
  137. "Wikipedia Views: ChatGPT". wikipediaviews.org. Retrieved 31 August 2023. 
  138. "Google Trends". Google Trends. Retrieved 31 August 2023.