We've updated our Privacy Policy to make it clearer how we use your personal data. We use cookies to provide you with a better experience. You can read our Cookie Policy here.


DeepMind AI Used To Develop New Math Techniques

DeepMind AI Used To Develop New Math Techniques content piece image
Listen with
Register for free to listen to this article
Thank you. Listen to this article using the player above.

Want to listen to this article for FREE?

Complete the form below to unlock access to ALL audio articles.

Read time: 2 minutes
Sydney researcher Professor Geordie Williamson is working with colleagues at Oxford using DeepMind's artificial intelligence to develop fundamentally new techniques in mathematics.

Professor Williamson is a globally recognised leader in representation theory, the branch of mathematics that explores higher dimensional space using linear algebra.

In 2018 he was elected the youngest living Fellow of the Royal Society in London, the world’s oldest and arguably most prestigious scientific association.

“Working to prove or disprove longstanding conjectures in my field involves the consideration of, at times, infinite space and hugely complex sets of equations across multiple dimensions,” Professor Williamson said.

While computers have long been used to generate data for experimental mathematics, the task of identifying interesting patterns has relied mainly on the intuition of the mathematicians themselves.

That has now changed.

Professor Williamson used DeepMind’s AI to bring him close to proving an old conjecture about Kazhdan-Lusztig polynomials, which has been unsolved for 40 years. The conjectures concern deep symmetry in higher dimensional algebra.

Co-authors Professor Marc Lackeby and Professor András Juhász from the University of Oxford have taken the process a step further. They discovered a surprising connection between algebraic and geometric invariants of knots, establishing a completely new theorem in mathematics.

In knot theory, invariants are used to address the problem of distinguishing knots from each other. They also help mathematicians understand properties of knots and how this relates to other branches of mathematics.

While of profound interest in its own right, knot theory also has myriad applications in the physical sciences, from understanding DNA strands, fluid dynamics and the interplay of forces in the Sun’s corona.

Human intuition meets machine


Professor Juhász said: “Pure mathematicians work by formulating conjectures and proving these, resulting in theorems. But where do the conjectures come from?

“We have demonstrated that, when guided by mathematical intuition, machine learning provides a powerful framework that can uncover interesting and provable conjectures in areas where a large amount of data is available, or where the objects are too large to study with classical methods."

Professor Lackeby said: “It has been fascinating to use machine learning to discover new and unexpected connections between different areas of mathematics. I believe that the work that we have done in Oxford and in Sydney in collaboration with DeepMind demonstrates that machine learning can be a genuinely useful tool in mathematical research.”

Lead author from DeepMind, Dr Alex Davies, said: “We think AI techniques are already sufficiently advanced to have an impact in accelerating scientific progress across many different disciplines. Pure maths is one example and we hope that this Nature paper can inspire other researchers to consider the potential for AI as a useful tool in the field."

Professor Williamson said: “AI is an extraordinary tool. This work is one of the first times it has demonstrated its usefulness for pure mathematicians, like me.

“Intuition can take us a long way, but AI can help us find connections the human mind might not always easily spot.”

The authors hope that this work can serve as a model for deepening collaboration between fields of mathematics and artificial intelligence to achieve surprising results, leveraging the respective strengths of mathematics and machine learning.

“For me these findings remind us that intelligence is not a single variable, like an IQ number. Intelligence is best thought of as a multi-dimensional space with multiple axes: academic intelligence, emotional intelligence, social intelligence,” Professor Williamson said.

“My hope is that AI can provide another axis of intelligence for us to work with, and that this new axis will deepen our understanding of the mathematical world.”

Funding: This research was funded by DeepMind, which is owned by Alphabet Inc, the parent company of Google.


Davies A, Veličković P, Buesing L, et al. Advancing mathematics by guiding human intuition with AI.
Nature. 2021;600(7887):70-74. doi:10.1038/s41586-021-04086-x

This article has been republished from the following materials. Note: material may have been edited for length and content. For further information, please contact the cited source.