OpenAI Solves 700 Problems, Math Community Alarmed

Key Takeaways

1OpenAI has solved over 700 math problems and published the results

2A mathathon at Caltech sparked a petition with over 700 signatures and the withdrawal of OpenAI

3Researchers are raising concerns about the impact on research and the need to recheck AI solutions

Why it matters : The massive influx of AI into mathematics is disrupting research and questioning the role of humans in the discipline.

Researchers criticize the mass effect of solutions published by OpenAI, a "mathathon" at Caltech triggers a petition and OpenAI's withdrawal, while Sam Altman assures that acceptance is progressing. Between Terence Tao's caution and questions about the opacity of reasoning, the use of AI in mathematics is divisive.

Letter from 700 signatories and OpenAI's withdrawal from the "mathathon"

At Caltech, a "mathathon" allowing the use of AI has sparked an outcry. A letter signed by over 700 individuals claims that the event could have destructive effects on the mathematical community. OpenAI has withdrawn, but the event is still scheduled to take place. Its organizer, 19-year-old student Caiman Moreno-Earle, says he wants to preserve a place for humans in the discipline. Anthropic and OpenAI each provided the equivalent of $1 million in computing power to support the initiative.

Research programs "annihilated," according to a NYU professor

After publicly releasing a solution related to the Navier-Stokes equations, OpenAI solved over 700 other problems and made the data available on GitHub. Tristan Buckmaster, a professor at NYU, believes that this flood of solutions has annihilated entire areas of research. The rapid dissemination of these results fuels a broader debate about the role of AI in mathematics.

Le brief IA que lisent les pros

Le brief IA que les pros lisent chaque soir

Les 7 actus IA du jour, décryptées en 5 min. Gratuit.

Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.

Choisissez votre rythme

Gratuit · Pas de spam · Désabonnement en 1 clic

Altman speaks of acceptance, Tao considers the "romantic" era of mathematics over

Sam Altman attempted this week to ease tensions, stating that mathematicians would largely move from anger to acceptance of AI. Meanwhile, Terence Tao, a professor at UCLA, argues that the "romantic era" of mathematics is over. He believes that new perspectives exist with these tools, while also noting losses. These statements come after OpenAI solved a problem considered historically difficult.

Solutions to be rechecked and suspicions about the genesis of ideas

The outputs of the models are not always readable or clear, forcing mathematicians to scrutinize the proofs. Questions remain about how the systems arrive at their solutions. Some speculate that, when researchers discuss their ideas with models, these models might exploit that material to preempt a solution. Others suggest that, similar to the transition from slide rules to calculators, the profession will need to adopt these tools, even if students describe the situation as demoralizing.