LogFAQs > #971775561

GastroFan posted...

Sunday Morning on CBS a little over a month ago did a story on ChatGPT and interviewed a couple of people, one of whom talked about how the company that produces ChatGPT takes samples from all the literary works available and tosses them into their database. Their concern was that, while some of the samples they use are from literary works that are in the public domain (that is are no longer under copyright protection), other samples are or might be, which becomes a totally different matter as far as compensation to the artists, etc.

But it's not just wholesale pasting a response, it's cutting up phrases together.
Which, unless you want to say all books are breaking the copyright of the dictionary, is not a strong argument.