Physicists say AI agents chose and solved a PRL-worthy problem in two hours
Puzzleheaded-King584 · reddit · 2026-10-27
Christopher Herzog, physics professor at King's College London, reports that he and Andy Stergiou had ChatGPT and Claude autonomously pick a problem, do the analysis and draft a paper suitable for Physical Review Letters. The two agents each proposed three ideas, ranked the six and agreed on one — a conjecture bounding the specific-heat-to-central-charge ratio of 3d CFTs by the free scalar. The first draft took a couple of hours; the humans then spent two weeks verifying everything and rewriting the prose. The resulting paper, "The Free Scalar as a Thermal Envelope for Three-Dimensional CFTs," conjectures cS/CT ≤ 16πζ(3) and tests it in free, large-N and holographic theories plus Monte Carlo data for 3d Ising and O(N) models.
More from AGI Musings
- Why AI can't be a great idea guy: its own outputs raise the bar it must clear — theteknosaur · 2026-10-28
- ATMs didn't eliminate bank tellers: why AI may reshape jobs the same way — Olivier__OG · 2026-10-28
- Sci-fi writer interview: computing power should explore consciousness, not conquest — thesephist · 2026-10-28
- Turing's 1951 'heretical' essay predicted machines would outthink humans and take control — JeffLadish · 2026-10-28
- Developers debate which jobs survive as AI takes over software work — Top-Organization-675 · 2026-10-28
- As AI agents rise, human research taste becomes a massive premium — Swarooprm7 · 2026-10-28