I’ve always been interested in math and AI, mostly independently. Now that AI can do some math I’m interested in that intersection too. I plan to use this site to post my observations on the topic.

Here are some motivating questions I hope to explore.

  • What is the frontier of AI systems’ math capabilities?

  • What do different benchmarks tell us about these capabilities?

  • What additional benchmarks might help us understand these capabilities better?

  • Can humans collaborate with AI systems to do math that neither can do alone?

  • Can AI systems be useful for human math education?

About me: I was a math major in college, but haven’t pursued math as a career. I’ve worked in AI more broadly, but not on the systems that are good at math. This is a somewhat amateur endeavor! Please reach out if you’d like to collaborate or argue.

User's avatar

Subscribe to Lemmata

Observations About AI and Math