Google DeepMind says it has achieved a breakthrough in building an AI system that can handle complex mathematical problems.
The research division, which is part of Alphabet-owned Google, announced Thursday that it has created a software system that combines multiple AI models and can score well-enough on the International Mathematical Olympiad (IMO), a global test of high school students’ mathematical talents, to be in the top quartile of contestants taking the test. This would be good enough to obtain a silver medal in the competition.
