How about this. I think it is pretty well known that pilots and astronauts are trained on simulations where some of the information they get from “tools” or gauges is wrong. On the surface it is just simulating failures. But the larger purpose is to improve critical thinking. They are trained to take each peice of information into context and if it doesn’t fit, question it. Sound familiar?

AI spits out lots of information with every response. Much of it will be accurate. But sometimes there will be a faulty basis in it that causes one or more parts of the information to be wrong. But the wrongness almost always follows a pattern. In context the information is usually obviously wrong. And if you learn to spot the faulty basis, you can even sus out which information is still good. Or you can just tell it where it went wrong and it often will come back with the correct answer.

Talking to people isn’t all that different. There is a whole sub for confidently wrong on reddit. But spotting when a person is wrong is often harder because the depth of thier faulty basis can be soo much deeper than an AIs. And, they are people, so you pften can’t politely question the accuracy of what they are saying. Or they are just a podcast… I think you get where I am going.