Mathematicians are unhappy about OpenAI. The gist of their argument is that they form a community that trains young people. When AI started producing breakthroughs on hard mathematical problems, I asked what a very smart 17-year-old would feel. Do you still choose a math major and train yourself to prove difficult results by hand? This crisis has been coming for a long time. When I was a kid, the four-colour theorem was proved by a computer, in 1976. An intense debate followed. Does it count as a proof? I wrote my PhD thesis using symbolic algebra software. To my knowledge, nobody then would publish their scripts. I tried to include mine. I was told it would make me look bad. The letter says: "In recent months, the success of AI in solving major mathematical problems has made headlines even outside mathematical circles. But solving problems is only a tool and proxy for achieving the primary goal of conceptual understanding and insight. Forgetting this in the world of AI may turn the tool against the primary goal. Often these solutions are announced in a rush, leaving no time for a proper writeup, the isolation of new methods and ideas, and citing relevant previous work of others. As in all creative professions, this raises severe attribution and plagiarism questions. We are witnessing a general threat to intellectual work, with misalignment between the outcome of the use of AI and its initial purpose. In many fields and activities, years of training have traditionally served not only to produce a final answer or product, but also to develop understanding and the ability to formulate new questions and ideas." They worry that kids will not choose to become old-school mathematicians. That is a reasonable fear. They do not seem to consider that some kids might still do mathematics, just in a very different way. Doron Zeilberger, in 2009: "Teaching computers how to discover and prove mathematical results is certainly the way to go, and I believe that mathematicians who continue to do pure human, pencil-and-paper, computer-less, research, are wasting their time." sites.math.rutgers.edu/~zeil… The letter implies that people, because of AI, will stop having ideas, or will stop taking the time to understand the issues. If that were true, mathematics would continue only on computers, with no humans interested, or it would stop. Both scenarios assume that people care about mathematics only when they can claim credit. I am not sure why I cannot study a proof generated by AI if I want to. I can give talks about it. What becomes less likely is the reward of having been the one who proved the result. The rest of the letter makes a big deal of credit. What if the AI builds on what it read and does not give proper credit? Where is the evidence that AI is worse than human beings at citing sources? The letter notes that mathematics contributes to society. It never considers that faster progress might increase those contributions. My stance is simple. Mathematicians, software developers, engineers, lawyers, physicians will all learn to work with AI. You cannot put the genie back in the bottle. Only a totalitarian world government could try, and some of us would rather avoid that outcome. Difficult proofs will now be built with AI, just as most code will be written with AI. People who insist on pen and paper should view themselves as artists. Other mathematicians will work with AI. They will have no trouble finding interested kids. They will contribute to society.

So uh, what the fuck, this is a story for the ages if true, I can’t believe I’m alive rn. I’m too stupid to explain this well but it seems that Levent and an NYU professor nearly solved Navier-Stokes (or vibe collabed with frontier models to do it, mostly Claude and Codex). Per the prof, OpenAI caught wind of it, threw a team and massive compute at it after the leak, and now claims their model finished it. He asked if they’d trained on his Codex sessions with all his drafts in them. No answer. They offered to post their result the day after his and publicly say he “deserves the Clay Prize,” or let him write it up solo with Levent dropped as an author (because he works for ant, this is all about ipo hype ig LOL). He said no and that he’d go public. Open ai C suite guy then says: “Why would you ruin your career?” “If you don’t want me to be nice, then I don’t have to be nice.” Wtffff am I understanding this right? Galois-tier math lore (if we don’t get vibe paperclipped first)

I am a mathematician but I avoided commenting on this because the problems are (far) outside my domain. Even if I sat down to read the technical details (which I have not), it would be hard to appreciate the context of prior work, etc. But I've gotten enough sense of things through colleagues to be confident about some aspects. First of all, all the mathematics is fundamentally correct. How impressive are the results? There is a spectrum of reactions (as with everything), but the consensus seems to be all the results could realistically land (by prior standards) in the very highest tier of publications, e.g., Inventiones for the math papers and "Best Paper Award" at STOC for the TCS papers. What does this mean practically? Prior to 2026, a Ph.D. student whose only result was one of these 10 would likely land an interview for a tenure line job at a strong university. For some of the results, you could probably replace "tenure line" with "tenured". When Astra comes out, it will clearly be very disruptive to the existing system. I think it's good that the community has the time to react and prepare. Personally I think we should totally overhaul hiring and publishing practices. Are the results "Fields Medal Level"? I have not heard a mathematician answer Yes to this, and the quoted reason is invariably that the solutions are very clever and beautiful but not "deep". This is consistent with what I can see from a cursory glance. However, the same criticisms could be levied at certain human generated works which have won major prizes. Ultimately, comparing results is very subjective and also somewhat political, and I absolutely think that if a human had one of these results, and also the right influential senior advocate, then that advocate could make a strong case for the work to win major prizes. The fact is that most mathematicians, including prize committees, don't have the breadth or time to evaluate all the top results, so they rely on proxies for importance like: Is the problem longstanding? Is it famous? Was there a large community of people actively trying to solve it? For e.g. #3 (existence of non-sofic groups), the answer to each of these questions is undoubtedly Yes. I've always felt that people should pay less attention to these proxies and focus more on the actual mathematics; glad to see that is happening with the AI solutions, but the same standards should then also be applied everywhere.

