DeepMind AI agent uses LLMs and Lean to solve nine Erdős problems, including two unsolved for 56 years
The proofs cost a few hundred dollars per problem.
AI just recently crossed a threshold where it now can solve some open research math problems that professional human mathematicians couldn't... and this wasn't true a month ago.
The mathematics field will be reshaped much the same way that software engineering was back in January and cybersecurity was back in April.
This phenomenon will come for additional professions soon.
Another 9 open Erdos problems solved, this time by DeepMind team. Interesting loop of LLM - Lean agents working autonomously, and only after it's verified formally, going through human review.