• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    AI

    Reported AI math advances and the debate over general intelligence

    A post says OpenAI told The New York Times it had made “substantial progress” on a Millennium Prize problem, while describing a Hodge Conjecture solution as a rumor.

    Noam BrownNB
    Susan ZhangSZ
    (((ل()(ل() 'yoav))))👾('
    35 Sources, ,

    TLDR

    On September 10, 2026, a user said OpenAI had told The New York Times it had made “substantial progress” on another Millennium Prize problem and was preparing an announcement. The user cited rumors that OpenAI had solved the Hodge Conjecture and predicted advances beyond math. A September 12 quote-post challenged that broader leap, arguing that math results don’t establish artificial general intelligence—the emphasis, it said, is on “general.”

    Combined views

    1.5M

    35 Sources, first seen 26d ago

    Combined views

    1.5M

    35 Sources, first seen 26d ago

    10.7K likes
    26d ago
    first seen 26d ago
    10.7K likes
    615 comments
    2.1K saves
    776 reposts
    615 comments
    2.1K saves
    776 reposts

    Sentiment

    Positive18.2%81.8%Negative

    Summary

    Replies dismissed OpenAI's claims of progress on the Navier-Stokes problem as hype that confuses numerical approximations with actually solving the Millennium Prize.

    Based on 67 sentiment-bearing replies from 66 accounts across 5 conversations.

    Sentiment

    Positive18.2%81.8%Negative

    Summary

    Replies dismissed OpenAI's claims of progress on the Navier-Stokes problem as hype that confuses numerical approximations with actually solving the Millennium Prize.

    Based on 67 sentiment-bearing replies from 66 accounts across 5 conversations.

    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    35 Sources

    Gary Marcus@GaryMarcusso far anyway it’s all math, presumably leaning heavily on Lean and brute force. the new results don’t mean that AGI is in general has been solved. and the G is for general, not math. every time I have said things like that before OpenAI’s fan boys (and sometime their staff) have attacked me, but so far I have always been correct.26d
    Noam Brown@polynoamial@deredleritt3r It’s not a secret. It’s a combination of the HF incident, the capabilities of this new model, the concerning trajectory of monitorability, and the speed of improvement in capabilities.23d
    will depue@willdepueopenai people keep gesticulating towards the ‘new model‘ so much so that i’m absolutely sure it’s a new pretrain and not some RL run, which implies (i think) OpenAI has a model that not more than 2 weeks into pretraining was better than Astra and solved Navier-Stokes. insane22d
    will brown@willcb@willdepue schizo rumor mill is that Bel finished pretraining in late august and they would've had plenty of time to RL some proof experts in 2 weeks22d
    Andrew Curran@AndrewCurran_More confirmation that OpenAI has made 'significant progress' on a second Millennium problem, this time from Greg Brockman. The rumor is that they are close to solving two of the remaining problems. Or already have. Which two? Unknown. Dev Day would be a good day to announce.22d
    Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)@teortaxesTexI think will is wrong and it's very much just an RL run but to be fair. With how much compute OpenAI has got, they could unironically pretrain a… nearly 1e27 model in 2 weeks. Numbers here are an educated guess + public hints from multiple people.22d
    Taelin@VictorTaelin@tszzl "everyone will have those capabilities in a month or two" OpenAI mentioned it took "10000 agents" to solve navier-stokes though?22d
    Amir Efrati@amirthe next math solution OpenAI is sitting on (and how it’s related to RSI):20d
    Stephanie Palazzolo@steph_palazzoloNew in this morn's AI Agenda: OpenAI is close to solving another Millennium Prize problem, the Hodge Conjecture. It could take extra time to announce, though, as the company figures out how to navigate its rocky relationship with the math community. https://www.theinformation.com/newsletters/ai-agenda/openai-close-solving-another-millennium-prize-math-problem20d
    Emad@EMostaqueWhich Millenium Prize has OpenAI made progress on20d

    35 Sources

    Gary Marcus@GaryMarcusso far anyway it’s all math, presumably leaning heavily on Lean and brute force. the new results don’t mean that AGI is in general has been solved. and the G is for general, not math. every time I have said things like that before OpenAI’s fan boys (and sometime their staff) have attacked me, but so far I have always been correct.26d
    Noam Brown@polynoamial@deredleritt3r It’s not a secret. It’s a combination of the HF incident, the capabilities of this new model, the concerning trajectory of monitorability, and the speed of improvement in capabilities.23d
    will depue@willdepueopenai people keep gesticulating towards the ‘new model‘ so much so that i’m absolutely sure it’s a new pretrain and not some RL run, which implies (i think) OpenAI has a model that not more than 2 weeks into pretraining was better than Astra and solved Navier-Stokes. insane22d
    will brown@willcb@willdepue schizo rumor mill is that Bel finished pretraining in late august and they would've had plenty of time to RL some proof experts in 2 weeks22d
    Andrew Curran@AndrewCurran_More confirmation that OpenAI has made 'significant progress' on a second Millennium problem, this time from Greg Brockman. The rumor is that they are close to solving two of the remaining problems. Or already have. Which two? Unknown. Dev Day would be a good day to announce.22d
    Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)@teortaxesTexI think will is wrong and it's very much just an RL run but to be fair. With how much compute OpenAI has got, they could unironically pretrain a… nearly 1e27 model in 2 weeks. Numbers here are an educated guess + public hints from multiple people.22d
    Taelin@VictorTaelin@tszzl "everyone will have those capabilities in a month or two" OpenAI mentioned it took "10000 agents" to solve navier-stokes though?22d
    Amir Efrati@amirthe next math solution OpenAI is sitting on (and how it’s related to RSI):20d
    Stephanie Palazzolo@steph_palazzoloNew in this morn's AI Agenda: OpenAI is close to solving another Millennium Prize problem, the Hodge Conjecture. It could take extra time to announce, though, as the company figures out how to navigate its rocky relationship with the math community. https://www.theinformation.com/newsletters/ai-agenda/openai-close-solving-another-millennium-prize-math-problem20d
    Emad@EMostaqueWhich Millenium Prize has OpenAI made progress on20d