another post was from a physicist who described how he got it to invent a property which definitely did not exist - with citations which also did not exist but sounded credible.
if I run into it I will link it, it was weeks ago, either on elon or pmarca’s twitters.
I think the key is to use it for basic foundational research of broad trends....not drilling down on specifics too much. basically journalistic depth.
I think journalists should be very afraid.
I don’t think scholars should be.
Used like that, I’ve found it to be a great tool.....