I disproved the long-standing Guy’s conjecture for subtraction games. By encoding a binary counter into a subtraction game with just 26 moves, I constructed win/loss sequences whose least eventual period grows exponentially the largest move! Best part, NO ai vibecoded proofs! 🔥
Looks like the proof took 6 weeks of “collaboration” with Claude.
Alpöge doesn’t say it outright but the tone suggest Anthropic DIDN’T use 10,000 agent swarms like OpenAI or spend $10M to brute-force it.
if that’s true, it would be 100x more impressive than OpenAI’s approach.
and it also dodges the main mathematician criticism that frontier labs just brute force the discoveries by dumping resources into it.
WOW!
🚨 JUST IN: Turkish-American mathematician Levent Alpöge, working with Claude, claims to have proved one of the hardest problems in mathematics: the Generalized Riemann Hypothesis!
The Riemann Hypothesis is one of the seven Millennium Prize Problems.
The announcement includes a massive Lean formalization, which, when written out in English, runs to over 250,000 pages! 🤯
If verified, this could be the most significant mathematical breakthrough achieved with AI to date.
Is this real? https://t.co/Q7YDMVYZPV
"I believe I have, working with Claude, discovered a proof of the Riemann Hypothesis..the method easily extends much further, resolving the Generalized Riemann Hypothesis, and I expect, with more time, the Extended Riemann Hypothesis too."
Oh my god: Anthropic mathematician Levent Alpöge (@__alpoge__) says he has proven the Riemann Hypothesis, open for 167 years.
"I believe I have, working with Claude, discovered a proof of the Riemann Hypothesis."
He says the method also covers the Generalized Riemann Hypothesis. The proof is out as Lean code (a proof language where a computer checks every step), so experts should be able to check it quickly, Scientific American reports.
Same Alpöge who was behind this year's Jacobian conjecture result and the Navier-Stokes work with Tristan Buckmaster.
I assume we will get an official statement very soon from him!!
BREAKING: A mathematician at Anthropic claims to have used the company's AI to solve the Riemann hypothesis, the most notorious open problem in all of math
https://t.co/nEXAyIkvSt
Holy shit. Levent Alpöge from Anthropic just claimed Claude has resolved the General Reimann Hypothesis!!!
Proof is 250,000+ pages long. Lean is released.
GRH is the 2nd of 6 Millenium prizes solved by AI, after Navier-Stokes. This may be the biggest AI math result of all time.
I disproved the long-standing Guy’s conjecture for subtraction games. By encoding a binary counter into a subtraction game with just 26 moves, I constructed win/loss sequences whose least eventual period grows exponentially the largest move! Best part, NO ai vibecoded proofs! 🔥
Our colleague Eva Yi Xie, a Princeton neuroscience PhD student, was detained by ICE at Reagan National Airport on her way home from SfN, where she’d just presented her poster. She’s being held in Chantilly, VA. Please help us bring her home. 📷
Our colleague Eva Yi Xie, a Princeton neuroscience PhD student, was detained by ICE at Reagan National Airport on her way home from SfN, where she’d just presented her poster. She’s being held in Chantilly, VA. Please help us bring her home. 📷
@AmericaGetBig That’s the next question I want to tackle! 26 is an upper bound from our construction, not a proven minimum. I suspect the counter can be encoded more efficiently with fewer moves. I may leave that as an exercise for the reader…
I’m curious to see whether Abbott Elementary gets renewed for another season. There’s still room to explore heavier themes, like death or grief, but I wonder how the writers will keep expanding the show w/o killing depth
Turing Award winner Andrew Yao just teamed up with Kimi founder Yang Zhilin for the first time ever.
They dropped a brutal theoretical computer science research exam built to break every model.
It’s called TCSExam, the first benchmark evaluating frontier LLMs on the full range of Theoretical Computer Science (TCS).
We're talking algorithmic complexity, cryptography, automata theory, and rigorous mathematical proofs.
The actual physics of computation.
And the results are brutal.
Frontier models, the smartest systems on the planet—spectacularly failed.
When asked to write boilerplate web code, they look like geniuses.
But when forced to reason through actual theoretical computer science, their performance completely collapses.
Here is what happens when you push them past their training data:
• They hallucinate proofs that look mathematically perfect but are logically broken.
• They fail at multi-step deduction that requires true reasoning instead of pattern matching.
• They rely heavily on memorization, masquerading as intelligence.
This exposes a massive illusion in the AI industry.
LLMs haven’t learned how to reason. They have just memorized the syntax of reasoning.
If a problem has been solved on StackOverflow a thousand times, the AI is flawless.
But the moment you step off the map into pure, unmapped theory, the illusion shatters.
One of the coolest uses of AI in education is making teachers’ lives easier. I recently built a tool for my English teacher that lets her upload essays and rubrics, review them side-by-side, and leave interactive feedback, combining Google Classroom with smarter grading tools.