Google trained Bard on ChatGPT chats from ShareGPT

Re:Google trained Bard on ChatGPT chats from ShareGPT by ar9av on Thursday March 30, 2023 @12:41PM (#35375488)

I love that OpenAI uses a ton of other peoples work to train their model, yet when someone uses OpenAI to train their model, they get all up in arms.

As far as I'm concerned, OpenAI has decided terms of use don't exist anymore.

[ Parent | on HN ]

Re:Google trained Bard on ChatGPT chats from ShareGPT by modernpink on Thursday March 30, 2023 @12:45PM (#35375553)

Where are they up in arms?

[ Parent | on HN ]

Re:Google trained Bard on ChatGPT chats from ShareGPT by jug on Thursday March 30, 2023 @12:49PM (#35375619)

OpenAI is training on data that is against their terms of use? That reads like a serious allegation. What is this all about?

[ Parent | on HN ]

Re:Google trained Bard on ChatGPT chats from ShareGPT by cycomanic on Thursday March 30, 2023 @03:34PM (#35377790)

OpenAI is training on copyrighted data without a licence. I would argue copyright law has much stronger legal standing than some ToS.

Now OpenAI is arguing their training is fair use, but that has certainly not been legally established so far and could just as much be used as a defence against ToS violation.

So in short yes OpenAI is pretty much doing the same thing.

[ Parent | on HN ]