Claude solves two open black-hole physics problems: a 20-year-old phase-transition calculation and an exact modular-bootstrap ceiling
Synopsis
Anthropic reports that Claude solved two long-open problems in black-hole and quantum-gravity theory. The first is a calculation physicists set up in 2003 but never completed: at weak coupling in N=4 super Yang-Mills theory, the transition that occurs as temperature rises is abrupt, matching what Hawking and Page found from a sum over gravitational histories. The second is the optimal ceiling of the simplest modular bootstrap, bounding the lightest black hole's mass at c/Λ with Λ about 9.1266, and proving that ceiling cannot be lowered. Neither result describes our own universe directly, and neither has practical applications.
Interpretation
In N=4 super Yang-Mills theory, the transition that occurs as temperature rises is abrupt rather than smooth. In 2003 Aharony and colleagues showed that whether this transition is abrupt or gradual hinges on a single number appearing only at three loops, but that calculation involves many thousands of diagrams and was never completed; in 2005 the same group finished only a simpler cousin of the theory. Claude ran a small research group of copies of itself, with almost every piece of the calculation done twice by agents acting independently and parts compared against known results in the literature, reaching an abrupt transition consistent with the Hawking-Page gravity-side calculation.
The simplest version of the modular bootstrap can push the lightest black hole's mass no lower than c/Λ, with Λ about 9.1266, and this ceiling is optimal. Numerical calculations had hinted the ceiling could reach about c/9.1, but they get harder as c grows and could not settle the question; analytically, Hartman, Mazáč and Rastelli proved a ceiling of c/8.5, clearly above what the numerics suggested. While reproducing literature numerics, Claude noticed how the ceiling shifted slightly as c grew, used a mathematical tool called a spectral curve to derive an exact formula, and proved this version cannot go lower; half the proof follows the strategy of a sphere-packing result proved by an OpenAI model, and the results were formally verified in Lean in September.
Perspective
Both results target simplified model universes: the weak-coupling end of N=4 super Yang-Mills theory in anti-de Sitter spacetime, and the simplest version of the 3D-gravity modular bootstrap that ignores black-hole spin. They do not describe our own universe directly and have no practical applications.
The abrupt transition holds at both weak and strong coupling, and the most natural guess is that it stays abrupt in between, but that is only a guess. Above the Hawking-Page temperature the hot gas can persist in a superheated state until a higher Hagedorn transition, and whether that transition's gravity-side properties also appear in N=4 super Yang-Mills theory remains to be checked. This version of the modular bootstrap cannot on its own push the ceiling down to c/12, where black holes first appear; that will take other ingredients.
