1/ What happens when the world runs on code no one understands?
In TIME, Founding Mathematician @KenOno691 joins @patrickshafto at @DARPA and @skominers at @HarvardHBS to argue that mathematical verification must become a national mission in the age of AI.
See op-ed in thread.
At the National Academies’ Mathematical Sciences Education Board meeting, I moderated a session on preparing for a changing workplace. I repeated my message from July: “Hope is not a plan, and nostalgia is not mentorship.” Real challenges, but new opportunities are emerging.
Our 2025 Nature Physics note aged quickly. OpenAI’s release today makes clear that AI-assisted mathematical discovery at the highest level is no longer hypothetical.
I think that when we published this note in Nature Physics with @KenOno691 last year, nobody expected the era of “superhuman mathematicians” to arrive in 2026. And it's a big mess. I think that's what surprises me the most.
3/ Great mathematics needs people with the time and freedom to think, and IHES has given them that for nearly 70 years. If you can, please join us and help support it: https://t.co/RVbukjtL9o
1/ On November 26, the IHES Trust UK Gala in London will raise funds for @Institut_IHES, one of the world's great research institutes. Nine of the 14 mathematicians it has appointed as permanent professors have won the Fields Medal, most recently Hong Wang. 🧵
2/ This year's theme is "From Turing to Artificial Intelligence." I'm honored to join Eugénie Rives of Google DeepMind and Sir Dermot Turing, author of Alan Turing Decoded, as a speaker, and to share a few thoughts on what AI might mean for mathematicians.
I am thrilled to share my recent preprint with Kenny Lau and @KenOno691@axiommathai
We proved the HJO conjecture, a q-series identity arising from the geometry of singular curves, revealing new connections to cylindric partitions, compositional Dyck paths, and shuffle theorem.
Why should we care whether humans still understand mathematics if AI can solve the problems for us? Fair question. But we can say this at least: Historically, understanding has been a route to predictive power — with enormous benefits for our lives.
https://t.co/44cUXi8sce
Welcome @powdr_labs to Axiom!
@powdr_labs has been rethinking the software stack from the bottom up around machine-checked correctness: from autoprecompiles for zkVMs to a ZK-native stack with crush + powdr-wasm.
We share this vision. Very excited to build together 🚀🚀🚀
17 countries are rewriting K-12 math curricula in response to the emergence of data science, computational thinking, and AI. The US is still deciding what to do.
A new @theNASEM report compares them, and we're discussing it live on Sept 17, 12 PM EDT, with @WIRED's Kate Knibbs moderating. Free and open to all.
The Ramanujan Journal is honored to publish one of the final papers by the late Prof. Florian Luca (1969-2026), whose unexpected passing in July was a great loss to the math community.
We are proud to share this contribution to his enduring legacy: https://t.co/ue5cZZddCw
@stevenstrogatz@yet_so_far I had Herstein as a freshman at Chicago in 1986. He was tough! He threw chalk at us when we got answers wrong. I loved it. He really made us learn.