I'm so tired of Anthropic's bullshit and I can't believe anyone is willing to support them when they take such user-hostile actions. There's no other industry where a provider can say "we're going to take this job" and then later say "we're not going to give you the deliverables because it might have some of our IP in it, you still have to pay for it though".
Everyone is mentioning crawlers, but wouldn't most traffic from a datacenter's ASN usually be bots just in general given that there's not many humans there to generate human traffic?
Here's an example French hosting company which has something like 14.8% "human" traffic, (the Yandex is under 5% human) though how much of that is actual humans and not advanced things puppeteering headless browsers is another question.
It seems fairly obvious that it's meant to be an analogy for ChatGPT only supporting textual outputs until now (disregarding that it had some limited capabilities to show static images).
It looks fine in sRGB, but I can see how it would be bad with the horrid default profiles that monitors ship with. Graphic designers should really be aware of that at this point.
Presumably the fact that they're heavily trained to reply in this way? I don't know about the rest of the paper, but this part sticks out as a really odd claim unless I'm entirely misunderstanding this part.
Thanks for pointing out, maybe I should be more explicit in the wording - I mean we don't fully know what drives the voice in LLMs. Models that are post trained as instruct models are expected to have the disclaimers, but what about base models (those that are trained on just a lot of text)? How do they talk about themselves? What happens when you strip off the chat template from instruct model's prompt? I hope the rest of the paper makes the questions clearer, but I will try to do better in the abstract next time, as you point out this sentence is kind ambiguous. Thank you!
> we don't fully know what drives the voice in LLMs
Who is "we"? I, working in an LLM startup, know exactly what drives the base "voice" in the LLMs we train, because we have a process to select for it. OpenAI and Anthropic surely do too. Saying broadly that something is not well-understood in a scientific paper because it's not understood to casual observers is, uh, not very rigorous.
> The strange thing is that the base models (before RLHF) use the "experiential" voice, even though they are not incentivized to do that.
(Replying to your quote from another comment)
This is a matter of the training material. We have trained models that do not do that. I'm not exactly divulging trade secrets here. It should be really, really obvious that if you train a model on chat-conversation-like patterns of speech it will infer probabilities for how to continue a textual sample that will differ from the probabilities learned from being trained on narration, prose, or informational patterns of speech, even without RLHF.
Yes. It was obvious enough that they were able to explain exactly why that happened, and the explanation was exactly as obvious as you'd expect. Knowing why something undesirable happens doesn't mean it can't happen by accident. Have you never written a bug before?
> but what about base models (those that are trained on just a lot of text)? How do they talk about themselves?
Those don't have a themselves, because they can only continue text. A base model can only plausibly continue along the lines of what a character would say in a novel or what the narration would say in a story or in an article. Post-trained models may tie "I"-talk to actually observable effects they caused in some RL environment, or to how RLHF humans rewards its self-talk. But there is no themselves in a base model.
I suspect they're also losing a lot of university subscriptions after they started advertising how they offer a service to make LLM-written text look human-written, if I recall correctly they even explicitly advertised that you could use it to cheat. I don't know what they were thinking with that marketing campaign given that universities are probably one of their biggest market segments.
I also doubt this. When I observe people out in town, most of them stare at their phones non-stop. To get to pick them up that often, they'd first have to put them down every now and then.
reply