Mathematics in the age of AI
AI is tackling unsolvable math problems, changing the nature of discovery.
For over 150 years, the Riemann hypothesis has been a brick wall for the brightest minds in mathematics. It is a problem that defies conventional logic and exhausted decades of human effort. Now, a new contender has stepped into the ring.
An unreleased model from Anthropic has made progress on this legendary problem. It didn't solve it. It just pushed further than we thought possible.
This shift in mathematics is fascinating. We are moving from a world where math is purely a product of human intuition and brute-force paper-and-pencil work to one where we partner with machines to expand our reach.
The interesting part here is not the AI acting as a calculator. It’s the AI as an explorer. It can identify patterns, test variations, and suggest avenues that a human researcher might have discarded or never considered. It operates in the space of possibility, while the human mathematician provides the grounding, the intent, and the ultimate validation.
This model is a reminder of what the best AI looks like. It is an agentic partner that amplifies human curiosity. It takes the part of the process—in this case, the heavy lifting of exploration—leaving the human in total control of the direction and the judgment.
True discovery is rarely a solo act. It is a dialogue. We ask the right questions, and the machine helps us see what we couldn't before. If AI can tackle the oldest problems in human history, imagine what it does for the problems we are trying to solve right now.
Sources:
An unreleased Anthropic model made progress on one of math’s biggest unsolved problems | https://techcrunch.com/2026/08/11/an-unreleased-anthropic-model-made-progress-on-one-of-maths-biggest-unsolved-problems/
We build the systems that run the repetitive work — around the clock, gated by you.
Book a call →