The interesting thing is how it becomes harder and harder over time to draw a line between what human brains do and what these models do. Once you have seen ch

The interesting thing is how it becomes harder and harder over time to draw a line between what human brains do and what these models do. Once you have seen chain-of-thought reasoning in action (does the book even cover that?) you realize that defense of human superiority is strictly a "God of the gaps" game. Arguments that work today won't hold up tomorrow, because they're built on unreliable definitions, special pleading fallacies, obviously-temporary tech limitations, and general hand-waving. It's not that the technology is magical or special, it's that we're not. That being said, finding new ways to study the nature and limits of cognition and consciousness after 2000+ years of unproductive navel-gazing feels very magical and special.