Don't applaud fake "creativity" in AI-generated garbage. You're applauding a parody of creativity, and

[Ms. Toad]

82. Not a single one of the hundreds of AI summaries with links that I have reviewed -

has gotten it right.

At least if there is a link, you can to to the site and confirm the site is generally reliable, and can check the facts against other sources.

But then, why bother using AI for facts because you cannot depend on its citations to be accurate. It either cites to sources which don't say what it claims they say (about half to 3/4 of the time), or to sources which are applicable to the factual summary for which it is cited (the remaining portion - e.g. citing to verifiable Ohio law when the legal issue is governed by Federal law).

And most AI "facts" don't contain links - so you have to check every "fact" it contains. I have quizzed ChatGPT on facts on which it was trained (at least on which it says it was trained). The narrative it creates (repeatedly, on different issues) is a mixture of fact and fiction - with no distinction in believability. I only knew which was which because I was testing it on subject areas in which I knew the underlying facts. There were no "tells" the way people often have "tells" and worse, it made crap up when there was no reason to make it up - other than whatever predictive algorithm