Changes

Jump to: navigation, search

Timeline of Wei Dai publications

1,020 bytes added, 18:11, 5 March 2020
no edit summary
| 2009 || {{dts|November 9}} || Blog post || LessWrong || [http://lesswrong.com/lw/1e8/reflections_on_prerationality/ "Reflections on Pre-Rationality"] || || 30 ||
|-
| 2009 || {{Dts|November 16}} || Blog post || LessWrong || [http://lesswrong.com/lw/1fu/why_and_why_not_bayesian_updating/ "Why (and why not) Bayesian Updating?"] || The post shares Paolo Ghirardato's paper "Revisiting Savage in a Conditional World", which gives axioms (based on [[wikipedia:https://en.wikipedia.org/wiki/Leonard_Jimmie_Savage|Savage]]'s axioms) which are "necessary and sufficient for an agent's preferences in a dynamic decision problem to be represented as expected utility maximization with Bayesian belief updating". The post points out that one of the axioms is violated in certain problems like counterfactual mugging and the absentminded driver; this motivates using something other than Bayesian updating, namely something like Wei Dai's own updateless decision theory. The postscript points out the fact that in Ghirardato's decision theory, using the joint probability instead of conditional probability gives an equivalent theory (i.e. it is possible to have an agent that produces the same actions, but where the agent's beliefs can only get weaker and never stronger), and asks why human beliefs don't work like this. (The comments provide some answers.) || 26 ||
|-
| 2009 || {{dts|November 24}} || Blog post || LessWrong || [http://lesswrong.com/lw/1gg/agree_retort_or_ignore_a_post_from_the_future/ "Agree, Retort, or Ignore? A Post From the Future"] || || 87 ||

Navigation menu