I was talking about specific queries in which people have falsely attributed an idea to an author. This is a common problem.

For me to assert that an author did not say something, I have to buy every single work of the author to show that the idea was misattributed.

It can take a year and several thousand dollars to answer this type of question.

Since AI has access to the life work of an author, I can ask "did author say 'x' or a logical equivalence of x"?

If the author did say "X" then it gives me a citation. If the author did not say "X" the ai is often able to find out who actually said "X"

The AI world is developing alternatives to generative AI that are even better at asking such questions.