Big brain tech dude got yet another clueless take over at HackerNews etc? Here's the place to vent. Orange site, VC foolishness, all welcome.
This is not debate club. Unless it’s amusing debate.
For actually-good tech, you want our NotAwfulTech community
A recent two hour interview with Kokotajlo I do not have the time to go through all this but if anyone does feel free to go ahead.
What I will say tho is that the thumbnail is definitely some kind of clickbait, because of the “uhm actually we were never confident in 2027” thing
Steven Bartlett barely has 3 brain cells to rub together, so this seems like the perfect outlet for Kokotajlo's message.
this Diary of a CEO channel has always struck me as the kind to fall for any kind of bs hook line and sinker
John Michael Greer writes:
Second, I’ve had various people try to launch discussions about AIs — that is to say, large language models (LLMs) and the utilities they power — on this and my other forums. The initial statements and their follow-up comments always end up reading as though they were written by LLMs — that is, long strings of words superficially resembling meaningful sentences but not actually communicating anything. That’s neither useful nor entertaining. Thus I’ve decided to ban further discussion of this latest wet dream of the lumpen-internetariat here, and have extended that ban to LLM-generated content of all kinds. https://ecosophia.net/july-2026-open-post/
Good for you, you peak-oil meme-magic druid, for keeping your corner of the net weird.
Yeah, I began losing interest in Greer as it became clear that he was perfectly happy squatting in the middle of the red-brown alliance during the Trump era. His critiques of industrialism and unquestioning belief in technological progress broadly align with what we discuss here, but he will always coddle MAHA types and tale a shrugging "well, what can ya do?" attitude towards people like Trump, as it fits his preference for cyclical theories of civilization.
I noticed a couple months ago that he actually managed to dig Nick Land out of whatever tweaker den that guy's been hiding in for a podcast, which says a lot about what he's willing to indulge these days.
That Greer/Land crossover podcast is one of the funnier things I've listened to in a while, thanks for the ref, I guess I had missed it. https://www.youtube.com/watch?v=q-pcSez32-s if others are interested.
Land opens with a Greer/Land yin-yang and we're off the to races. Greer definitely brings up Spengler. Then Land starts talking about "Father Chad the Colorado Exorcist" and Greer is just yes-and-ing about the Solomonic tradition of binding demons.
... Half way through, that's enough for one sitting.
Relatedly, C2C had a guy on talking about the singularity plus the bible. https://kfiam640.iheart.com/featured/coast-to-coast-am/content/2026-07-22-1100-the-best-of-coast-to-coas-the-singularity-best-of-coast-to-coast-am-72/
Seems like America is definitely getting ready for this stuff to cross over to the Left Behind crowd, I predict more Tucker segments. I'll post some of the other guys I've found, digging into the "AI is actually the demonic spiritual energy of surviving nephilim" theory rn.
I like to keep my analysis separate from my word-magic (making things happen by saying them). Greer is trying to remake the world and not just work up his courage which is a harmless and consensual form of word-magic.
I still respect the definition he gives for magic, which he probably pulled forward from one of the many past ritual magicians he's studied. Magic is the change in consciousness in accordance with will... but we must always remember that that doesn't say whose consciousness, nor whose will.
"Any Sufficiently Transparent Magic…" by Damien P. Williams talks about chatbots as sorcery https://pages.charlotte.edu/damien-p-williams/wp-content/uploads/sites/1342/2024/01/DPW_Any-Sufficiently-Transparent-Magic_Edits-PrePrint.pdf
and covid vaccine denier
Debian has a new general resolution to ban LLM use in Debian packaging: https://www.debian.org/vote/2026/vote_002
Coincidentally, the second proposal in favor of conditional LLM use in Debian verily reads as generated to me. :)
Not to mention its singular argument literally is "many Debian contributors find AI tools helpful", right after "recognizing that AI-assisted contributions raise many concerns".
Concerns that were clearly laid out in the counter-proposal.
The "conditions" listed to address these concerns are hilariously toothless. Contributors should (sic!) stay accountable and responsible for any legal, functional, and procedural fuck-ups, and, like, really not be one of those thousands of dicks who are the reason for this discussion in the first place, and things will be fine! It's so simple!
Yeah, cool. I'm sure every contributor thoroughly checks if any of the generated output violates any existing license or infringes on someone's copyright. I'm sure you can just prompt an LLM to check that for you, though!
Is this naïveté, or deliberate disregard? I don't know, and that makes me mad.
Coincidentally, the second proposal in favor of conditional LLM use in Debian verily reads as generated to me. :)
I used AI tools to improve my initial draft, and (7.) was actually suggested by an AI tool.
@djb @dgerard It' could be undermined by Linus's tolerance of AI slop in the kernel, and by other upstream projects accepting LLM code, notably (to me) vim and pandoc.
The unknown original provenance of code generated by LLMs means that many open source projects may soon be in violation of their own license terms.
