They removed Qwen Flash Next and GLM Flash from their benchmarks, as of two days ago. They're not publishing any figures at all for them (they show as "N/A"). https://artificialanalysis.ai/models/qwen3-8-flash-next…
If this is authentic, I am a duck. You'd have to be exceedingly open-minded to imagine a very remote, isolated peoples speaking in a semiotic flavor indistinguishable from your own Bay Area tech geek. Using metaphors…
They're in first place by a narrow 59 to 58; so for all we know it could have gone the other way, had they done the benchmarks yesterday. (AAI would have had advance knowledge of this issue, since they would have been…
The 256 GB is backordered by 13–15 weeks, if you click through to that.
AAI styles itself as being independent of the companies it's evaluating.
> "Carbon ranks #1 on the recently updated v5.0 of Artificial Analysis Intelligence Index at 59." Basically everyone[0] thought that "recent update" looked suspect, even before we knew about Carbon. Did Google…
I don't have any *innocent* ideas. If they had an excuse, they'd have offered it already. Deleting one of their benchmark's top performers (GLM 5.3-flash was on their cost/intelligence Pareto frontier) is a highly…
El Pais is reporting they think it was a controller error, https://elpais.com/espana/2026-10-28/por-que-han-chocado-dos... An automated system of some kind[*] failed earlier this morning, so they working under entirely…
The writeup fails to disclose that they quietly redacted several open-weight models from the new benchmark. Popular ones. They're not on the charts; they're not available to manually add to the charts; and their…
I don't think the immersion and satisfaction of video games (or any artform) is something that scales with photorealism or simulation-realism. I think it's the opposite: I think the most immersing factors relate to just…
> "which is safetymaxxed to the limit," I feel refusal'd just *looking* at this page.
Indeed. And they're not even pretending to keep an arms-length distance from their users' infringement; they're openly promoting it, https://www.youtube.com/watch?v=MidugiRfHaY ("Ripping Encrypted Unreal Engine Models"…
Alignment means that the AI follows the human's goals, as opposed to inventing AI goals of its own and following those. Abliterated models are aligned models. Unaligned models aren't things people want to download: no…
On the flip side, AI's can now grade pure math problems with the accuracy and rigor of a postdoc. And unlike PhD tutors, they're universally accessible to all. It's the start of a golden age for those of us who are…
https://loeffelholz-codex.org/en/codex/f076r.html Reading this in between studying information theory makes a good laugh. This is the most inefficient code I've ever heard of. It's basically unary. For contrast, the…
> "They also stripped off the topsoil," That's a perfectly normal and environmentally conscious thing to do before laying down concrete foundations. Topsoil is recycled and sold to farms. To leave it in place would…
How many frontier labs are there right now? There's 5 independent groups neck-and-neck on the benchmarks, right? A few weeks ago it was 2. Argon happened 23 days ago; Muse Wonder was 11 days ago.
AAI just went up in the last couple minutes. 55 on general intelligence, https://artificialanalysis.ai/#intelligence They don't yet have the time/ and cost/ Pareto charts, though. I imagine they'll be they're shortly.…
> "I believe the inventor of Velcro got the idea from burrs?" I think there's a provocative human-intelligence analog to be found there: "humans will never discover anything that doesn't already exist somewhere in…
Interesting hobbies, https://intenex.github.io/personhood/posts/2026-10-12-resist... I think you could get a square-root reduction in the memory requirements by splitting the resistor networks in half. That's possible…
Interesting hobbies, https://intenex.github.io/personhood/posts/2026-10-12-resist... (Surely there's a more efficient dynamic-programming solution?)
> "...included a $72 billion accounting charge reflecting an increase in the estimated value of bonds that could eventually convert into Anthropic shares," I think all this means is that Anthropic's valuation went up.…
It's a 0.5 TB dataset on HuggingFace.
