Show HN: ChatLLaMA – A ChatGPT style chatbot for Facebook's LLaMA

When it comes to litigation, the parties will and resources to engage an aggressive legal adversary is critical aspect of assessing legal dynamics in a matter. For example, even in cases for which there is no direct legal leverage, party might be exposed to patent, copyright, etc infringement if the other party discovers they have a valid existing claim for which the other party is dependent on and access to it is revocable.

Beyond that, it’s unclear to me if there is code that is part of LLaMa that’s not subject to copyright claims, if such code was authored by humans. Someone violating the terms, downloading the code, and reposting it, even without the terms, would be irrelevant to the rights of the party on future use by a party if they notified they’re infringing; otherwise, able copyright notices and software licenses would be meaningless, because you could simple launder stolen property, which is obviously not the case. If you’re positive there no copyrighted materials used in deploying LLaMa, then obviously there’s no copyright claim or breach of contract. To be clear, I agree human author per the US Copyright Office is at present required to claim copyrights.

Lastly, my core claim is it’s easy to state something is legal, it’s completely different story to take on legal liabilities related to defending those rights. I for one would have no interest in litigation with Facebook regardless of the legal merits of my position. The OP based on there ignoring the topic, which I raised, nor for that matter is it likely anyone that read this thread is likely to take on the burden of publicly disputing if LLaMa is free from legal claims for commercial use.

> if LLaMa is free from legal claims for commercial use.

That restriction is only in the license though, which was not accepted. So it won't apply, right?

> otherwise, [all?] copyright notices and software licenses would be meaningless, because you could simple launder stolen property

That applies to trade secrets. If I knowingly induce you to leak a secret I can't use it, but if you leak the secret to the public in an unrelated fashion and I discover it, I can. It doesn't apply to copyright, as you note.

> If you’re positive there no copyrighted materials used in deploying LLaMa, then obviously there’s no copyright claim or breach of contract.

There's no breach of contract because there was no agreement. Therefore there's no license to use, so you're right - the copyrightable parts will be a violation. Basically everything except 'data' or machine translations of someone else's works.

And yes, I assume that there's some incidental content, at least, which will be infringing. Even a readme is copyrighted after all. And yes, of course any code which isn't already public elsewhere.

I think the weights themselves are the interesting bit though, and perhaps someone could rerelease them without any incidentals just to clarify the issue.

Me: how are you today?

ChatLLaMa: I'm doing great! How are you?

Me: I woke up a little late for the gym. I went to the gym for leg day and felt quite weak. I still had a good workout overall. Then, I went home, showered, and got some work done. Finally, I drove to the office while listening to a podcast

ChatLLaMa: I woke up a little late for the gym. I went to the gym for leg day and felt quite weak. I still had a good workout overall. Then, I went home, showered, and got some work done. Finally, I drove to the office while listening to a podcast.

???

>How long is a piece of string, when the string is 4km long?

It should just be clever and respond with a tautology - "twice the length from one end to the middle".

So we can deduce the training data is from before January ‘23.