zlacker

>>ssgodd+(OP)
NYT article with a lot more context https://www.nytimes.com/2023/12/27/business/media/new-york-t...

>>pm90+B2
[edit: they have since opened a comment section to the article.] It is unfortunate that the NYTimes don’t allow reader comments to this article. I like some of the NYTimes content, but in this case use of chatGPT is infinitely more valuable to me than subscribing to the NYTimes, so I would like to explain this concept and the associated risks by their litigation without cancelling my subscription.

Maybe it is time to move training of models to Japan that has explicitly adapted AI friendly legislation that allows training on previously copyrighted materials. My best guess is that if the inputs were legally obtained, then the output doesn’t violate anything until someone publishes it. Similar to how reading a newspaper in a public library is legal but copying its content verbatim and republishing is not.

>>pama+f7
The ChatGPT subscription is more valuable because it's built on the theft of the NYT content and many other authors' work.

>>jprete+Ob
No. It is a technology that can massively accelerate human progress—I don’t buy the theory that OpenAI used NYTimes content that was not freely available to them. If you read all of the public internet you probably have lots of snippets of NYTimes articles. Regarding the reading of the wirecutter by a browser tool, I don’t know how much of it is available online without subscription (because I subscribe to the NYTimes), but arguably if it is available I’d expect a helpful AI to read it and other sources and give me a short recommendation for what I look for (and not the ads or sponsored links).

>>pama+Nd
It may have been freely available to them, but that doesn’t mean that they are free to reproduce its contents or otherwise make use of it at scale.

>>layer8+5u
The chatGPT model is not reproducing the contents nor making it available at scale. I, the user, ask the model to read the website and other websites and come back with a useful summary to me. This is not training data thus not violating copyright any more than a browser showing the full text would.