Parisi, Zamponi Solve 2014 Jamming Puzzle With Claude-Guided Proof
Updated
Updated · Gizmodo · Jul 1
Parisi, Zamponi Solve 2014 Jamming Puzzle With Claude-Guided Proof
3 articles · Updated · Gizmodo · Jul 1
Summary
A paper published July 1 says Giorgio Parisi and Francesco Zamponi finally proved why two parameters in their 2014 jamming model always sum to 1.
Claude supplied an initial line of attack after the physicists hit a stalemate, but its proof contained errors and needed repeated human checking and revision.
Francesco Zamponi said the AI's core idea was "essentially correct" and pointed the team to a simpler case they had missed while searching for a deeper structure.
The result adds to evidence that generative AI can help spot patterns or promising approaches in hard mathematics, even when experts must still validate the final proof.