Want to dive into Scott Alexander's work and his thousands of blog posts? This fan website lets you sort and do semantic search through the whole codex. Enjoy!
| Date | ||
| Length | ||
| Likes | ||
| Comments |
Scott explains how the 'you ignore outside criticism' accusation works as an unwinnable Kafkatrap, and proposes a power-law approach to deciding how much attention to give critics at different distances from your views.
Scott presents an AI-driven "autograntor" that forecasts the value of AI safety grants against a milestone point system, shows sample runs, and reflects on how well it works and where it may mislead.
Scott Alexander analyzes EA's approach to criticism, arguing that specific, targeted critiques are more valuable than vague, paradigmatic ones for driving real change.
Scott argues that rationalists should test their objections against 'the least convenient possible world,' where convenient excuses are removed, in order to find their true reasons and avoid dodging hard questions.
Read Scott like a free app: tap then Add to Home Screen. Read Scott like a free app: tap then Add app to Home screen. Add Read Scott Alexander to your home screen to read it like an app.