AI grabs silver medal at human maths competition.

An artificial intelligence (AI) developed by Google DeepMind has achieved a silver medal score at the International Mathematical Olympiad (IMO), marking the first time an AI has reached the podium. The IMO is a prestigious competition for young mathematicians, and answering its test questions requir...

Descripción completa

Detalles Bibliográficos
Publicado en:New Scientist Vol. 263; no. 3502; p. 11
Autor principal: Wilkins, Alex
Formato: Journal Article
Publicado: New Scientist Ltd. 8/3/2024
Acceso en línea:Ver este registro en EBSCOhost
Descripción
Sumario:An artificial intelligence (AI) developed by Google DeepMind has achieved a silver medal score at the International Mathematical Olympiad (IMO), marking the first time an AI has reached the podium. The IMO is a prestigious competition for young mathematicians, and answering its test questions requires mathematical ability that AI systems typically lack. Google DeepMind has released a new AI called AlphaProof, which can solve a wider range of problems, and a new version of AlphaGeometry. The AI's performance was surprising and it appeared to find "magic keys" to answer problems in a similar way to humans. However, the AI's method of reinforcement learning requires a large set of problems written in a language it can understand and verify, which is a challenge for most IMO-like problems written in English. The AI's solutions were translated into a programming language called Lean using Google's Gemini AI, and it took up to three days to find some solutions. The hope is that AlphaProof will help improve the accuracy of Google's large language models.