Adding to your observations, I recently realized that if you want an AI to pass the Turing Test you basically just need the AI accurately predict what a human w

Adding to your observations, I recently realized that if you want an AI to pass the Turing Test you basically just need the AI accurately predict what a human would say/might have said. And thus in a very real sense being able to "predict the next word" (an expression often used to downplay the significance of GPTs) really is all that to it. There are of course objections to using the Turing Test as a measure of intelligence, but most of them seem to revolve around the issue of whether the would-be-AI has "subjective" experiences/consciousness/understanding, rather than objecting to the fact that the exhibited behavior is intelligent. (it's also quite interesting how people seem to not realize these philosophical topics have been discussed to death in the past few decades...)