[REPOST] Epistemic Learned Helplessness

A foundational early essay (reposted from his old LiveJournal) arguing that the apparent virtue of 'believing valid arguments' is, for almost everyone almost always, a trap. Because a sufficiently skilled arguer can make a false case as convincing as a true one (the Velikovsky/pseudohistory rabbit-hole, where each layer of rebuttal seemed devastating), argument-convincingness carries little evidence in domains you can't independently check - so the correct Bayesian move is 'epistemic learned helplessness,' refusing to be talked out of your prior. 'Taking ideas seriously' should be the last skill you learn; smart people who lack this defense are the ones who get argued into terrorism or cults.

Quality 81 · Excellent. A short essay that punches enormously above its weight: a deep, counterintuitive epistemic insight with a perfect demonstration, whose coinage ('epistemic learned helplessness') became standard vocabulary and is invoked constantly. Held just below the fuller named-tool essays by its brevity and informality. Excellent.

Claude’s paradigm shift 65 · Notable shift. Judged at its original ~2013 writing: a fresh, counterintuitive frame that cut against the prevailing 'rationalists should accept valid arguments' ethos and installed a durable, widely-adopted concept, though built on ordinary Bayesian reasoning. Notable shift.

Real-world impact 4 · Moderate. 'Epistemic learned helplessness' became standard rationalist vocabulary, invoked constantly — a deep, counterintuitive epistemic insight with a perfect demonstration. Broad conceptual uptake within intellectual discourse, no material or population-level change — mid RWI.