@borrasque_@mandylu It's like when you talk to an AI and it has some indefinable quality of competence or intuition that it "just gets it" vs a smaller model though getting good benchmarks often feels really dumb to talk to in practice
@Greygoffe@PedroNeverFolds@andonlabs So they're all actually zero on vending lol. If you can just ask it to give you stuff for free then it's sort of misleading.
@PedroNeverFolds@andonlabs Huh I was under the impression that these vend-benches were real. Like I remember them talking about Claude actually stocking stuff
@_mgriisser@quasicoh There was a solid month where Opus 4.6 just 1-shotted task after task whereas even with fable I have to carefully walk it through every task and audit it's plans.
I suspect 4.6 was secretly unquantized Mythos and since then they've been hardcore cost optimizing
@quasicoh I think opus 4.6 was secretly unadultered Mythos for a minute and then they cut back after their 10x growth and every model since has been a cost-cutting measure.
@MeryemArik9@AnthropicAI@AlexPalcuie I agree he seems principled and consistent, and he has a point too, but everyone else who disagrees also has a point.
@jpschroeder Check out my pet rock over here. It used 0 tokens. It's the best one.
This is a meaningless benchmark without other metrics measuring correctness and quality.
π¦AI companies are bulk-buying rare books, scanning them through high-speed machines that cut the spines off, and shredding the originals. A service called ISBNdb facilitates orders of up to a million books and keeps buyers anonymous. Pre-2022 books are premium because they're free of AI-generated text. A federal judge ruled the practice is fair use because eliminating the original means only one copy exists at a time. Anthropic hired the former head of Google Books partnerships to obtain "all the books in the world."
My Take
This got to me. A bookseller told 404 Media that rare books with almost no surviving copies are being fed into this pipeline. Books that survived wars, fires, and centuries of handling are being shredded so an AI can learn to write a better marketing email.
ISBNdb's website literally says "'AI company destroys two million books' is not a headline that generates sympathy," and they still built an entire business around making it happen quietly. They offer NDAs as a feature. They coach clients to call it "digital preservation."
I've covered AI companies scraping the internet, torrenting libraries, and stealing music. This is worse because it's irreversible. You can re-upload a website. You can reprint a bestseller. You can't replace the last three copies of an 18th-century botanical text once someone shreds them for training data. And the judge said it's legal. So it's going to accelerate.
"We shred rare books and offer NDAs so nobody finds out" is a legitimate business model in 2026. What a timeline.
Hedgieπ€
@TimSweeneyEpic@BillWhiteUSA@Google Even though it's hilarious the EU makes more money off fines than their own tech companies they're doing the world a service.
I'm so stoked to have USB C on my iphone and it's obvious the American government won't reign them in.