Anthropic is right! AIs trained without direct consent from the owners of the training material should be blacklisted!
We demand that our governments criminalize all AIs developed without full permission for use of training material, in support of Anthropic!
Steal from a thief and get 100 years of relief
Alien Vs Predator
Whoever wins, we lose.
Oh yes the classic thief crying about being robbed
Alibaba : please Claude, tell me what you can do. No secrets please.
Claude : I can do this.
Alibaba : Nice. But how? I don’t steal, it is only for research. Pretty please. No mistakes
Claude : Here you go …
Anthropic : :O Attacks!!
They stole out fairly stolen content…
I know the US government and legal system will side with Anthropic, because that’s what these fuckers do, but I hope they fuck off and, if they intend to escalate, China retaliates. These Silicon Valley companies are full of shit and full of themselves.
It’s not like the Chinese are any better. I want both sides to lose.
They’re a little better. Many of their models are open-source and can be run offline.
there are no open source models. There are open weight models.
No models make available the way in which they were produced, aka the source.
There is one: Apertus by ETH Zürich.
Everything is open-source if you know assembler
Weights are basically algorithm. It is not even obfuscated. Companies successfully produce derivative works from them, what is the problem?
One does not require git history for any Foss project, the exact way the code was produced IA not the requirement, it is the reproducibility that matters.
When I can train the ai on my own in the same way I can compile a program (yes I understand the huge difference in computing power necessary, that’s not what I’m talking about) then I’ll consider it equivalent.
As far as I’m concerned open weights are effectively compiled code. I can technically modify it, but I cannot rearrange the base components that were used to create it.
Even AI developers themselves cannot train AI in the same way as one compiles a program. Storing all the necessary data is just not feasible. To achieve results comparable to modern flagship models, one needs to obtain data in real time, use large amounts of generated data that is deleted after single use. And the training algorithm itself is not deterministic.
And I still fail to see how this is any different from requiring a git commit history. There is an algorithm present - that’s it. It has no other forms in which it can be presented.
From an open source perspective, this is how things are supposed to work, humanity making progress by standing on each other’s shoulders… Of course they want the rest of humanity to help them, with government subsidies, investment from our retirement funds, etc.
Oh NOW you shitheads care about intellectual property?
“Illicit” huh? They can fuck right off with that, their models were illicit to begin with.
Ah yes, the classic tale “Ali baba and the 40 thieves”
I thought it ceased to be yours when you train ai on it
Right?
… LOL
Even if it’s true, whacha gonna do about it? Call the cops?
