Seven weeks ago, mathematicians said AI could execute but not decide direction. Then a Fields medalist won his medal and left academia for OpenAI the same day.
This is an expanded and updated version of “AI in math is going exponential: A working mathematician’s view”, an online talk I gave to the Madrid effective altruism and AI safety community on 15 June 2026. A Czech version, “Stroj na věty: Co umělá inteligence dělá s matematikou a proč by nás to mělo zajímat”, followed on 29 June, delivered from Madrid to mostly in-person audiences in Prague and Brno. Given how much happened since then, the post is substantially updated. I am a postdoctoral researcher at ICMAT in Madrid; in my own time I volunteer for AI Czech. These views are my own.
https://kubuondr.substack.com/p/ai-in-mathematics-2026-assistant-researcher