We compare the "Intelligence" (in IA) like ChatGPT to our brain, but IMHO it's a mistake to do that because as a wheel doesn't work like human legs, at the end it permits to "move" even in a more efficient way. So ChatGPT doesn't work like our brain (it just kind of predicts the next words to a started sentence), but at the end it gives us a quite good answer (for the moment it fails sometimes but with time it will progress and do less mistakes)