Notion design engineer Geoffrey Litt: understanding is the new bottleneck in the AI coding era
EchoShao8899 · x · 2026-09-19
In EP7 of the Stanford-organized "Software With Taste" podcast, Notion design engineer and malleable-software researcher Geoffrey Litt argues that once code is free to write, human understanding becomes the bottleneck: coding agents make programming lonelier, and the process should be opened up to be inherently collaborative. He describes using AI to quiz himself during code reviews, warns against optimizing for speed at the expense of understanding, and discusses what infrastructure changes are needed once writing code is democratized.
Related event: Notion engineer: understanding is the new bottleneck in AI era(3 posts)→
More from AGI Musings
- Schmidhuber's AI Blog Surveys Recursive Self-Improvement from 1987 and the 1991 Roots of Today's AI Boom — SchmidhuberAI · 2026-09-20
- iamtrask: Concentrating more power than the AI risk you're mitigating is a bad strategy — iamtrask · 2026-09-20
- Greg Brockman: AI tips cybersecurity's balance in favor of defenders — pmddomingos · 2026-09-20
- Ruxandra Teslo: AI's most neglected safety risk is humans surrendering thinking — clarejtbirch · 2026-09-20
- Dario and Sam said we must slow down last week — so do they believe it? — _TheWolfOfWalmart_ · 2026-09-20
- How AI-Era Intermediaries Could Become Systemically Dangerous Power Brokers — iamtrask · 2026-09-20