Final week, I attended a chat at MIT by Neil Thompson, who research progress in computing and synthetic intelligence. Over the course of simply half an hour, Thompson made a convincing case that a number of scientific fields are about to be upended by AI.
As mathematicians reel from the way in which AI is automating their area, Thompson’s newest examine means that chance, statistics idea, financial idea, high-energy idea, gravitation and cosmology, and quantum physics could also be subsequent, due to the quickly evolving capabilities of frontier fashions. Between mid-2024 and mid-2026, the examine exhibits, one of the best AI fashions went from with the ability to accurately clear up about 2 % of theoretical issues in these fields to about 35 % of such issues. And there aren’t any indicators that this progress will decelerate.
“Inside just some years, we count on a variety of idea inquiries to be completely reworked when it comes to AI’s means to do them,” says Thompson.
Broadly, Thompson’s lab digs into how advances in pc science influence society and have an effect on innovation. This examine—which Thompson performed alongside Gabriel Manso, a PhD pupil, and Claudio Mayrink Verdun, a analysis scientist—took issues acknowledged in analysis papers posted on-line between December 2025 and Might 2026, after which examined whether or not totally different OpenAI fashions—o1, o3, GPT-5, GPT-5.1, GPT-5.4, and GPT-5.5—might clear up them.
Crucially, options to those issues had been present in papers printed after the coaching cutoff for GPT-5.5 and previous fashions. Because of this the fashions they examined needed to work out options for themselves, reasonably than counting on solutions fed in as coaching information. (Such papers served as floor reality for testing the fashions’ fixing abilities.) The researchers additionally blocked web entry, which prevented the fashions they examined from trying up options on-line.
The areas lined by the examine had been restricted—the trio centered on fields wherein AI fashions might determine issues out while not having to run simulations or experiments. However the work means that the sorts of shock waves presently being felt in math are prone to ripple out to different math- and theory-heavy areas of scientific inquiry.
In September, OpenAI rocked the maths world by asserting that an unreleased mannequin had solved a long-standing, grand problem in arithmetic referred to as Navier-Stokes, which considerations equations that designate the conduct of fluid movement. The episode proved further controversial after a mathematician at NYU—who had been working with one other educational related to Anthropic—accused OpenAI of gazumping them after studying of their progress.
OpenAI and others have since introduced a whole lot extra options to longstanding math issues, triggering one thing of a disaster within the area. Some mathematicians have referred to as for limits on the usage of AI, warning that the method of fixing issues is usually as helpful as having the ultimate reply.
