In a groundbreaking development, Google’s DeepMind, renowned for its advancements in AI, has successfully tackled complex geometry problems traditionally reserved for the brightest mathematical minds competing in the International Mathematics Olympiad (IMO). This marks a significant milestone in the intersection of AI and mathematical problem-solving.

Geoff Smith, Chair of UKMT, said: “Among maths topics, geometry was always the lowest hanging fruit because it is a closed world in a sense, a bit like the games Go and Chess where AI systems are already stronger than any human. Even so, this is an astonishing piece of work.”

“We teach little geometry in UK secondary schools, so this will have limited implications for our education system. In countries which have rich geometry school syllabuses, this will (in time) cause problems.”

While we’re not quite at the point where AI can outshine human intellect in every aspect, DeepMind’s success is a promising step forward. It highlights the symbiotic relationship between human ingenuity and AI capabilities, paving the way for exciting new discoveries and innovations in the future.

Read more about this discovery here:

New Scientist Article
Deepmind Article

Find out more about XTX Market’s $10mn AI Mathematical Olympiad Prize.