I agree with Socrates and too many people have the wrong memory of him, making his prediction come true. There was a great philosophical book last year, Open Socrates [1], that explains his methods and ideas are the opposite direction of how most people use AI. Socrates believed we can only get closer to knowledge through the process of open, inquisitive conversation with other beings who are willing to refute us and be refuted in turn. He claimed ideas can only be expressed and shared in dialogue & live conversation. The one-direction communication of all the media since books have lacked this, and AI's version of dialogue is sycophancy and statistical common patterns instead of fresh ideas.
I'm sure you could train an AI to be skeptical/critical by default. The "you're absolutely right!" AIs are probably always going to be more popular, though.
[1] https://www.nytimes.com/2025/01/15/books/review/open-socrate...