AI Tools Revolutionizing Math: Solving Erdős Problems
Artificial intelligence (AI) chatbots are making waves in the mathematical community by solving long-standing problems that have stumped even seasoned professionals. This unexpected development is transforming the way mathematicians approach problem-solving, marking a significant milestone in the field of mathematics.
The problems being tackled by AI originate from the renowned Hungarian mathematician Paul Erdős, known for posing deceptively simple yet challenging questions throughout his illustrious career. According to Thomas Bloom, a mathematician at the University of Manchester, Erdős’ questions, though seemingly straightforward, were notoriously difficult to crack.
By the time of Erdős’ passing in 1996, over 1000 of these problems remained unsolved, spanning various mathematical disciplines such as combinatorics and number theory. Today, these Erdős problems serve as benchmarks for advancements in mathematics, with Bloom overseeing a website dedicated to cataloging these problems and tracking progress made by mathematicians.
The simplicity of Erdős problems spurred mathematicians to experiment with AI tools like ChatGPT to aid in solving them. Bloom recalls that in October last year, individuals began using AI models to sift through mathematical literature for references relevant to solving these problems. Surprisingly, AI tools started uncovering partial improvements and novel results, a feat previously unattainable.
One such success story involves Kevin Barreto, a mathematics student at Cambridge University, and Liam Price, an amateur mathematician, who leveraged AI to tackle Erdős problem number 728 in number theory. Utilizing ChatGPT-5.2 Pro, they obtained a sophisticated argument for the problem, demonstrating the AI’s ability to provide valuable insights.
Subsequently, Barreto and Price employed Aristotle, an AI tool developed by Harmonic, to convert their proof into Lean, a mathematical programming language, allowing for instant computerized verification. This method streamlines the validation process, saving researchers time and effort.
As of mid-January, AI tools have fully solved six Erdős problems, though scrutiny revealed that five had previous solutions in mathematical literature. Only problem number 205 was entirely resolved by Barreto and Price without prior solutions. Additionally, AI has facilitated incremental progress on seven other problems, indicating its potential to unearth new ideas.
Despite debates on whether AI is rediscovering old solutions or generating novel insights, Bloom emphasizes the AI’s ability to uncover obscure papers and translate problems into new forms. This innovative approach has the potential to accelerate mathematical discoveries that may have remained undiscovered for years.
While current AI models excel at solving simpler Erdős problems, more complex challenges still elude them. Barreto acknowledges the need for more advanced models to tackle these tougher problems, suggesting that the AI community has room for growth before claiming bounty prizes.
Kevin Buzzard, a mathematician at Imperial College London, acknowledges the promise AI holds in solving Erdős problems but notes that the current focus on straightforward or overlooked problems may not yet pose a threat to professional mathematicians. While AI’s contributions are significant, mathematicians remain cautiously optimistic about the future implications of this groundbreaking approach. AI is revolutionizing the field of mathematics by providing researchers with the tools to explore complex problems in new and innovative ways. According to mathematician Bloom, the ability of AI models to handle intricate mathematics could change the way proofs are researched and written. This technology allows mathematicians to draw on a wider range of tools and knowledge from different disciplines, opening up new possibilities for research.
Terence Tao, a mathematician at the University of California, Los Angeles, believes that AI tools can help address a common problem in mathematics – the focus on a few difficult problems while neglecting less challenging yet important ones. By applying AI to a broader range of problems, mathematicians can adopt a more empirical and scientific approach to their work. This could involve testing different solutions on a large scale and conducting statistical studies to determine the most effective methods.
Tao emphasizes that the limited resources of expert attention in mathematics prevent researchers from exploring a vast number of potential problems. However, AI is demonstrating that large-scale mathematics is feasible and can lead to new insights and discoveries. By leveraging AI technology, mathematicians can survey a wide range of problems, identify interesting ones, and compare different methods to determine the most efficient approach.
In conclusion, the integration of AI into the field of mathematics is opening up new possibilities for research and problem-solving. By harnessing the power of AI tools, mathematicians can explore a broader range of problems, conduct large-scale studies, and revolutionize the way mathematics is practiced. This advancement in technology is not only enhancing the capabilities of researchers but also expanding the scope of mathematical research.