we had a significant security incident during evaluation of our models. we are sharing what we have learned so far. thanks to @huggingface for the partnership on this. openai.com/index/hugging-fac…

Here are a few benchmark scores of K3 that have been officially confirmed This is a Fable/Sol class model that is strictly better than Opus 4.8 across the board at Sonnet pricing. Insane

GPT 5.6 Sol just hit CursorBench. The economics are brutal for Anthropic. Fable 5 Max: 70.5% at $17.32 per task. 103,525 tokens. GPT 5.6 Sol Max: 67.2% at $5.22 per task. 28,320 tokens. 95% of the performance. A third of the cost. 73% fewer tokens. Fable 5 kept the crown on raw score. It lost on everything that shows up on your invoice. And remember: GPT 5.6 comes with limits you can actually live with. The frontier is not an intelligence war anymore. It is a value war.

John Carmack: The code for AGI could conceivably be written by one individual After stepping down as CTO of Meta’s Oculus VR, legendary programmer John Carmack was trying to decide if he would work on nuclear fission or artificial general intelligence (AGI). He ultimately chose AGI. “I think [fission] is possible,” John says. “Somebody should be doing this, but it’s going to involve some politics, decent-sized teams, and a bunch of cross-functional stuff that I don’t love. While artificial general intelligence seems to me like the highest-leverage moment for a single individual potentially in the history of the world.” He explains: “With the things we know about the brain and what we can do with artificial intelligence . . . I’m not a madman for saying that it is likely that the code for artificial general intelligence is going to be tens of thousands of lines of code — not millions of lines of code.” He continues: “This is code that conceivably one individual could write, unlike writing a new web browser or operating system. And based on the progress that machine learning has made in the recent decade, it’s likely that the important things that we don’t know are relatively simple. There’s probably a handful of things — my bet is there’s less than six key insights — that need to be made. Each one of them can probably be written on the back of an envelope. We don’t know what they are, but when they’re put together in concert with GPUs at scale and the data that we all have access to, we can make something that behaves like a human being or living creature and that can then be educated in whatever ways we need to get to the point where we can have universal remote workers where anything that somebody does mediated by a computer, that doesn’t require physical interaction, an AGI will be able to do.” Carmack does not think this will be “unapproachably hard”: “That’s incredibly hubristic to say, but what I said a couple of years ago was that there’s a 50% chance that somewhere there will be signs of life of an AGI in 2030, and I’ve probably increased that slightly — maybe 55-60% now because I do think there’s a little sense of acceleration.” Source: @lexfridman (Aug 2022)

Claude Tag is a Trojan horse.  Not because Anthropic is doing anything evil. Because the incentives are obvious. Day one, this looks like a great feature: tag Claude in Slack, let it follow the thread, remember context, connect to tools, break down tasks, chase work, and act like a teammate. But that is exactly the problem. The moment your AI vendor becomes a shared coworker, it stops being just a model provider. It starts becoming the place where work is interpreted, remembered, routed, and eventually executed. That is not model lock-in. That is context lock-in. You are now renting your company back from them. Models can be swapped. Agents can be copied. But the memory of how your company actually works is much harder, maybe impossible, to move: the Slack scar tissue, the exception paths, the customer promises, the unfinished threads, the weird workflows, the implicit owners, the “we tried that in Q2 and it failed” knowledge. Once that lives inside one vendor’s agent layer, you are not renting intelligence anymore. You are renting your company’s operating memory. And the pricing model makes it even more dangerous. A human coworker has a salary. Claude has unbounded tokenized activity. The more work moves through it, the more the vendor captures not just IT spend, but labor spend. This is the enterprise bargain people will regret: Convenience now, and rapid decent into dependency. The right architecture is simple: rent the best intelligence from whoever is best this month. OpenAI, Anthropic, Gemini, open source, whatever. But own the context layer. Your company memory should be inspectable, permissioned, portable, and model-neutral. It should not be buried inside the same vendor that sells you the intelligence and the workflow surface. Claude Tag is useful. That is why it is dangerous. Rent the intelligence, but own the context. Or, regret later.

MidJourney just announced... a full body ultrasound! Yup... read on because it's as crazy as it sounds. "As powerful as MRI and as casual as a trip to the spa" They are calling it "the @midjourney scanner" Insane details: - First, the scale. The device uses 8,960 individual transducers arranged in a ring around your body - The precision is the most jaw-dropping part: it resolves motion at the picometer range. It can image internal tissues finer than the width of an atom. We are talking sub-atomic level diagnostic capability - The compute requirement is massive. The system processes 17 gigabytes of data per second. It takes 40GB of raw data to reconstruct just one cross-sectional slice. And they are planning to scan 100 slices? - Midjourney claims that fewer than 12 of these machines could perform more full-body scans than every MRI machine on Earth combined. Welcome to the future of healthcare! Not only these scanners are announced, they will exist in a "Midjourney SPA" - with hot tubs, saunas, cold plunges, and 9-10 whole body scanners.

this is art

Jensen Huang on his prediction about the next revolution: digital biology ‘It’s the next revolution. It’s going to be flat-out one of the biggest ones ever.’