@cstross @djb @dgerard the proposal starting with copyright was such a welcome change, I find it wild how many projects just have their head on the sand over the copyright status of LLM generated code. You can be unethical as you like generating code, open source licences permit that, but they are all about copyright, they are built on it.
long rant about math
The recent big AI results in math have left me in quite a bad mood. I believe the main ingredient is Lean, which is a formal language resembling a programming language. Math proofs written in Lean can be verified deterministically with a computer, which really helps mitigate the hallucination problems of LLMs. Back in the days of pure scaling LLMs and Sam Altman talking about Dyson spheres, I was skeptical that LLMs would do math, but I did think that perhaps in the future, techniques using these formal languages could contribute to math. Well, it seems like OpenAI and Anthropic had the same idea and I underestimated their limitless checkbooks. Many of the biggest results were announced by mathematicians directly working for them (and presumably being paid a handsome amount).
For what it's worth, after the last of these big announcements, I decided to try one of these AIs on one of my small problems that I couldn't figure out. The AI did give a solution. That is, until I checked it thoroughly and realized that the it had a subtle but severe mistake that made it useless. I reprompted it, it failed again, and I ran out of tokens. I'm sure someone will tell me to shell out $200/mo for a pro subscription.
In the math and computer science research community, this is all anyone can really talk about right now. Honestly, after watching this whole AI bubble starting from the very beginning, I think the AI companies want to use marketing to stoke fear that all mathematicians will be replaced. But now, I am just too tired to argue. The amount of alarm and the extraordinary social pressure to use LLMs has soured me to this whole research thing. If becoming a researcher will one day require supporting these evil AI companies, I would rather just not. My dream job now is Factorio developer.
A lot of annoying people in technical areas view the world in terms of an intelligence hierarchy: the smartest people do math and physics, the slightly less smart people do coding, and the dumb people do everything else. So if AI can do math then it can do anything else. But, as an example, it is abundantly obvious now that AI is not replacing filmmaking. The techbros might be moved by arguments about how hilariously expensive video generation is, and how all these videos are 2 second clips stitched together so you won't feel the uncanny valley. But the real reason is that nobody wants to watch slop made with no intention or feeling. Also, nobody wants to support the AI companies, which could not act more evil even if they tried.
The mania in math right now quite resembles the mania in software engineering back in December-February, when Claude Code definitely solved all coding. I don't think the boosters expected that by April, everyone would be complaining about how expensive it all was while seeing an endless parade of vibe coding disasters (and no increase in productivity). Even if math research works out perfectly well (which is a still big if), it's not going to pay the bills. They would need to find a use case in the real world, where hallucinations can cause serious damage and cannot be formally prevented. And they have certainly tried. Math will not change the fact that all of this will collapse.
I reprompted it, it failed again, and I ran out of tokens. I’m sure someone will tell me to shell out $200/mo for a pro subscription.
One of the things that’s never clear from the reporting on ai successes is exactly how much actual paid human time went in to achieving those successes. This was especially notable in the fable-based security work… a huge amount of person-hours went into turning fable-detections into actual meaningful vuln reports.
A lot of demonstrably clever and capable people are involved with the llms-for-maths work, and a lot of money was spent on their time and supporting their work. Replicating it without your own stable of mathematicians and computer scientists and all the tokens they can eat is probably impractical.
I believe the main ingredient is Lean, which is a formal language resembling a programming language. Math proofs written in Lean can be verified deterministically with a computer, which really helps mitigate the hallucination problems of LLMs.
Fwiw, lean is a general purpose programming language, though despite microsoft’s efforts no-one uses it for that. I think its popularity with mathematicians came as a bit of a surprise.
Anyway, the other important thing that didn’t get reported on is that building the formal definition of the problem is not trivial! Obviously I don’t need to tell you that, but from the reporting you’d think that an llm solved all these problems, when in fact it was an llm in the hands of some very capable people who absolutely did not just prompt the thing in plain english.
Anyone hoping for self-marking homework here is going to be disappointed… lean slop confirming to formal spec slop is just expensive slop. Reviewing regular genai code is awful, even the thought of reviewing genai dependently-typed code makes me want a new career.
It is surprising how many exceptionally strong mathematicians have started working for OpenAI and Anthropic. These people would have easily become professors at top universities if they stayed in academia. I think many mathematicians, especially the competitive ones at the top, have a "progress at any cost" attitude (and I'm sure the paychecks helped). As for the results, you still need good mathematicians to sift through all the output to identify that the proofs are valid.
I would honestly be positive about universities developing their own specialized math AI (in an ethical manner) to help mathematicians get these kinds of results, but right now, AI is inseparable from these evil companies. Thankfully, I believe this is a likely outcome in the future because the AI companies will one day implode.
From what I've seen, most prompts are in plain English. I suppose the part where the AI parses the statement correctly is much easier than the part where it boils a couple lakes in the process of bashing its head against the wall trying millions of different combinations of random shit from the literature to slap together a proof. For one of the big results (cycle double cover), the prompt specified that the AI could use 64 subagents and was required to not give up for at least 8 hours. The tokenmaxxers would be proud, we didn't need that forest anyway. Thank god math doesn't have a CTO to look at the expense reports.
