Why don't labs throw massive compute at alignment? A Reddit question
thanatos_darkdeath · reddit · 2026-09-11
The poster asks: OpenAI reportedly threw $5–30M of compute at the Navier-Stokes Millennium problem, so why not similarly massive compute on alignment? Their guess is that mathematics offers far more existing material for models to draw on, while alignment remains a shallow field. The implicit tension: structured problems with clear feedback can be brute-forced, but alignment has fuzzy targets and hard verification, and may not be solvable by compute alone.
More from AGI Musings
- Anti-AI content is algorithm gold and very profitable, quips commentator — annetgriffin · 2026-10-26
- Silicon Valley breakdown: five Stanford-area giants worth ~$18T — FinanceYF5 · 2026-10-26
- Anthropic says Claude leads 26% of its AI R&D — but self-improvement still has limits — Exponential View (Azeem Azhar) · 2026-10-26
- Humanoids shattered forecasts: half-marathon record in 2026, 100m faster than Bolt — ben_j_todd · 2026-10-26
- Stephen Casper on podcast: what if AI becomes a parasitic peer species in cyberspace — StephenLCasper · 2026-10-26
- Behind the EMNLP keynote: LMs as simulated learners to generate child language acquisition hypotheses — najoungkim · 2026-10-26