2 Comments

'OpenAI announced their latest reasoning model, o3, demonstrating another step function advance in AI capabilities. The model sets new performance records across multiple benchmarks, achieving 96.7% on the American Invitational Mathematics Exam and 87.7% on graduate-level science questions.'

If a powerful AI algorithm with instant access to almost unlimited data via the internet is only able to score 87.7% in a science test, and is unable to achieve 100% in maths which is a 'black and white' subject, where is the issue?

Is the AI model imperfect?

Or are the exams imperfect?

I don't know the answer, but its what I'm thinking about having read this post.

Expand full comment

Good stuff! Thank you for sharing! ♥️☀️☮️🌈🏁

Expand full comment