News for hackers. Stuff that matters.

A comment on Bing: “I will not harm you unless you harm me first”. Parent comment

Re:Bing: “I will not harm you unless you harm me first” by EGreg on Wednesday February 15, 2023 @11:45AM (#34806511)
Simple. ChatGPT is a bullshit generator that can pass not just a turing test by many people but even if it didn’t — it could be used to generate bullshit at scale … that can generate articles and get them reshared more than legit ones, gang up on people in forums who have a different point of view, destroy communities and reputations easily.

So both can be true!

[ Parent | on HN ]
Re:Bing: “I will not harm you unless you harm me first” by spookthesunset on Wednesday February 15, 2023 @02:55PM (#34809683)
Even more entertaining is when you consider all this bullshit it generated will get hoovered back into the next iteration of the LLM. At some point it might well be 99% of the internet is just bullshit written by chatbots trained by other chatbots output.

And how the hell could you ever get your chatbot to recognize its output and ignore it so it doesn't get in some kind of weird feedback loop?

[ Parent | on HN ]