Ex-Tesla AI Chief Predicts Labs Won't Need New Pretraining Researchers by 2027
jasondeanlee · x · 2026-11-21
Jason Dean Lee (former Tesla Autopilot lead) claims AI labs are building a benchmark to evaluate research-agent automation and hill-climbing on it — the highest-value target, in his view. He predicts labs will stop hiring new pretraining researchers by 2027, with posttraining hiring to follow soon after. In the thread, AcerFur responds that rather than millennium-problem proofs, he'd most like to see AI crack Goldbach or the TPC. The exchange captures an aggressive industry timeline for AI automating AI research itself.
More from AGI Musings
- Vishal Sikka on reported Riemann Hypothesis proof with Claude: 'an incredible day' — vsikka · 2026-11-21
- Tamay offers 2:1 odds that RH won't be solved within two years — panickssery · 2026-11-21
- Dev Laments: AI Coding Offloaded the Fun Parts, Leaving Only Reviewing and Correcting — pritisinghhhh · 2026-11-21
- Chollet: AI made staggering leaps since 2024, yet NVDA buyers are down 5% — fchollet · 2026-11-21
- Independent mathematicians attest AI Riemann hypothesis proof is almost certainly true — soumitrashukla9 · 2026-11-21
- Pretraining vs RL debate is silly — evidence says you need both, argues voooooogel — voooooogel · 2026-11-21