A lot of annoying people in technical areas view the world in terms of an intelligence hierarchy: the smartest people do math and physics, the slightly less smart people do coding, and the dumb people do everything else. So if AI can do math then it can do anything else. But, as an example, it is abundantly obvious now that AI is not replacing filmmaking.
Going by those annoying peoples' logic, filmmakers are smarter than coders, because LLMs can (allegedly) program, but they can't make a good film. I have no wider point to this, I just find this really, really funny
I believe the main ingredient is Lean, which is a formal language resembling a programming language. Math proofs written in Lean can be verified deterministically with a computer, which really helps mitigate the hallucination problems of LLMs.
100% this. Also, looking back at an earlier example that was actually written up in more detail, AlphaGeometry 1 got 28/30 problems, but entirely stripping out the LLM from the system, the symbolic logic proportion alone could get 14/30, and replacing the LLM with different heuristic methods could get 18/30 and 21/30 (for different methods).
Even if math research works out perfectly well (which is a still big if), it’s not going to pay the bills. They would need to find a use case in the real world, where hallucinations can cause serious damage and cannot be formally prevented. And they have certainly tried. Math will not change the fact that all of this will collapse.
The boosters and LLM companies still believe LLMs get their current level of performance by generalizing and not just memorizing facts (and maybe a wide shallow pool of weak heuristics). So they are hoping by pushing the LLM performance up in some narrow domain they can churn out synthetic data for, they will see some large general improvements in LLM performance.
"The mania in math right now quite resembles the mania in software engineering back in December-February, when Claude Code definitely solved all coding. I don’t think the boosters expected that by April, everyone would be complaining about how expensive it all was while seeing an endless parade of vibe coding disasters (and no increase in productivity). Even if math research works out perfectly well (which is a still big if), it’s not going to pay the bills."
^MBAs at Open AI desperately trying to figure out who is willing to buy a counter example for 100 billion USD . pee en gee
Am I right in understanding that almost all the big name results in LLM-derived math recently come from big publicity projects in which someone spent ungodly amounts of money to have the thing nondeterministically fuzz huge numbers random seeds leading to independent random outputs around a topic, putting out simulacra of ideas which could be then deterministically algorithmically checked? In fields where something like finding one counterexample to a conjecture would be a big deal, or where you just need to try a huge number of possible solutions until you happen to hit on one that works, rather than follow a long train of logic?
Among the three big results I've looked at (unit distance problem, cycle double cover, Jacobian), two were counterexamples and one of them had a short 3 page proof using ideas from the 1970s. The Jacobian conjecture is an extreme case because a single counterexample is enough (for unit distance, you technically need a family of counterexamples), and it is easy to check with very basic computations. It is telling that all of these announcements came from OpenAI or Anthropic employees, who presumably have unlimited access to their AI. Nobody really knows how many resources they spent on this, or what else they tried. Nobody really seems to care about this question, either.
I think there is a phenomenon where supposedly hard questions are much easier than expected, because by chance nobody found the right approach for a while, and eventually it becomes famous as a "hard problem" which makes nobody want to attempt it.
What I'm more worried about is many people starting to use AI to try and prove small lemmas for them in their projects. Of course, a $200/mo subscription is absolutely necessary to them. This honestly feels like a repeat of Claude Code back in February. The software engineers eventually realized that AI is absurdly expensive after the AI companies realized that spending $14000/mo to service a $200/mo subscription is a bad idea. If the AI vendors couldn't squeeze money out of rich software companies, what exactly are they gonna get out of poor mathematicians and universities? Also, there is the cognitive decline caused by overuse of LLMs that has yet to set in.
guys, the reason Codeberg decided to ban "AI" is because they host the "open slopware" list, and that inspires "meatpuppets"!
Multi-trillion-dollar (-self-valued) industry that’s the future of all work and that everyone who doesn’t use it gets left behind and everyone who does use it evokes superheroically productive turns out to be helpless in the face of a small group of outspoken and minimally organised opponents?
I see.
"AI bet goes awry: Oracle fires 21,000 employees, then hit a $7 billion power hurdle"
https://www.msn.com/en-us/money/news/ai-spending-spree-hits-600b-as-oracle-fires-21000-employees-to-fund-boom/ar-AA28vWuD
"Oracle’s Worst Stock Crash in 25 Years" "Has Cost Larry Ellison $213 Billion in 10 Months"
https://finance.yahoo.com/markets/stocks/articles/oracle-worst-stock-crash-25-113002772.html
It's going to be Oracle to collapse the house of cards, isn't it. Come on Oracle, die and take down the USA economy with you. Make the people happy, Oracle.
"In short, Oracle's 65% decline is historic, but the stock's future depends less on its past and more on whether its AI investments produce durable cash flow"
well good luck with that, Oracle! :D
guess the pentagon is footing the power bill?
"Up to 7 billion" over a period of 10 years can hardly dig Oracle out of the hole it has dug itself into, though...
they had big corporate contracts for their other, boring and profitable part of the business, but that won't save them from openai going broke. it's like 50x less than they need