I’m pretty sure because of the black box problem, we actually -don’t- understand how generative ai works…

We also don’t know how our minds work, but don’t give these models more weight than they deserve, we, collectively, have no idea how they arrive at the conclusions they do. Maaaaaaybe some of the people who develop them do, but honestly? Probably not. For the same reasons we don’t understand the human mind. It’s too complex.