J
The predictive model he uses as a baseline for his approach is consistent with many of the other books I'll suggest. His emphasis on cortical columns is not something I've seen before, so I'll need more information to comment on the science there. However, as a model for an actionable approach to AI, I like the groundwork he's laying. I also think, though published before the explosion of LLMs, that he does a great job of articulating some of the limitations of the approaches that they're based on and are why I don't think they'll ever scale to intelligence.
The predictive model he uses as a baseline for his approach is consistent with many of the other books I'll suggest. His emphasis on cortical columns is not something I've seen before, so I'll need more information to comment on the science there. However, as a model for an actionable approach to AI, I like the groundwork he's laying. I also think, though published before the explosion of LLMs, that he does a great job of articulating some of the limitations of the approaches that they're based on and are why I don't think they'll ever scale to intelligence.