This book, about 7 different aspects of the brain, is one of the shortest “pop-sci” book I’ve read in recent memory. Each chapter is about 10 minutes long, and the whole book could probably be read in under an hour in one sitting.
I’ve always read long winding tomes – looking at you, pop business psychology books – and wonder if it would have sufficed as an essay. In this case, the reverse is true. This book would have benefited a lot by citing sources and disagreements in context of the discussion. Arguments are often presented as is and while based on my existing knowledge nothing it’s saying is outrageous, I doubt they are far from being as certain as the presentation might suggest. That said, the meat and the nuance of the discussions are tucked behind in an appendix, which is surprisingly very readable for an appendix, but who reads appendixes? I would chalk it down to the intention of boiling things down and making them accessible especially for a science topic, but in this case I have to wonder if it’s a little too syrupy.
The summaries on websites do a pretty good job of laying out Lisa’s discussion so I’m not going to repeat it here. A few things stood out to me however: Her assertion (chapter 2) that the triune brain (i.e., the lizard layer, limbic system and neocortex) was not correct. The brain did not develop new layers that introduced functions and abilities but instead is a function of a differently evolved prehistoric brain structure. The parts that gave us evolutionary advantages grew denser and larger while unhelpful parts turned anemic (e.g., for flying). The theory makes sense and the implications intriguing to say the least (e.g., Annie Murphy Paul’s discussion about embodied thinking). I do, however, keep going back to Jeff Hawkins’s A Thousand Brains where he discusses his theories about how knowledge is created, retained, and revisited inside the neocortex. The latter (as I recall) discusses a theory of how neocortex brain cells work and how it can a model for computation and AI while this book argues that both “logic” and emotions are tightly intertwined with the needs of our body and survival. The discussions above do not necessary overlap and can simultaneously be true; there’s an interesting divergence here around whether intelligence can be harnessed or has to be fit to purpose (e.g., reproduction and survival).
My grasp of the sciences here is slippery and I am definitely treading into unfamiliar waters. A way to converge the two is to assume that our human brains have evolved into a certain state and “real” AI is simply a way of recreating both the circuitry and a subset of the environment that allows for the production to happen. While the verdict is probably still out on whether the our current implementation of neural networks are a step in the right direction when it comes to mimicking our circuitry or function, by all accounts of neuroscience it seems like a plausible step. The structures will have to change to account for newer findings, and the way we build the layers will change – less layers and more network and even more scarily so less hand tuned and more evolved – but it seems to me that functional, adaptable, AI is within the realm of possibility after a few more AI winters.
Anyway, back to the book. I like it. It’s a short read, very interesting especially for newer readers of neuroscience and would suggest to read the book in two halves: the chapters and the appendix.