Jaeden Schafer

data set.

And like that was what they were using to train the models.

And the judge said that is okay.

But because they, you know, pirated them first, he made them pay a fine.

And the fine on that was $1.5 billion, which is pretty massive.

And then I think the authors received about $3,000 per book, there's about 500,000 copyrighted titles.

So I think that was pretty big.

But, you know, that wasn't financially crippling to the company, they have $183 billion valuation, they've raised a lot of money.

So I think Anthropic was like, and I honestly, I think they, they did wrong.

But also those getting those 500,000 copyrighted titles and using it to train their model made their model the best model for like a lot of literary stuff.

So like,

Honestly, I think as far as a cost of business, it was worth it for Anthropic.

I'm not saying it was morally right.

I'm just saying I think it was worth it because it gave them a competitive edge over OpenAI for quite a while as far as like tone and style of writing.

So they probably viewed that as worth it.

They paid the fine, which I mean, $1.5 billion is still a pretty massive penalty.

They paid the fine.

But what's interesting is the same people involved in that lawsuit was like, sweet, we won this lawsuit.

Now we're going to go for music.

And what's interesting is we're going from, you know, getting $1.5 billion for 500,000 books to $3 billion for the lyrics of 20,000 songs.