Cultivating Instinct
summary
Argues that perceptual learning—short, repeated classification trials with immediate feedback and massive, messy datasets—lets brains compress instincts into days or hours instead of years. Gives concrete results: 1,200 chicks/hour 97% accuracy after apprenticeship, three hours → 80% correct on pilot visual navigation (versus ~50% baseline), and middle-school fraction problem scores rising 40% → 70% with retention. Shows practical payoff in code: a patch that cuts integration DB calls by ~40% and uncloaks an INT vs bigint truncation bug.
AI-written summary · about
chapters (14)
- 0:00Start
- 4:00Expertise
- 6:40How do you teach intuition?
- 10:17Differentiation
- 15:28Visual navigation
- 20:24Troubleshooting
- 22:06Profiling
- 22:40Refactoring
- 23:06Debugging
- 26:10Instinct (explained)
- 26:45Amateur's Guide
- 27:37Complex variation
- 28:30Identifying a target skill
- 29:06Well-defined activity
