At Monday’s Fireside Talk, Professor Yang-Hui from the University of Oxford, took us on a fascinating journey into how artificial intelligence is reshaping mathematical discovery and what that means for the future of the subject. The talk explored three complementary approaches to doing mathematics. First, a bottom-up approach, where AI excels at formal logic, pattern-spotting, and automated proof-checking, tasks that would take humans years to complete. Second, a top-down approach, where intuition, creativity, and experience still matter most, with AI acting as a powerful guide rather than a replacement. And finally, meta-mathematics, where large language models analyse the structure and language of mathematical research itself, revealing hidden patterns in how ideas develop.
Along the way, we heard some eye-opening anecdotes. In 2022, ChatGPT passed the Turing Test, sparking debate about whether machines can truly “think”. Even more striking, recent AI systems have reached a level comparable to a 12-year-old Terence Tao—and are now capable of achieving Gold-medal-level performance in mathematical Olympiads. Yet, as Professor He stressed, AI still struggles with deep understanding and interpretation without human insight.
The key message was clear: the future of mathematics lies not in humans versus machines, but in human-AI collaboration, combining creativity, intuition, and ethical judgement with computational power on an unprecedented scale. A thought-provoking, inspiring evening, with students driving the discussion through their curiosity and asking sharp, insightful questions.
Next Monday, we welcome BBC correspondent and author Ben Chu, who will be discussing Exile Economics: What Happens If Globalisation Fails?