LESSWRONG
LW

3461
Wikitags

Postmortems & Retrospectives

Edited by Ruby, gustaf, Kaj_Sotala, Yoav Ravid last updated 7th Oct 2025

A Postmortem or Retrospective is a reflection on past actions with an eye to what went well, what didn't, and the cause of any failures. Retrospectives are crucial for improving one's rationality: they are the opportunity to the grade both one's direct decisions as well as the decision-procedure and epistemic algorithms that one was employing. Sharing such accounts publicly is prosocial and allows others to learn from one's experience too.

One can ask "could I have done better had I reasoned better with the information available?" Often the answer is "yes", and one can apply lessons learnt going forward. It can feel painful to reflect on one's mistakes, but doing so is how one grows. 

Updated Beliefs tagPostmortems & Retrospectives (this tag)
Changes in general beliefs about the world – what you believed before, what you believe now, and causes of the change.Focus on reporting actions and outcomes together with an evaluation of the choices/thinking patterns used.
e.g. 'Other people are wrong' vs 'I am right'e.g. Arbital Postmortem
e.g. Six Economics Misconceptionse.g. Eliezer's retrospective on The Sequences

Related pages: Updated Beliefs (examples of), Growth Stories, Progress Studies
See also: Premortem

Subscribe
Discussion
2
Subscribe
Discussion
2
Posts tagged Postmortems & Retrospectives
1What were mistakes of AI Safety field-building? How can we avoid them while we build the AI Welfare?
Güney Türker
2d
0
5Experience Report - ML4Good Bootcamp Singapore, Sep'25
NurAlam
23d
0
40What Happened After My Rat Group Backed Kamala Harris
Blake
1mo
3
95The Center for AI Policy Has Shut Down
Tristan W
1mo
2
15Reflections from Ooty retreat 2.0
Aditya, bhishma
3mo
2
81Situational Awareness: A One-Year Retrospective
Nathan Delisle
4mo
4
29Lessons from a year of university AI safety field building
yix, afterless, Parv Mahajan, Andersehen, Tuna, neverix
5mo
3
227Jaan Tallinn's 2024 Philanthropy Overview
jaan
6mo
8
58How I switched careers from software engineer to AI policy operations
Lucie Philippon
7mo
1
90How To Believe False Things
Eneasz
7mo
14
30Meetups Notes (Q1 2025)
jenn
7mo
2
26Report & retrospective on the Dovetail fellowship
Alex_Altair
8mo
3
76AI for Epistemics Hackathon
Austin Chen
8mo
12
78Evaluating “What 2026 Looks Like” So Far
Jonny Spicer
8mo
7
26MATS Spring 2024 Extension Retrospective
HenningB, Matthew Wearden, Cameron Holmes, Ryan Kidd
9mo
1
Load More (15/204)
Add Posts