Delip Rao: scaling was necessary but not sufficient, OpenAI got lucky
deliprao · x · 2026-09-07
AI researcher Delip Rao argues scaling was a necessary but by no means sufficient condition for today's AI. OpenAI's faith in scaling laws was misplaced, and the field got lucky because later approaches—many invented outside OpenAI, like function calling and ReAct—worked synergistically with blind scaling.
Without those inventions, Kaplan's "scaling will solve everything" thesis would have hit a ceiling quickly, far short of current reality, Rao says.
More from AGI Musings
- LEAP survey: 194 experts put >50% odds on a major AI-driven harm event by 2050 — ben_j_todd · 2026-09-07
- AI X-Risk Advocate Fires Back at Skeptics: Warning the World Is the Logical Response — nabla_theta · 2026-09-07
- Writer's Handwritten Essay Scored 100% AI-Generated by Pangram Detector — austinc3301 · 2026-09-07
- Economist: AI has net-created over 1M US jobs, postponing the jobs apocalypse — QuintinPope5 · 2026-09-07
- Hameroff revives quantum consciousness theory: aromatic amino acids form a 'Quantum Underground' — JosephJacks_ · 2026-09-07
- Vibe coding a prototype in two nights, and nobody uses it next week: on fake demand — sujingshen · 2026-09-07