I think we should start calling it "distillation terrorism" just to make it sound even more absurd.

It's pure model murder, and if you call it anything else, you're an anti-American communist.

> Translation: Alibaba will continue distillation attacks using accounts that aren't directly attributable to it's own corporate infrastructure.

What's a "distillation attack"? How is it different from simply distillation?

It's pretty much the same as when "installing programs on your computer" is called "sideloading". Deliberately deceptive, weaponized language to make it seem like a bad thing.

using infected machines as proxies would be a fair line in the sand

You wouldn't distill a car.

I would distill all the cars.

> The target doesn't want to be distilled.

So?

Fraudsters don't want to be jailed, their victims don't want to be scammed, employees don't want to be laid off, etc.

What the target wants is irrelevant - what society wants as enforced by laws is what is relevant, and as the leading AI providers have demonstrated, simply grabbing other people's copyrighted stuff for learning purposes is perfectly fine!

If they already think this practice is fine, why would I believe that their concerns about this are real?

I was only describing the difference not taking a side.

(Mis)anthropic already performed "distillation attacks" on the internet.

i can see why they want to stop it but 1. you have to pay for the "attack" 2. these AI companies trained on copyrighted content without permission or attribution to anyone who's data was used to train.

As long as they're paying for the tokens, there's no attack . Otherwise you have to call training on copyrighted material theft.

They are not paying for most tokens. The actual users in China do. All they need is the logs.

Anthropic still gets paid.

Unlike the vast majority of people Anthropic stole from.

In that case it's already bought and paid for by the users, is it not?

Did Anthropic perform "distillation attacks" when they hoovered up the entire internet?