PSI Leverages GPT and Claude to Tackle Quantum Challenges, Unveils 5 Research Findings at Once

Beating AI News Flash: AI physics lab Physical Superintelligence (PSI) announced that with the help of models such as GPT-6 Astra and Claude, it has made new progress on 5 long-standing unresolved problems in quantum information.

For example, researchers proved that in a specific quantum field, even if two regions are far apart, quantum entanglement still exists, but once the distance exceeds a certain range, nonlocal correlations can no longer be detected through the CHSH Bell inequality test. Another study proved that after a certain type of quantum operation is performed continuously enough times, entanglement will definitely be destroyed.

During the research process, physicists were responsible for guidance and verification, while models such as GPT and Claude participated in the derivations. The team also had AI Agents in Codex and Claude Code write Lean formal proofs, with computers checking whether the reasoning held. All 5 results publicly released papers, proof code, and verification materials.

However, these 5 results did not all completely solve the original problems. Some conclusions apply only under specific conditions, and some problems remain unresolved.