Stop using generative AI as a search engine

Ana Navarro-Cardenas used ChatGPT to make an erroneous claim about presidential pardons. It turns out that just scratches the surface of what Gen AI gets wrong.

The Verge
@verge why do people keep being surprised at this? LLMs paraphrase text, they do not search for information. If there is no text to paraphrase regarding the issue in a prompt, they combine elements into similar patterns and thus produce fictitious results.
@tanyakaroli @verge I think the problem is that an extremely large portion of humanity doesn't know this, including journalists.
@jfml I think that’s right. And pretty worrisome. @verge