OpenAI’s release of 719 research papers generated by artificial intelligence has reignited debate over the future of mathematics. But while AI may produce results at unprecedented speed, one mathematician argues that research is about more than calculations and proofs: it is a creative, personal pursuit akin to writing poetry or fiction.
On 6 October, OpenAI drew attention in specialist mathematical circles by publishing 719 research papers covering a broad range of abstract mathematics. According to the account, the results were produced in an average of three hours each using ChatGPT Pro alongside an unreleased OpenAI model. None of the papers had been verified by human researchers.
The development raises two distinct questions: whether AI will threaten the livelihoods of mathematicians and whether it will make the human pursuit of mathematics redundant. These concerns are often treated as one, but they are not the same, the author argues.
The economic impact of automation is a legitimate concern for researchers who depend on their work for an income. A universal basic income (UBI), the author suggests, could help address the problem by ensuring people can support themselves even if AI renders large numbers of jobs obsolete.
But the deeper question is whether people will still want to do mathematics if machines can do it faster.
Mathematics is more than finding the right answer
The author compares mathematical research to writing poetry or fiction rather than performing arithmetic. Calculators can reliably handle sums, but research mathematics involves more than obtaining a correct answer or proving a theorem. It also requires developing ideas, exploring relationships and constructing compelling intellectual narratives.
In this sense, mathematical concepts play a role similar to characters in a novel. Shapes, functions and abstract relationships exist within systems governed by their own rules. For a mathematical argument to be convincing, those relationships must be logically consistent, much as characters and imaginary worlds must make sense within a work of fiction.
The author’s own research involves interpreting high-dimensional mathematical structures through the lower-dimensional patterns they produce. The process is compared to the central character in Peter Høeg’s novel Miss Smilla’s Feeling for Snow, who reads footprints to reconstruct the movements that created them.
Such work, the author argues, is not primarily about calculation. It is about interpreting patterns and developing an understanding of structures whose meaning may not be immediately apparent.
The role of intuition in mathematical discovery
A crucial part of research is turning vague mathematical intuitions into precise ideas. These intuitions can be difficult to communicate because researchers may not yet fully understand them themselves.
AI may be able to generate proofs and explore problems, but the author questions whether it can reproduce this deeply personal process of discovery. A system cannot simply be given an intuition that its creator has not yet managed to articulate.
This does not mean all mathematics operates in the same way. Some research focuses on clearly defined optimisation problems, such as determining how spheres can be packed together in higher-dimensional spaces. Ukrainian mathematician Maryna Viazovska won the Fields Medal in 2022 for her groundbreaking work on sphere packing in eight dimensions.
Yet the author sees a distinction between solving a defined problem and pursuing research driven by emerging ideas and personal intellectual curiosity.
Why human mathematics will endure
The author acknowledges that AI could make enough jobs obsolete to force societies to consider measures such as universal basic income. Even in that scenario, people would continue to pursue mathematics, just as they continue to cook at home despite the availability of faster, convenient alternatives.
Home cooking can express personal experience, family traditions and cultural identity in ways that go beyond the simple production of food. Mathematical research, the author argues, can similarly reflect the interests, intuitions and creativity of the person doing it.
AI may increasingly outperform humans in speed and computational capacity. But the central argument is that efficiency does not determine the value of every human activity.
Mathematics, like poetry, cooking and love, can be a form of personal expression. Even if machines become capable of producing mathematical results faster than people, that does not necessarily mean humans will stop wanting to discover mathematics for themselves.
Be First to Comment