361–370 of 909 posts
This would be a great "AGI" test. See if it can derive biohazards from first principles
Earlier quoted context omitted.
Slavery is still legal and widespread in most of the US, including California. There was a ballot measure to actually abolish slavery a year or so back. It failed miserably.
The slavery of free humans is illegal in America, so now the big issue is figuring out how to convince voters that imprisoned criminals deserve rights. Even in liberal states, the dehumanization of criminals is an endemic behavior, and we are reaching the point in our society where ironically having the leeway to discuss the humane treatment of even our worst criminals is becoming an issue that affects how we see our…
They choose to because extra money = extra commissary snacks and having a job is preferable to being bored out of their minds all day.
That's the part that's frequently not included in the discussion of this whenever it comes up. Prison jobs don't pay minimum wage, but given that prisoners are wards of the state that seems reasonable.
Earlier quoted context omitted.
Maverick and Scout were not great, even with post-training in my experience, and then several Chinese models at multiple sizes made them kind of irrelevant (dots, Qwen, MiniMax) If anything this helps Meta: another model to inspect/learn from/tweak etc. generally helps anyone making models
There's nothing new here in terms of architecture. Whatever secret sauce is in the training.
If you even glance at the model card you'll see this was trained on the same CoT RL pipeline as O3, and it shows in using the model: this is the most coherent and structured CoT of any open model so far.
Having full access to a model trained on that pipeline is valuable to anyone doing post-training, even if it's just to observe, but especially if you use it as cold start data for your own training.
I was hoping these were the stealth Horizon models on OpenRouter, impressive but not quite GPT-5 level. My bet: GPT-5 leans into parallel reasoning via a model consortium, maybe mixing in OSS variants. Spin up multiple reasoning paths in parallel, then have an arbiter synthesize or adjudicate. The new Harmony prompt format feels like infrastructural prep: distinct channels for roles, diversity, and controlled aggrega…
Fascinating, thanks for sharing. Are there any specific kind of problems you find this helps with? I've found that LLMs can handle some tasks very well and some not at all. For the ones they can handle well, I optimize for the smallest, fastest, cheapest model that can handle it. (e.g. using Gemini Flash gave me a much better experience than Gemini Pro due to the iteration speed.) This "pushing the frontier" stuff wo…
And it obviously works for code and math problems. My first test was to give the llm-consortium code to a consortium to look for bugs. It identified a serious bug which only one of the three models detected. So on that case it saved me time, as using them on their own would have missed the bug or required multiple attempts.
Earlier quoted context omitted.
I would guess the “secret sauce” here is distillation: pretraining on an extremely high quality synthetic dataset from the prompted output of their state of the art models like o3 rather than generic internet text. A number of research results have shown that highly curated technical problem solving data is unreasonably effective at boosting smaller models’ performance. This would be much more efficient than relying…
> research results have shown that highly curated technical problem solving data is unreasonably effective at boosting smaller models’ performance. same seems to be true for humans
Earlier quoted context omitted.
> Once someone hits AGI/SGI I don't think there will be such a unique event. There is no clear boundary. This is a continuous process. Modells get slightly better than before. Also, another dimension is the inference cost to run those models. It has to be cheap enough to really take advantage of it. Also, I wonder, what would be a good target to make profit, to develop new things? There is Isomorphic Labs, which seem…
> I don't think there will be such a unique event. I guess it depends on your definition of AGI, but if it means human level intelligence then the unique event will be the AI having the ability to act on its own without a "prompt".
The lede is being missed imo. gpt-oss:20b is a top ten model (on MMLU (right behind Gemini-2.5-Pro) and I just ran it locally on my Macbook Air M3 from last year. I've been experimenting with a lot of local models, both on my laptop and on my phone (Pixel 9 Pro), and I figured we'd be here in a year or two. But no, we're here today. A basically frontier model, running for the cost of electricity (free with a rounding…
Earlier quoted context omitted.
I don’t think they trained it for fact retrieval. Would probably do a lot better if you give it tool access for search and web browsing.
What is the point of an offline reasoning model that also doesn't know anything and makes up facts? Why would anyone prefer this to a frontier model?
I'm out of the loop for local models. For my M3 24gb ram macbook, what token throughput can I expect? Edit: I tried it out, I have no idea in terms of of tokens but it was fluid enough for me. A bit slower than using o3 in the browser but definitely tolerable. I think I will set it up in my GF's machine so she can stop paying for the full subscription (she's a non-tech professional)