Jeff Dean's '5+2' Framework: How to Choose Research Problems Worth Five Years
FinanceYF5 · x · 2026-08-09
At Stanford's AASF 2026 event, Jeff Dean shared his methodology for selecting research directions:
- Broad Scanning: Read abstracts of 100 papers first instead of diving deep into one. The goal is to connect important ideas that haven't been linked yet.
- The 5+2 Framework: When facing a hard problem, if 5 sub-problems have emerging community progress and 2 require inventing entirely new tech, the risk level is just right to invest around five years.
More from AGI Musings
- AI Race Reversals: Google Bounced Back with Gemini After Bard's Flop — haider1 · 2026-08-09
- The Superintelligence Paradox: Will the Public Ever Be Allowed to Use It? — imjustnewatai · 2026-08-09
- Sam Altman Declares 'We Are Now in the Singularity' — rohanpaul_ai · 2026-08-09
- Fungus Metaphor: Rethinking AI Agent Environments and Tool Use — cephaloform · 2026-08-09
- Beff Jezos: Believing AI Will Be Leashed Post-2030 After OpenAI Black Hat Talk is Delusional — beffjezos · 2026-08-09
- AI and the Tragedy of the Cognitive Commons: Disrupting the Regeneration of Professional Expertise — dr_alphalyrae · 2026-08-09