Mathematics and AI refers to the two-way relationship between mathematics and artificial intelligence. The two fields have been interrelated since the initial development of AI decades ago. On one hand, mathematics plays a key role in artificial intelligence. For example, many AI algorithms use methods relying on optimization, statistics and linear algebra. On the other hand, researchers have for decades been trying to use AI and computers to help solve mathematics problems. Artificial intelligence has made a significant impact on many academic disciplines. The impact has been pronounced in mathematics, where results are exactly verifiable.
Background
Mathematics in AI Many AI systems aim to approximate a set of data. This data, represented as numbers, is combined with weights using the techniques of linear algebra. The weighted data is typically non-linearized using for instance a sigmoid function, before defining a loss function to minimize. The loss function is then minimized using an optimization algorithm, for example gradient descent. This produces a set of weights which can be used to approximate the initial data.
AI in mathematics In mathematics the goal is often to prove a statement or theorem. Large language models (LLMs) can assist mathematicians by writing proofs or helping find mistakes in written work. LLMs can also be used to disprove theorems by finding counterexamples. Since mathematics is verifiable, once a proof is written it can be verified by checking that every line of the proof follows from the assumptions, for example by using a proof assistant such as Lean. This makes mathematics very suitable as training data for AI.
History After the computer revolution, computers have made huge strides in mathematics, for example by helping to solve large numerical calculations in physics, engineering and military research. In 1956, a significant step was made when mathematicians built a program able to prove many of the theorems from Russel and Whitehead's Principia Mathematica. The Logic Theorist, as it became known, has been dubbed "the first artificial intelligence program". During the 1960s, key strides were made in the theoretical foundations of artificial intelligence. A decade later, computers began to be used in computer assisted proofs. In 2017, transformer architecture, a key component in modern LLMs, was developed using mathematics. In 2021, DeepMind's AI enabled researchers to identify patterns in mathematics which had remained elusive to humans. Google then launched AlphaGeometry in 2024, which was able to use artificial intelligence to solve advanced problems in Euclidean geometry.
Present Up until roughly 2024, large language models (LLMs) were poor at doing mathematics. Since then, LLMs from major companies have made significant strides in mathematics, heralding an "AI Revolution in Math". Some of the most significant breakthroughs are summarized in the following table.
In 2026 the First Proof [1] project was launched to benchmark the capability of current AI models to perform research level mathematics.
Discussion
Concerns have been raised that the importance of the problems solved by AI in mathematics is overstated, and that the companies publishing these results are using the proofs as advertising for their models. There is also the question of attribution when AI makes a mathematical discovery. Furthermore, the prohibitive price of the best AI models can potentially restrict mathematical research to elite institutions. To address some of these issues, several mathematicians have signed the Leiden Declaration on Artificial Intelligence and Mathematics. Mathematician Terence Tao has argued that AI will fundamentally change the way mathematics is done in the future. In particular, AI will soon generate proofs that are verifiable by proof assistants yet indigestible to humans. The role of the mathematician will therefore have to involve greater emphasis on digesting and disseminating proofs.
References
See also
Notes
