Rationally Speaking Podcast How to reason about COVID, and other hard things (Kelsey Piper)
9 snips
Sep 14, 2021 AI Snips
Chapters
Transcript
Episode notes
Do Focused Self-Study And Check Back
- Try to learn key questions yourself when stakes are high and evidence is limited by investing 10–20 hours of focused research.
- Check later evidence and admit mistakes to calibrate whether you should keep doing independent evaluation.
Think For Yourself — But Monitor Your Error Rate
- Do your own thinking but test whether you can admit being wrong by checking future evidence.
- Stop doing independent analysis if you notice you cannot concede errors when new data arrives.
Judge Whether More Research Is Worthwhile
- Your first hour of research often misses dominant factors; plan to explore whether a question's plausible range could be decisive.
- Prioritize questions where the 90% confidence interval would change behavior significantly.