Here's the proposition: "ChatGPT is fuelled by our intimate online histories. It's trained on 300 billion words, yet users have no way of knowing which of their data it contains." Well, I've had this conversation with people before. If it has been posted on the public internet, it shouldn't be thought of as private or intimate any more. Never mind what the law says - different laws in different jurisdictions say lots of things. The fact of the matter is that if, say, you posted a comment on Twitter, it is there for everybody to read, including an AI, which (just like a human) will take these words and the other (maybe) 300 billion it has read, just like you, and create something new, building on that. Pretending that these words are 'private' and 'intimate' doesn't make them so.
Today: 1 Total: 85 [Direct link] [Share]
