Forum Discussion
Bing Chat / Copilot "Invents" fake "facts" - "I don't know" should be a valid answer!
Thank you for your reply, too. I have absolutely no confidence that Microsoft will stop patting itself on the back long enough to read the feedback OR this post, but we can hope! 🙂
This discussion opens up an interesting topic,( Trust in AI answers, in my opinion should be limited )
Only when it provides sources of information and I can verify it - it is valuable information.
Deleted HollyJ2022 Yeah, it's often hard to 100% agree with an AI generated answer. I've gotten false facts before. But remember that it doesn't always make things up, as it reads the web and articles. It does make it's opinions, based on the web, but they aren't always accurate.
- It used to claim it couldn't use Search and had no data more recent than 2021, but now it does - proven because I asked it about attribution for a fairly recently posted poem that I wrote and published on my own website, and it found that.
Bard, at least, says "I don't know" or "I don't have enough information to answer that" when asked the same question I asked Bing the other day. Sometimes, that is the BEST answer; Bing should be taught to say "I don't know" rather than inventing non-factual answers.
- If "fake news" was a problem when humans created it and disseminated it, it is an exponentially worse problem with AI and deep fakes, as mis- and dis-information can be created and spread around at lightning speed. Because it can present this authoritatively and using correct spelling and competent grammar, it will be seen as "credible" by the people who never checked a fact, and much more challenging for us compulsive fact-checkers - and that will prove deadly.
This is very disturbing and I don't know how the developers have set up security, in this case I agree with you!
"Because it can present this authoritatively and using correct spelling and competent grammar, it will be seen as "credible" by the people who never checked a fact, and much more challenging for us compulsive fact-checkers - and that will prove deadly."
- In my experience, it has no problem citing non-existent sources in its footnotes, or sources that do not back up its assertions.
