Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Anthropic says Alibaba illicitly extracted Claude AI model capabilities

reuters.com

Post reply on HN
Print view View original

441–450 of 1001 posts

Re: Anthropic says Alibaba illicitly extracted Claude AI model capabilities

#441

Post by parineum »

fnord77 wrote: post #344
0xbadcafebee wrote: post #11

There's two basic kinds of distillation: 1) the massive [and dumb] method where you ask a question and use the answer as reinforcement (Black Box), and 2) more targeted distillation where you use one model to directly inform/train/guide another model (RLAIF). The latter is basically fine-tuning the model with direction from another model. Thousands of businesses do this every day to fine-tune. This is almost certainl…

Can you reach into the model and "transplant" weights directly?

If you have access to the weights, you can just use them as is...

Re: Anthropic says Alibaba illicitly extracted Claude AI model capabilities

#442

Post by nutjob2 »

Hamuko wrote: post #434
trymas wrote: post #431

Earlier quoted context omitted.

> Anthropic's IP was created by harvesting and "distilling" other people's IP. Copyrighted materials, and the commons... which they have essentially privatized. Anthropic and others argue that because LLMs don’t output full copyrighted works word for word - hence their LLMs aren’t infringing on copyright laws. I think (if this ever comes to that) Chinese lab should use same arguments against Anthropic. UPDATE: this i…

Isn’t the output of LLMs completely copyright-free in the US?

One lower court has said that the output of AI models is uncopyrightable.

But the real unsettled issue is if model training is fair use, and where copyright infringement might creep in to model output.

Re: Anthropic says Alibaba illicitly extracted Claude AI model capabilities

#443

Post by JumpCrisscross »

0xbadcafebee wrote: post #11

There's two basic kinds of distillation: 1) the massive [and dumb] method where you ask a question and use the answer as reinforcement (Black Box), and 2) more targeted distillation where you use one model to directly inform/train/guide another model (RLAIF). The latter is basically fine-tuning the model with direction from another model. Thousands of businesses do this every day to fine-tune. This is almost certainl…

> These complaints of distillation are inflating the problem

They’re also missing the point. What would have happened to a member of the Manhattan Project who, through personal pursuit of profit, neglected their duty enough to let the bomb leak?

Re: Anthropic says Alibaba illicitly extracted Claude AI model capabilities

#445

Post by camgunz »

nutjob2 wrote: post #442
Hamuko wrote: post #434

Earlier quoted context omitted.

Isn’t the output of LLMs completely copyright-free in the US?

One lower court has said that the output of AI models is uncopyrightable. But the real unsettled issue is if model training is fair use, and where copyright infringement might creep in to model output.

The copyright office itself also says this when it talks about determining authorship.

Re: Anthropic says Alibaba illicitly extracted Claude AI model capabilities

#446

Post by mcintyre1994 »

handoflixue wrote: post #419
0xbadcafebee wrote: post #11

There's two basic kinds of distillation: 1) the massive [and dumb] method where you ask a question and use the answer as reinforcement (Black Box), and 2) more targeted distillation where you use one model to directly inform/train/guide another model (RLAIF). The latter is basically fine-tuning the model with direction from another model. Thousands of businesses do this every day to fine-tune. This is almost certainl…

> But if you show them a jailbreak of their model that bypasses their safety, they'll tell you that any model can eventually be jailbroken so don't worry about safety. They claim two things: 1) The specific, available jailbreak for Fable 5 is not dangerous - this has been confirmed by multiple experts, and there is no credible evidence against this claim (in other words, Anthropic is probably correct) 2) It is imposs…

Supposedly the details of the ‘jailbreak’ are that you give it insecure code and say “fix this code”, and it does, and then you ask it for test scripts and that’s effectively an exploit against the unfixed code.

If true then I have no idea how anyone’s going to release a useful model that doesn’t have the same jailbreak. https://www.theregister.com/security/2026/06/15/feds-freaked...

Re: Anthropic says Alibaba illicitly extracted Claude AI model capabilities

#447

Post by dualvariable »

toobulkeh wrote: post #262
dualvariable wrote: post #236

Earlier quoted context omitted.

"Information wants to be free" Anthropic profited from training its models on all kinds of copyrighted information, live by the sword, die by the sword... Their model weights, training data, training methods, etc are all going to leak to China over time. Nobody on a site named _Hacker_ news should be all that upset about this.

You’re mistaking the original term hacker, a tinkerer of systems, for the black hat variety.

Hackers didn't use to spend a lot of time defending trillion-dollar corporations and their intellectual property rights.

Re: Anthropic says Alibaba illicitly extracted Claude AI model capabilities

#448

Post by da_grift_shift »

p4coder wrote: post #416
anon373839 wrote: post #380

Earlier quoted context omitted.

> These complaints of distillation are inflating the problem to make it sound worse than it is Unfortunately, the Reuters piece itself is complicit in this dramatization. The lede paragraph parrots Anthropic's talking point that distillation is an "attack", without using quotes that would alert the reader that this framing is a corporate talking point. Distillation is NOT an attack.

Agreed! I had to do a double take and check the URL. I thought I am reading a press release rather than actual reporting.

Same thing nowadays :^)

Re: Anthropic says Alibaba illicitly extracted Claude AI model capabilities

#449

Post by soperj »

p4coder wrote: post #416
anon373839 wrote: post #380

Earlier quoted context omitted.

> These complaints of distillation are inflating the problem to make it sound worse than it is Unfortunately, the Reuters piece itself is complicit in this dramatization. The lede paragraph parrots Anthropic's talking point that distillation is an "attack", without using quotes that would alert the reader that this framing is a corporate talking point. Distillation is NOT an attack.

Agreed! I had to do a double take and check the URL. I thought I am reading a press release rather than actual reporting.

That's exactly what they pay the publicist for.

Re: Anthropic says Alibaba illicitly extracted Claude AI model capabilities

#450

Post by killingtime74 »

suzzer99 wrote: post #415
killingtime74 wrote: post #244

Earlier quoted context omitted.

Maybe Fable is not as capable as thought? On the one hand they talk it up as world ending and on the other hand they can't manage bot accounts on their own service. I want to hear how this can be rationalised. From the article "every layer of control frontier US AI companies have added (geoblocking, phone verification, credit card requirements, and now live biometric KYC checks) has produced a corresponding layer of…

No system is foolproof. They'd have to be willing to throw out some % of good customers along with the bots. Amazon can do that because they have a monopoly already. Anthropic can't risk it when they're trying to grab market share.

In this case, being distilled is sort of existential to them. The false positives would just be losing some revenue (depending if profitable, not even losing profit).
Post reply on HN