Try, even if they have you cold

In a recent LessWrong post titled 'Try, even if they have you cold,' author WalterL challenges the tendency of smart people to avoid action when they foresee a guaranteed failure. The argument draws on counterintuitive examples: Stockfish, a top chess engine, can still beat human players even when giving knight odds; the Go AI Sensei regularly defeats kyu players with a 6-stone handicap. These AIs succeed not because the odds are in their favor, but because their imperfect adversaries make mistakes over time. The principle extends to real-world scenarios: panhandlers, scammers, and even e-girls succeed despite seemingly hopeless odds because some targets will still engage. WalterL recalls selling Netflix stock early because mailing DVDs seemed absurd—yet Netflix thrived, proving that 'doing a dumb thing' often beats sneering at it.

The core message is that counter-counter-move is rare in practice. Smart people enjoy theoretical if-then reasoning and assume adversarial agents will have perfect counterplay. But in reality, opponents are often unprepared, distracted, or inconsistent. The author borrows a phrase: 'just do things.' Even if initial attempts fail, persistence can reveal a low hitrate that requires multiple tries to detect, as commenter cousin_it notes. This essay is a rallying cry for professionals who overanalyze risks and underestimate the power of taking action, even when logic suggests it's futile.