Well, I saw all of these technological steps you are talking about, but so far I think people are jumping to conclusions on generative AI.

What exactly is the use of a program that spits out plausible text that may or may not be true? And what is the path from generative AI to a system to understands the difference between truth and falsehood?

I do not see the obvious path from generative AI to accurate question answering. Certainly, more training is not going to do it.