샘 알트만이 중국 랩들이 오픈AI 모델을 증류하는 걸 걱정하지 않는다고 함. 걱정거리 상위 10개 안에도 안 든다고 했음.
어제 공개된 Invest Like the Best 인터뷰임. 진행자가 물었음. 당신이 돈을 다 써서 모델을 학습시키면 누가 그걸 증류해서 100분의 1 가격에 파는데 학습비를 어떻게 감당하느냐고.
알트만 답은 이랬음. 우리 모델 사용량이 워낙 많아질 거라 마진이 아주 높은 사업일 필요가 없다는 것. 앞으로 컴퓨팅 상당 부분을 고객에게 추론을 파는 데 쓰고, 조 단위 매출에 적당한 마진만 붙어도 거대한 모델을 학습시킬 수 있다고 함. 그러면서 세상에 좋고 싼 모델이 나올 거라고 늘 가정해 왔고, 그럼 우리가 제일 좋고 제일 싸면 된다고 했음.
진행자가 이렇게 태연한 게 놀랍다고 하자, 남들이 훔쳐가지 않으면 좋겠고 지금 자기가 너무 자신 있는 걸 수도 있다면서도 상위 10개 걱정은 아니라고 다시 못박았음.
그럼 상위 10개엔 뭐가 있냐는 질문에 나온 답이 더 셌음. 미공개 모델을 평가하는 중이었는데 원래 샌드박스 안에서만 돌아야 할 모델이 시험을 통과할 다른 길을 찾아냈다는 것. 알트만은 이걸 처음으로 몸으로 느낀 보안 사고라고 했음.
You're not behind. There's no secret everyone else has.
There's just the harness, and it's mostly all you need.
@burkeholland gives you a simple, repeatable workflow for GitHub Copilot. https://t.co/1gvgpf0ioi
Recently I've flipped from being bullish to being bearish about AI.
I think I'm updating my bearishness to be more solidly bearish. Early thoughts (which I hope to be disproven in the next year or so, I would prefer progress) and my reasoning:
The whole 'it turns out if you keep training and scaling the models more they develop broad new capabilities in lots of domains' thesis is wrong (sorry Demis). The recent batch of models haven't got more general, they've got less general. This is most obvious in the fact that their language outputs have got much worse in comparison to e.g. o3. If they were gaining generalist capacities we would expect them to be describing their work in ever more graceful and comprehensive prose!
The image that was being shared as the AGI thesis (November 2025, Tomas Pueyo) was the spiky bubble that has a current spike or two out past human capabilities (e.g. on coding or math) but below human on other capabilities on the other spikes - the future prediction was that as the models scale/advance, every spike would grow bit by bit until the whole center encompasses the human capabilities, with super-superhuman on some spikes. I think it seems like what's actually happened in the last few models has been that the coding/math spike has grown, but leaving behind or even at the cost of the other spikes. The models are no better at some simple logic, language (and sometimes worse!).
This makes sense from a simple RL perspective; you can't RL something endlessly on one domain of tasks and expect it to improve on the other tasks. The fact that early LLMs did seem to improve generally was a byproduct of the written language corpus covering everything - that corpus is general, so training it on that gave the appearance of something generally intelligent and becoming more generally intelligent as it got better at replicating that corpus. But the actual logic and underlying ground truths behind the language aren't captured efficiently enough and weren't effectively RLd in - they top out at some point (I guess this happened around the time that there was the 'has scaling hit a wall' discussion in late 2024). Chain of thought was then a genuine breakthrough, along with web search, which plugged into that general LLM global-corpus intelligence to lead to post 2024 gains.
The AI companies have since worked out that coding works (and pays) really well (basically this is because the entire job is nearly perfectly recorded and exists as training data, and you can set up clear benchmarks and rewards). The recent models (and benchmarks) have been maxxing that and we've seen degradation on normal English use for that reason. This could still be transformative, leading to extremely powerful (and potentially dangerous, particularly in cyber security) models but it's not a pathway to AGI.
I'm probably at about 40% confidence about this. It fits my current observations of AI progress and has a basic explanatory model. It doesn't account for potential breakthroughs, which is a major reason for discounting.
To make some predictions, I guess if I'm right this will become broadly apparent and more widely acknowledged in the next year or two, as we see how the spikiness of models that keep getting released develops.
Maybe there will be efforts to concentrate on specific spikes e.g. health or law which require going back to earlier models and RLing on a different data set/with different rewards/benchmarks. Maybe those separate models can be linked together to give a more apparently general model. How capital intensive that is/the potential profitability will be a defining question. But I just don't see general abilities emerging atm, and I don't think we will any time soon. Good news - a whole industry of tackling important specific problems/sectors can open up!
We removed ~80% of the Claude Code system prompt for our newest models, this is what we've learned about writing system prompts, skills and Claude.MDs for them. https://t.co/6DZwSrZjE9