https://m.youtube.com/watch?v=EzPPmiKFf5I

Right, because then we have to decide at what point our use of AI becomes slavery.

If you think my behavior here was truly ban worthy than do it because I don't see anything in the I would change except for engaging at all

That isn't true.

The copyright to derivative works is owned by the copyright holder of the original work. However using illegaly obtained copies to create a fair use transformative work does not taint your copyright of that work.

> Even if not, you agree that they infringed on copyright of something close to all copyrighted works on the internet and this sounds fine to you?

I agree that they violated copyright when they torrented books and scholarly arguments. I don't think that counts at "close to all copyrighted works on the Internet".

> The consequences and fines from that would kill any company if they actually had to face them.

I don't actually agree that copyright that causes no harm should be met with such steep penalties. I didn't agree when it was being done by the RIAA and even though I don't like facebook, I don't like it here either.

"Learning" is an established word for this, happy to stick with "training" if that helps your comprehension.

> LLMs don't "learn" but they _do_ in some cases, faithfully regurgitate what they have been trained on.

> Legally, we call that "making a copy."

Yes, when you use a LLM to make a copy .. that is making a copy.

When you train a LLM... That isn't making a copy, that is training. No copy is created until output is generated that contains a copy.

I don't think this is legally true. The law isn't fully settled here, but things seem to be moving towards the LLM user being the holder of the copyright of any work produced by that user prompting the LLM. It seems like this would also place the enfringement onus on the user, not the provider.

> If someone hires me to write some code, and I give them GPLed code (without telling them it is GPLed), I'm the one who broke the license, not them.

If you produce code using a LLM, you (probably) own the copyright. If that code is already GPL'd, you would be the one engaged in enfringement.

> Please respond to the strongest plausible interpretation of what someone says, not a weaker one that's easier to criticize. Assume good faith.

https://news.ycombinator.com/newsguidelines.html

> We have evidence of LLMs reproducing code from github that was never ever released with a license that would permit their use. We know this is illegal.

What is illegal about it? You are allowed to read and learn from publicly available unlicensed code. If you use that learning to produce a copy of those works, that is enfringement.

Meta clearly enganged in copyright enfringement when they torrented books that they hadn't purchased. That is enfringement already before they started training on the data. That doesn't make the training itself enfringement though.

Right, because you would have done more than learning, you would have then gone past learning and used that learning to reproduce the work.

It works exactly the same for a LLM. Training the model on content you have legal access to is fine. Aftwards, somone using that model to produce a replica of that content is engaged in copyright enfringement.

You seem set on conflating the act of learning with the act of reproduction. You are allowed to learn from copyrighted works you have legal access to, you just aren't allowed to duplicate those works.

I am talking specifically about the ideas you are disputing:

>> partially signed off to corporate entities who are more than happy to consent away their access into our effects.

I haven't conflated anything. You may be confused and think we're talking about ownership or physical access though.

There are almost 2000 disbarments annually in the US.

The california bar recieves 1 compliant for every 10 law licenses in the state every year.

There's a wikipedia page on notable disbarments.

Legal malpractice suites are on the rise.

If you are going to assert that legal malpractice is not legitimate concern, I think the burden of evidence is on you.

Good thing that isn't what happened. It is called an "analogy" and is not a factual statement of equivalence.

Do you self host your own email? No? Those are "papers" that your email hosting provider can consent to providing law enforcement access to without a warrant.

Do you use search engines? Your search history is in the same boat with the search engine company.

Don't use a VPN? All of your internet traffic is in the same boat with your ISP

You use a VPN? All your internet traffic is in the same boat with the VPN.

The list goes on and on. It is almost certainly true that some company has private information about you that they can turn over without a warrant.

As is, you are being politely called out as incorrect because you are asserting someone people don't believe and not providing any argument, evidence or justification.

There are, of course, limitations to synthetic characters. Even with those limitations there are plenty of entertaining experiences to crafted.

The real challenges are around maintaining and safely operating automous robots around children in a way that isn't too expensive. These constraints place far more limits than those on synthetic characters in video games.

There is a strong population on HN that dislikes walled gardens. In my experience there are also plenty of people who disagree. There's also a large population that doesn't like EU tech regulations.

The ratio between different parts of the HN population can change significantly depending of stuff like time of day and headline draw. I don't find it particularly surprising, it isn't like HN is a monolith with internally consistent views across the entire population.