Eliezer Yudkowsky
14 notes in the collection. Most of these are from lesswrong.com (opens in a new tab)↗
First saved Leave a Line of Retreat on July 31, 2026.
Yudkowsky argues that an explanation able to fit any outcome predicts nothing, so saying "because of heat conduction" without real expectations is just magic dressed in scientific words.
Related
- Think Like RealityEliezer Yudkowsky
- Anchoring and AdjustmentEliezer Yudkowsky
- Hindsight biasEliezer Yudkowsky
- Hug the QueryEliezer Yudkowsky
Yudkowsky argues that calling quantum physics weird exposes bad intuitions, since surprise lives in the model, so you should reshape them until reality feels normal.
Related
- Fake ExplanationsEliezer Yudkowsky
- Anchoring and AdjustmentEliezer Yudkowsky
- Hindsight biasEliezer Yudkowsky
- Hug the QueryEliezer Yudkowsky
Yudkowsky argues school trains students to recite answers like "light is waves" that earn gold stars but predict nothing, and that real understanding means a belief that controls what you expect to see.
Related
- Paying AttentionMorgan Housel
- Reality has a surprising amount of detailJohn Salvatier
- Fake ExplanationsEliezer Yudkowsky
- Think Like RealityEliezer Yudkowsky
- NewSotW: Be Specific 15 min
Starting from a vague founder pitch at YC office hours, a call for exercises that train people to replace abstract claims with concrete, specific examples.
Related
- Taboo Your WordsEliezer Yudkowsky
- Illusion of Transparency: Why No One Understands YouEliezer Yudkowsky
- Fake ExplanationsEliezer Yudkowsky
- Think Like RealityEliezer Yudkowsky
- NewTaboo Your Words 4 min
Yudkowsky argues that banning a disputed word and describing what you actually expect to observe dissolves fake disagreements, like the tree falling in an empty forest, and exposes real ones hiding behind shared labels.
Related
- SotW: Be SpecificEliezer Yudkowsky
- Illusion of Transparency: Why No One Understands YouEliezer Yudkowsky
- Fake ExplanationsEliezer Yudkowsky
- Think Like RealityEliezer Yudkowsky
Explains anchoring: people start from an arbitrary number and adjust too little, as Tversky and Kahneman's wheel-of-fortune experiment showed.
Related
- Fake ExplanationsEliezer Yudkowsky
- Think Like RealityEliezer Yudkowsky
- Hindsight biasEliezer Yudkowsky
- Hug the QueryEliezer Yudkowsky
- NewHindsight bias 3 min
Yudkowsky argues that knowing an outcome makes it look far more predictable than it was, so after disasters like Challenger and 9/11 we learn overly specific lessons and underestimate the real cost of caution.
Related
- Fake ExplanationsEliezer Yudkowsky
- Think Like RealityEliezer Yudkowsky
- Anchoring and AdjustmentEliezer Yudkowsky
- Hug the QueryEliezer Yudkowsky
- NewHug the Query 2 min
Yudkowsky argues the strongest evidence sits closest to the question itself, so checking the Wright brothers' calculations or watching the plane fly beats weighing Lord Kelvin's authority or anyone's biases.
Related
- Fake ExplanationsEliezer Yudkowsky
- Think Like RealityEliezer Yudkowsky
- Anchoring and AdjustmentEliezer Yudkowsky
- Hindsight biasEliezer Yudkowsky
Yudkowsky uses Boaz Keysar's experiments to show that we know what we mean and assume others do too, so speakers overestimate how well they're understood, as Chamberlain did in his letter to Hitler.
Related
- SotW: Be SpecificEliezer Yudkowsky
- Taboo Your WordsEliezer Yudkowsky
- You can't tell people anythingChip Morningstar
- Fake ExplanationsEliezer Yudkowsky
Uses Wason's 2-4-6 task to show how people only test cases that confirm their hypothesis, and why you should look for evidence that could prove you wrong.
Related
- Fake ExplanationsEliezer Yudkowsky
- Think Like RealityEliezer Yudkowsky
- Anchoring and AdjustmentEliezer Yudkowsky
- Hindsight biasEliezer Yudkowsky
- NewTrying to Try 5 min
Why Yoda was right: setting out to "try" means aiming at having made an effort rather than at success, and that lower target quietly guarantees weaker plans.
Related
- How to be more agenticCate Hall
- Strong Opinions, Weakly HeldAbraham Thomas
- Risk-takers decide fasterSebastian Bensusan
- On the importance of infinitesimal actionsJakob Greenfeld
Yudkowsky shares his trick for getting out of bed by counting down from ten, which turns an effortful decision into a default plan that runs on its own, while warning it's shallow advice without deep theory behind it.
Related
- ProductivitySam Altman
- Trying to TryEliezer Yudkowsky
- Actually doing thingsBen Kuhn
- How to be more agenticCate Hall
Yudkowsky argues you should admit a big mistake all at once rather than concede it in small grudging pieces, since only acknowledging a fundamental error drives real change, as Enron's executives never did.
Related
- Fake ExplanationsEliezer Yudkowsky
- Think Like RealityEliezer Yudkowsky
- Anchoring and AdjustmentEliezer Yudkowsky
- Hindsight biasEliezer Yudkowsky
Before questioning a belief you care about, picture calmly how you'd live if it were false, so the evidence doesn't feel like a threat you have to fight off.
Related
- Fake ExplanationsEliezer Yudkowsky
- Think Like RealityEliezer Yudkowsky
- Anchoring and AdjustmentEliezer Yudkowsky
- Hindsight biasEliezer Yudkowsky
If you like Eliezer Yudkowsky, try
- LessWrong2 notes · also on thinking
- Sebastian Bensusan7 notes · also on thinking and writing
- Morgan Housel48 notes · also on thinking and writing