A technical dialogue between Scott Alexander and Eliezer Yudkowsky exploring why AI alignment is difficult, covering the inapplicability of human moral development analogies, the nature of AI consequentialism, and skepticism about acausal trade as a safety mechanism.

Scott announces the Bay Area Winter Solstice 2022 event and calls for volunteers for both logistics tasks (AV, lighting, ushering) and creative tasks (speeches, songs).

Scott explains why superintelligent AI poses serious risks, how a 'fast takeoff' could leave insufficient time to solve safety problems, and why even simple-sounding goals could lead to catastrophic outcomes without proper alignment research.

Scott presents the results of the 2014 Less Wrong survey covering demographics, beliefs, community participation, and various psychological measures from 1503 respondents.

Scott announces the 2014 Less Wrong Census/Survey, encouraging all community members including lurkers to participate in the main survey and optional extra credit questions, with a monetary reward available.

Scott is preparing the 2014 Less Wrong Census/Survey and asks for feedback on the draft survey and suggestions for additional questions to include.

Scott uses Less Wrong survey data to show that experienced rationalists are actually less likely than newcomers to believe cryonics will work, yet more likely to sign up for it, demonstrating that rationalist training teaches people to act on beliefs rather than making them more credulous.

Scott Alexander presents comprehensive results from the 2013 Less Wrong community survey, covering demographics, beliefs, and behavioral patterns of 1636 respondents.

Scott announces the 2013 Less Wrong Census/Survey, urging all readers to participate in the main survey (10-15 minutes) plus optional extra credit questions, with clear instructions for standardized responses and a deadline of December 31st, 2013.

Scott announces the 2013 LessWrong Census/Survey and requests community input for minor changes, particularly seeking better IQ and Big Five tests and asking CFAR about their previous questions.

Scott announces a new wiki page cataloguing blogs by Less Wrong community members, asking readers to add their own blogs and help maintain the list.

Michael Vassar is visiting European cities in April-May 2013 and wants to meet with Less Wrong community members.

Scott presents comprehensive results from the 2012 Less Wrong Survey of 1195 respondents, covering demographics, beliefs, IQ measures, and a concerning calibration test showing poor performance at avoiding overconfidence bias.

Scott announces the 2012 Less Wrong Census/Survey and encourages all readers to take it, emphasizing that the optional extra credit questions shouldn't prevent anyone from completing the main survey.

Scott announces the draft of the 2012 Less Wrong Census/Survey and requests community feedback on question clarity, suggestions for additional questions, and recommendations for an internet IQ test before the official launch.

Scott Alexander introduces 'Guilt by Association' as a fallacy where people judge things negatively based on category membership rather than actual shared features, using examples like calling MLK a 'criminal' or abortion 'murder' to illustrate when this reasoning fails.

Scott proposes the "noncentral fallacy" as a candidate for the worst argument in the world: applying emotional reactions from archetypal category members to non-archetypal examples, illustrated through examples like calling Martin Luther King Jr. a "criminal."

Scott examines neurological theories explaining delusions like Capgras syndrome through Bayesian reasoning, arguing that damage to the right prefrontal cortex causes patients to ignore priors and base their beliefs solely on explanatory adequacy.

Scott presents six game theory scenarios demonstrating how strategic thinking can force rational actors into seemingly irrational outcomes, ranging from theoretical examples like evil plutocrats and hostile takeovers to real-world applications like dollar auctions and the Golden Balls game show.

Scott explains why no perfect voting system exists, using examples to show how different voting procedures lead to different outcomes, and discusses mathematical proofs demonstrating that all reasonable voting systems must allow some tactical voting.

Scott explains the economics of bargaining and auctions, showing how prices emerge from negotiations between buyers and sellers when both parties lack perfect information about supply and demand.

Scott defines signaling and explains how it works through examples ranging from college degrees to diamond necklaces, emphasizing both its necessity for conveying credible information and its potential for wasteful resource destruction.

Scott examines experimental data on how people actually behave in Prisoner's Dilemma and Ultimatum Game situations, finding that people use various heuristics like tit-for-tat and fairness considerations rather than purely rational strategies.

Scott explains how real-world solutions to Prisoner's Dilemmas work through social institutions, reputation systems, and evolved emotions like friendship and anger.

Scott explains backward reasoning in game theory through decision trees, culminating in a counterintuitive example where giving the US president line-item veto power could actually reduce the president's ability to achieve desired outcomes.