Performance problems in theorem provers is an old topic. I remember watching this and it was fun. https://youtu.be/m-iGCCuHBvY [Talk] 10 years of superlinear slowness in Coq (2022)
This is pretty interesting in a lot of non-surface-level ways. I can see OpenAI pushing for this as a sort of more durable moat compared to the now huge number of agentic harnesses that run on your own machine. This…
Replace hire with marry and this is *exactly* what politics was globally for all of human history. This is not something "behind us" btw.
> hey Claude vibeslop me a CUDA Anubis solver" route is on its way to being fundamentally dead. Lmao yeah no. I don't think a little argon2 is going to change shit all. I mean the thesis of Anubis itself is "scrappers…
In real life, there is always this feedback edge from the results to the methodology. Theoretically it's unscientific to do tweaks like this but in reality this is what actual science is because you need to see the…
Its also available finally to Pro users! Just took 24 hours.
The AI labs have out considerable effort in trying to find and patch lean exploits. They explicitly set agents and have them try to prove false. > Daniel used OpenAI internal models to discover new soundness issues in…
Are you guys just posting what you think might end being the link so you can farm upvotes or something?
This seems to be the announcement w/ the trust me bro numbers. https://x.com/Alibaba_Qwen/status/2094968708288680276
They are simply measuring in base 1.7 :^) Uber has grown 2 orders of magnitude!
Hosting is probably ~0% of the cost lol. The costs of the writers is what makes journalism expensive.
The blog post is gone but I can currently use it in the gemini chat website.
[dead]
Do API prices not affect usage limits for subscriptions? They do in Codex.
Haiku would have to be a banger, with a significant price drop, to make any sense. It's currently priced 33% above Gemini 3.7 Flash, and several multiples of 5.6 Luna.
I think the jobs thing will be a slowly then all at once thing. Firstly, AI is sort of like this to begin with. It's useless for something until it gets hill climbed and is useful for something. AI agents for chip…
Man that claude writing is bad. Really bad.
Smaller models + more effort has strong diminishing returns, especially if your goal is to save money. Sol already lacks judgement. It will absolutely add idiotic tests and comments. Luna is that but worse so if you…
Italy too. It's bizarre trying to hop on the trenitalia trains while a bunch of random people are standing on the platform right in front of the doors trying to get a few last drags in before the train leaves.
A random blight in Canada in that Bell still hasn't enabled IPv6 for residential plans. You can get 8 Gigabit symmetric home internet from Bell but not IPv6.
Whats stopping you? You could buy puts right now. Get a 210 strike put contract and if your thesis is that nvidias current 10 day slide continues you could make some money.
> Data centers in the states analyzed depend on about 3.4 trillion gallons of freshwater annually for electricity, around 12 times the annual water use of Los Angeles, Phoenix, and Washington D.C., combined. > Most…
If you think of LLMs as programs. The weights and inference code are very much a binary. While the training code and data are the true source. Since if you want to robustly modify the LLM that's actually what you need.…
You don't need to seperate the clean and dirty air. Mixing them isn't as efficient but it should still be roughly 63% efficient (1-1/e).
> Overall my view is that AI is structurally a technology that tends to concentrate power > Open-weights do help some with this but are nowhere near a sufficient solution Which is exactly why Anthropic contributes…
Performance problems in theorem provers is an old topic. I remember watching this and it was fun. https://youtu.be/m-iGCCuHBvY [Talk] 10 years of superlinear slowness in Coq (2022)
This is pretty interesting in a lot of non-surface-level ways. I can see OpenAI pushing for this as a sort of more durable moat compared to the now huge number of agentic harnesses that run on your own machine. This…
Replace hire with marry and this is *exactly* what politics was globally for all of human history. This is not something "behind us" btw.
> hey Claude vibeslop me a CUDA Anubis solver" route is on its way to being fundamentally dead. Lmao yeah no. I don't think a little argon2 is going to change shit all. I mean the thesis of Anubis itself is "scrappers…
In real life, there is always this feedback edge from the results to the methodology. Theoretically it's unscientific to do tweaks like this but in reality this is what actual science is because you need to see the…
Its also available finally to Pro users! Just took 24 hours.
The AI labs have out considerable effort in trying to find and patch lean exploits. They explicitly set agents and have them try to prove false. > Daniel used OpenAI internal models to discover new soundness issues in…
Are you guys just posting what you think might end being the link so you can farm upvotes or something?
This seems to be the announcement w/ the trust me bro numbers. https://x.com/Alibaba_Qwen/status/2094968708288680276
They are simply measuring in base 1.7 :^) Uber has grown 2 orders of magnitude!
Hosting is probably ~0% of the cost lol. The costs of the writers is what makes journalism expensive.
The blog post is gone but I can currently use it in the gemini chat website.
[dead]
Do API prices not affect usage limits for subscriptions? They do in Codex.
Haiku would have to be a banger, with a significant price drop, to make any sense. It's currently priced 33% above Gemini 3.7 Flash, and several multiples of 5.6 Luna.
I think the jobs thing will be a slowly then all at once thing. Firstly, AI is sort of like this to begin with. It's useless for something until it gets hill climbed and is useful for something. AI agents for chip…
Man that claude writing is bad. Really bad.
Smaller models + more effort has strong diminishing returns, especially if your goal is to save money. Sol already lacks judgement. It will absolutely add idiotic tests and comments. Luna is that but worse so if you…
Italy too. It's bizarre trying to hop on the trenitalia trains while a bunch of random people are standing on the platform right in front of the doors trying to get a few last drags in before the train leaves.
A random blight in Canada in that Bell still hasn't enabled IPv6 for residential plans. You can get 8 Gigabit symmetric home internet from Bell but not IPv6.
Whats stopping you? You could buy puts right now. Get a 210 strike put contract and if your thesis is that nvidias current 10 day slide continues you could make some money.
> Data centers in the states analyzed depend on about 3.4 trillion gallons of freshwater annually for electricity, around 12 times the annual water use of Los Angeles, Phoenix, and Washington D.C., combined. > Most…
If you think of LLMs as programs. The weights and inference code are very much a binary. While the training code and data are the true source. Since if you want to robustly modify the LLM that's actually what you need.…
You don't need to seperate the clean and dirty air. Mixing them isn't as efficient but it should still be roughly 63% efficient (1-1/e).
> Overall my view is that AI is structurally a technology that tends to concentrate power > Open-weights do help some with this but are nowhere near a sufficient solution Which is exactly why Anthropic contributes…