Recent MIT grad solves 18-year quantum mystery using AI

Tanuj Joshi
Tanuj Joshi - Editorial Contributor
2 Min Read
Listen to article3 min
0%

CAMBRIDGE – David Turturean stared at the freshly printed mathematical proof on his desk and felt the ground shift beneath his feet. A recent graduate of the Massachusetts Institute of Technology, he had just used an unreleased artificial intelligence model to crack a puzzle that had baffled quantum physicists for nearly two decades.

Working at the startup laboratory Physical Superintelligence, Turturean leveraged early access to Anthropic’s Fable 5.2 software. Within days, the system helped him solve the Aaronson-Ambainis conjecture, an open problem concerning how quantum computers process information faster than classical machines. The solution closes a specific theoretical avenue for high-speed computing, providing a clear boundary for researchers.

“I’m totally accepting the idea that this might be the last paper I ever put out,” Turturean said, reflecting on how quickly artificial intelligence is transforming academic publishing. “In a few months, like, why would you even put out a paper?”

Verifying Machine-Generated Breakthroughs

Scott Aaronson, co-author of the original 2008 conjecture, admitted mixed feelings of awe and fatigue upon learning of the solution. Because the proof was autoformalized in Lean, a specialized coding language for mathematics, independent experts could verify its accuracy even when the underlying mechanics challenged human intuition.

Physical Superintelligence relies on specialized harnesses and memory systems rather than general-purpose chatbots. Matt Pines, chief executive of the startup, explained that engineering effective research systems requires combining simulation infrastructure with deep domain expertise.

Translating Theory Into Real-World Infrastructure

While theoretical physics remains an ultimate goal, the startup currently generates revenue by applying its computational models to data center design. Simulating heat, power, water, and acoustics allows engineers to optimize physical performance before breaking ground on new facilities.

Academic advisors are now scrambling to adapt to the accelerating pace of machine-driven discovery. Jim Halverson, a Northeastern University professor advising the startup, noted that professional standards and peer-review processes are struggling to keep up with algorithms that generate complex proofs overnight.

Share This Article
Leave a Comment

Leave a Reply

Your email address will not be published. Required fields are marked *