PSI is trying to make a business out of AI-driven scientific discovery

Physical Superintelligence (PSI), a startup lab, used early access to Anthropic's Fable 5.2 model to solve the Aaronson-Ambainis conjecture, an 18-year-old open problem in quantum computing. David Turturean, an MIT graduate and PSI staff member, led the effort; the proof was autoformalized in Lean and validated by experts. The solution is a negative result that rules out one avenue for quantum speedup, reflecting AI's growing impact on science. PSI CEO Matt Pines explained that the company focuses on specialized model harnesses, tools, memory systems, and simulation infrastructure to tackle problems in breadth and depth. Currently, PSI generates revenue by optimizing data centers, but its long-term goal is to convert theoretical physics breakthroughs into technologies. Pines noted that physics is harder than math because "nature gets a vote," requiring experimental validation. PSI shared its result with researchers before publication, and advisor Jim Halverson emphasized transparency and expert consultation as best practices in this emerging field.

(Source：TechCrunch)