OpenAI and ChatGPT Lawsuit List
originality.ai
originality.ai
It’s neat to be involved. Though I wonder what they could possibly want, since books3 was public. That was the whole point of making it.
They’ll request whatever information they think I might have that could help their case, and might call me in for a deposition.
And they’ll probably lose. It’s pretty clear that the output matters, not the input.
What sucks is that we can’t open source any training data anymore. So no one can make a competitor to ChatGPT. Even eleuther was cowed into silence with legal threats. They’re trying to make a non-copyrighted version of The Pile, but it’ll be hard to amass the data quality necessary. Meanwhile private companies can just vacuum up everything.
> Riehl never published the false information generated by ChatGPT but checked the details with another party. It’s not clear from the case filings how Walters’ then found out about this misinformation.
Amusingly the root cause of this was they "asked ChatGPT to summarize a real federal court case by linking to an online PDF". This is one of the most frustrating ChatGPT bugs: it can't fetch URLs (at least not the default verison) but it will happily pretend that it can. I wrote about that here: https://simonwillison.net/2023/Mar/10/chatgpt-internet-acces...
Also, I feel like the summaries in the OP could be improved since I tried to give it a fair reading and I got it wrong:
"This alleged defamation emanates from the purported dissemination of misleading information to a member of the journalistic profession, subsequently leading to the promulgation of inaccurate and erroneous news coverage pertaining to a federal lawsuit concerned with matters of civil rights."