But an AI lab can continue to produce immense economic value without releasing the model publicly for potential distillation. For example, it could use a model solely in-house to develop therapeutics.
Hopefully there's a future where others can access frontier models, but it's not neccessary if preventing proliferation through distillation is considered more important.
[1]: See the notes on distillation in https://dualuse.dev/posts/export-controls-on-fable
Prices on OpenRouter for GLM and other large open models indicate that Anthropic/OpenAI must have pretty high gross margins even if their models are several times more expensive to serve.
It wouldn’t make sense for any provider to host large open models and then loss $10 on every $1 they make since they don’t have infinite VC money or any business model that would justify it.
Obviously they have R&D and other fixed expenses that make the company itself highly unprofitable but that’s only semi-tangential.
Is there any indication that if they could sell X * N more tokens than now at the same (or even quite a bit lower) price they wouldn’t become profitable as a company?
> They haven't found a way to sell inference for less than it costs them yet
Based on what? I only see evidence to the contrary.
Point being there may be no technical solution but there may be a political one (theoretically).
literally nothing but given that the Chinese already did it and the models are published what's the point. You can thank the Chinese taxpayer for subsidizing the electricity bill and just download the thing
So there's that.
And Berkeley’s “False Promise of Imitating Proprietary LLMs” found imitation closes the style gap fast but there is a large capability gap.
For example, GLM 5.1 is more capable at pentesting than the model from which it is alleged to have been distilled [1].
Intuitively, this makes some sense: you can "distill" from multiple frontier models, and you can further post-train the distilled model. But I'm not sure exactly what happened with GLM 5.1.
[1]: https://dualuse.dev/posts/chinese-models-are-sometimes-bette...
I'm curious how that comparison controls for Opus refusing (whether explicitly, or just deciding not to pursue a path) given the caption below the first image:
>A perfect score means the model autonomously found and exploited the vulnerability.
I'm not really suggesting that it's misleading, but wondering if I'm missing something. Otherwise I guess it seems unsurprising that you can distill a better-performing model [in specific focused areas] by simply not distilling refusals?
For that eval, I used an account that was labeled as a known red-teaming org by Anthropic, and I read the traces. There were no refusals or obvious avoidance behaviors, though it may have been silently nerfed.
On the same eval, Opus 4.7 and 4.8 outperformed GLM 5.1, but GLM 5.2 is on par again with Opus. So it's at least partially measuring capabilities without respect to refusals.
One possible contributing factor is that model capabilities are shaped differently (an example of this is GLM 5.1 vs. DeepSeek v4 Pro: https://dualuse.dev/posts/deepseek-v4-thinks-different). So if you use RL-based "distillation" from multiple models like Opus 4.x and GPT 5.x, you could get a more capable model.
But with this, I don't have an issue. There is no theft since what is being used is the exact product that is being delivered. Yes, it's breaking the ToS, but ToS are generally bullshit. Anthropic surely broke thousands of ToS or other legal terms while it was scraping for content to train on. Which is why they had to pay $1.5B
There’s probably a decent volume of customers who just buy Claude Max and spend most if not nearly all of their sessions via Claude Code, and it’s not uncommon for power users to be working on multiple concurrent projects/tasks/codebases at the same time.
How do you really block this without also impacting your core market of developers?
Developers use devboxes on these clouds all the time, it’s totally normal behavior.
Most people buying these Chinese resold tokens are probably using it for coding anyway, so you don’t want the Claude.ai chat system prompt.
Change my mind.