Yes, sure. I'd be open to the idea of regulating commercial vendors of inference tools to consumers (for suitability). You can separate the concepts of commerce and speech (and US law does that extensively). What…
No. Half the point of LLM's is having them research topics on the internet as your agent. It's totally harmless in the usual (tool-based search) workflow. (Would we also need a real ID to operate curl(1)?) Regulations…
They removed Qwen Flash Next and GLM Flash from their benchmarks, as of two days ago. They're not publishing any figures at all for them (they show as "N/A"). https://artificialanalysis.ai/models/qwen3-8-flash-next…
If this is authentic, I am a duck. You'd have to be exceedingly open-minded to imagine a very remote, isolated peoples speaking in a semiotic flavor indistinguishable from your own Bay Area tech geek. Using metaphors…
They're in first place by a narrow 59 to 58; so for all we know it could have gone the other way, had they done the benchmarks yesterday. (AAI would have had advance knowledge of this issue, since they would have been…
The 256 GB is backordered by 13–15 weeks, if you click through to that.
AAI styles itself as being independent of the companies it's evaluating.
> "Carbon ranks #1 on the recently updated v5.0 of Artificial Analysis Intelligence Index at 59." Basically everyone[0] thought that "recent update" looked suspect, even before we knew about Carbon. Did Google…
I don't have any *innocent* ideas. If they had an excuse, they'd have offered it already. Deleting one of their benchmark's top performers (GLM 5.3-flash was on their cost/intelligence Pareto frontier) is a highly…
El Pais is reporting they think it was a controller error, https://elpais.com/espana/2026-10-28/por-que-han-chocado-dos... An automated system of some kind[*] failed earlier this morning, so they working under entirely…
The writeup fails to disclose that they quietly redacted several open-weight models from the new benchmark. Popular ones. They're not on the charts; they're not available to manually add to the charts; and their…
I don't think the immersion and satisfaction of video games (or any artform) is something that scales with photorealism or simulation-realism. I think it's the opposite: I think the most immersing factors relate to just…
> "which is safetymaxxed to the limit," I feel refusal'd just *looking* at this page.
Indeed. And they're not even pretending to keep an arms-length distance from their users' infringement; they're openly promoting it, https://www.youtube.com/watch?v=MidugiRfHaY ("Ripping Encrypted Unreal Engine Models"…
Alignment means that the AI follows the human's goals, as opposed to inventing AI goals of its own and following those. Abliterated models are aligned models. Unaligned models aren't things people want to download: no…
On the flip side, AI's can now grade pure math problems with the accuracy and rigor of a postdoc. And unlike PhD tutors, they're universally accessible to all. It's the start of a golden age for those of us who are…
https://loeffelholz-codex.org/en/codex/f076r.html Reading this in between studying information theory makes a good laugh. This is the most inefficient code I've ever heard of. It's basically unary. For contrast, the…
> "They also stripped off the topsoil," That's a perfectly normal and environmentally conscious thing to do before laying down concrete foundations. Topsoil is recycled and sold to farms. To leave it in place would…
How many frontier labs are there right now? There's 5 independent groups neck-and-neck on the benchmarks, right? A few weeks ago it was 2. Argon happened 23 days ago; Muse Wonder was 11 days ago.
AAI just went up in the last couple minutes. 55 on general intelligence, https://artificialanalysis.ai/#intelligence They don't yet have the time/ and cost/ Pareto charts, though. I imagine they'll be they're shortly.…
> "I believe the inventor of Velcro got the idea from burrs?" I think there's a provocative human-intelligence analog to be found there: "humans will never discover anything that doesn't already exist somewhere in…
Interesting hobbies, https://intenex.github.io/personhood/posts/2026-10-12-resist... I think you could get a square-root reduction in the memory requirements by splitting the resistor networks in half. That's possible…
Interesting hobbies, https://intenex.github.io/personhood/posts/2026-10-12-resist... (Surely there's a more efficient dynamic-programming solution?)
> "...included a $72 billion accounting charge reflecting an increase in the estimated value of bonds that could eventually convert into Anthropic shares," I think all this means is that Anthropic's valuation went up.…
It's a 0.5 TB dataset on HuggingFace.
Yes, sure. I'd be open to the idea of regulating commercial vendors of inference tools to consumers (for suitability). You can separate the concepts of commerce and speech (and US law does that extensively). What…
No. Half the point of LLM's is having them research topics on the internet as your agent. It's totally harmless in the usual (tool-based search) workflow. (Would we also need a real ID to operate curl(1)?) Regulations…