ACX
My Bet: AI Size Solves Flubs
Scott Alexander bets that AI models will quickly overcome current limitations, based on how GPT-3 improved on GPT-2's shortcomings identified by Gary Marcus.
25 min read 3,787 words
The reader’s shortlist
Worth a second thought.
Essays ranked by captured source likes. Essays without a saved count are listed last.
Most liked essays
2,008 essaysACX
Things I Learned Writing The Lockdown Post
Scott Alexander shares insights gained from writing a comprehensive post on lockdown effectiveness, highlighting the complexities and challenges of analyzing such a multifaceted issue.
24 min read 3,645 words
LW
2011 Survey Results
Scott Alexander presents comprehensive results from the 2011 Less Wrong survey of 1090 respondents, covering demographics, beliefs, probability estimates on various topics, and a calibration test that shows the community performs better than average but has some calibration issues.
16 min read 2,383 words
LW
The Mystery of the Haunted Rationalist
Using a haunted mansion thought experiment, Scott argues that one mind can hold separate levels of belief about ghosts: what it professes, what it bets on, and what it fears.
6 min read 890 words
LW
The Zombie Preacher of Somerset
Scott recounts the case of Simon Browne, an 18th-century preacher who claimed to have lost his soul and consciousness after killing a robber in self-defense, yet continued to write and reason brilliantly, potentially representing a real-world philosophical zombie.
12 min read 1,772 words
ACX
Book Review: The Revolt Of The Public
Scott Alexander reviews 'The Revolt of the Public', a 2014 book that predicted the rise of social media-driven mass revolts against elites, finding its insights now obvious but acknowledging its prescience at the time of publication.
29 min read 4,492 words
ACX
Metis And Bodybuilders
Scott examines how scientific research has challenged traditional bodybuilding wisdom about rest periods, questioning the reliability of practical wisdom ('metis') compared to academic studies.
16 min read 2,403 words
ACX
The Unbearable Semiheaviness Of Being
Scott Alexander examines the concept of 'light water' and its potential health benefits, critically analyzing various arguments while expressing fascination with this unexpected area of scientific inquiry.
11 min read 1,668 words
LW
The limits of introspection
Scott reviews psychological research showing that people don't have introspective access to their own mental processes and instead guess at their reasons for behavior in the same way they would guess at others' motivations.
12 min read 1,741 words
ACX
Book Review Contest Rules 2026
Scott announces the 2026 book review contest with a May 20th deadline, requiring anonymized submissions via Google Form, offering prizes of at least $2,500/$1,000/$500 for the top three winners.
5 min read 651 words
ACX
Does Anaesthesia Prove Ketamine Placebo?
Scott Alexander examines a study challenging ketamine's effectiveness as an antidepressant, discussing potential confounding factors and maintaining his belief in ketamine's utility despite the study's results.
12 min read 1,855 words
LW
2012 Survey Results
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.
35 min read 5,419 words
ACX
Ask Me Anything (2/2025)
Scott announces a new subscriber-only AMA session, setting clear boundaries about which questions he will and won't answer.
1 min read 111 words
ACX
Clarification On "Fake Tradition Is Traditional"
Scott Alexander refines his thoughts on tradition formation, proposing a three-part model that emphasizes the success of practical actions and those tied to tradition.
4 min read 476 words
ACX
Deceptively Aligned Mesa-Optimizers: It's Not Funny If I Have To Explain It
Scott Alexander explains mesa-optimizers in AI alignment, their potential risks, and the challenges of creating truly aligned AI systems.
23 min read 3,483 words
ACX
Your Book Review: God Emperor Of Dune
A review of 'God Emperor of Dune' by Frank Herbert, analyzing its themes of power, AI risk, and human evolution, while drawing parallels to modern concerns about artificial intelligence.
34 min read 5,165 words
LW
Techniques for probability estimates
Scott describes six techniques for translating vague probability feelings into accurate numerical estimates, explaining the strengths and weaknesses of each method.
13 min read 1,936 words
ACX
Birth Order Effects: Nature vs. Nurture
Scott Alexander explores new data on birth order effects among his blog readers, finding social factors likely explain firstborns' overrepresentation.
19 min read 2,869 words
ACX
Contra Drum On The Fish Oil Story
Scott Alexander defends his criticism of the FDA's approval process in the infant fish oil case, arguing that systemic issues cause harmful delays even when the FDA follows its mandate.
24 min read 3,628 words
ACX
From The Mailbag
Scott Alexander answers reader questions on various topics, including his projects, views, and personal preferences in a mailbag format.
17 min read 2,532 words
ACX
Ukraine Warcasting
Scott Alexander evaluates predictions about the Russia-Ukraine war, analyzing the performance of prediction markets and individual pundits.
53 min read 8,064 words
ACX
Journalism and Legible Expertise
Scott explores why journalists struggle to write controversial articles about topics like COVID-19 due to experts' reluctance to be interviewed.
4 min read 552 words
ACX
Links For July 2023
A diverse collection of links and news items from July 2023, covering topics from historical curiosities to current technological and social developments.
24 min read 3,581 words
ACX
On Hreha On Behavioral Economics
Scott Alexander argues that despite some replication failures, the core of behavioral economics remains valid and valuable, with the field continuing to evolve and refine its understanding of human decision-making.
47 min read 7,248 words
ACX
Instead Of Pledging To Change The World, Pledge To Change Prediction Markets
Scott proposes politicians pledge to change prediction markets' forecasts instead of making direct outcome pledges, using Biden's emissions pledge as an example.
6 min read 877 words