Lately, AI has been claiming to solve one open problem in mathematics after another. Even in computational complexity, it claims to have solved long-standing open problems like L = BPL.

I asked an AI about something I’d been thinking about for several years, and within minutes it pointed out a mistake. (sob)

For now, I’ve signed up for Claude Max, and I’m going to boost my math with AI. How many more months will humans still be able to contribute to mathematics? If I can’t do math anymore, maybe I’ll do philosophy.

I’m also at a loss about what assignments to give students, for bachelor’s theses and so on. AI can do a programming assignment in an instant, but work that needs conceptual understanding or ideas is out of reach for fourth-year students, who don’t have the study or the experience behind them yet.