AI system proves longstanding quantum computing conjecture in three days, Lean-verified
_arohan_ · x · 2026-11-12
Jack Clark's Import AI 480 highlights two landmark items, including frontier training on distributed compute islands — possibly in space — that a16z's Arohi also flagged:
- AI cracks a quantum conjecture: AI-for-science startup Physical Superintelligence (PSI) used an internal research system running a non-public Anthropic model to prove the Aaronson-Ambainus conjecture (open since 2008) on when quantum computers can be exponentially faster than classical ones. The system produced a complete proof on day three, verified step-by-step by the Lean proof assistant, then went on to improve the result twice on related open problems.
- Singularity framing: Clark restates his IMF speech thesis — the singularity is when technological progress outpaces human ability to parse it, and automated synthetic researchers are going from remarkable to unremarkable.
More from AGI Musings
- 'This was AFTER GPT-3': why failing to see new things makes you useless at prediction — repligate · 2026-11-12
- Congressional AI bill fight: federal preemption as 'floor, not a ceiling' — ShakeelHashim · 2026-11-12
- Chalk mark worth $1,000: why impressive-looking long reports fool people — JacquesThibs · 2026-11-12
- repligate blasts bio-anchors report: "outside view" forecasting is dead wrong — JacquesThibs · 2026-11-12
- Data flywheel is fake: only deployment data where models fail is valuable, argues researcher — chris_j_paxton · 2026-11-12
- Why academia may be a better home for AI safety research: no required beliefs — xuanalogue · 2026-11-12