Genes v. knowledge

At the recommendation of my friend and debate partner I read and have finished the book A Thousand Brains, a New Theory of Intelligence by Jeff Hawkins. After a short career in computers (remember the Palm device?) he turned to his first love of trying to understand how the human brain works. This book is written with the average person in mind. The author wrote in a style that is easy to understand. At the end he lists several papers for further reading, though warns papers about neuroscience can be difficult to read. The first section of the book delves into this new theory. Hawkins says the brain has two major parts, the old brain (I think that others have called the lizard brain) and the new brain. The old brain actually controls muscles and is where the basics of life (including sex drive) are carried out. The new brain, the neocortex, the wrinkly part we think of as the brain, is where intelligent thinking happens. If the neocortex was flattened out it would be the size of a napkin and not much thicker. It is made up of about 150,000 cortical columns about the size of a grain of rice. Each are essentially separate and nearly identical bits that do their own intelligence thing. Yeah, some are connected to the nerves from the eyes, others to nerves from the ears, from the skin to feel, from the nose and from the tongue (all routed through the old brain). So those columns do different kinds of intelligence things, but are still structured nearly the same. These columns do their thing through constantly predicting what will happen next, comparing that to what is actually perceived, and doing it all through reference frames. There is constant learning. Then the columns communicate their conclusions to other columns that put together an entire model of the world around us. For any more explanation than that please read the book. Hawkins does a good job of explaining this theory. The second part of the book is about machine intelligence. Hawkins says that’s not the same as AI or artificial intelligence as we understand it now. The current AI systems can learn how to do one thing and once they’ve learned that they can’t learn to do another thing. True machine intelligence means a machine can continuously learn to do an increasing number of things and then think about what it has learned, much like a human can. Hawkins then discusses why he believes machines, though they may become more intelligent than humans, cannot turn on us and enslave or kill us off. The third and last section of the book discusses the future of human intelligence and parts of this were the most interesting to me. The brain doesn’t actually see, hear, feel, taste, or smell. It receives signals from nerves. From those signals it constructs a model of the light that enters our eyes, the vibrations that hit our ears, and so forth. A big puzzle, even with this theory of the brain, is how do some nerve signals get interpreted as red, others as the rustle of leaves, as a smooth surface, as sweetness, as the fragrance of a rose, or as pain? They’re all nerve signals. What we experience is not the world, but our brain’s simulation of the world. And that means the simulation can be sometimes wrong. Since the brain is constantly learning – constantly comparing its model with input it receives – the simulation is usually corrected. And sometimes it isn’t. Hawkins explains how false beliefs persist. The earth as a flat surface was a useful model for a very long time in human history. There was little to contradict it. When there was something to contradict it most people didn’t experience it directly. We had to take someone else’s word for it. And some people refuse to do that. If they don’t experience it themselves they don’t believe it. I’ve shared a few social media memes, small pictures with catchy sayings that make a specific point. The original definition of a meme is similar to a gene, an idea that replicates through a population. Yeah, a social media meme is a subset of the original definition, an idea that replicates through the population through social media. An example of a meme is “all children should be educated.” Another meme, with a few parts, is “Everything in this book is true. Ignore evidence that contradicts it. Help others who believe this book is true. Banish or kill those who don’t believe.” This meme, with its extra sentences, has spread far. Now add a few more sentences. “Women should have as many children as possible. Don’t allow children to encounter contradictions to the book.” And the memes of the book spread as the genes of the people are spread. The way out of this situation, the way to correct our mental model is to seek out information that supports and information that contradicts our model. It is only through the contradictions that our mental model revised. Seeking to disprove our mental model is the scientific method. I add: One way to do that is to read books from a wide variety of types of authors. Another is to associate with a wide variety of types of people. Throughout time the old brain has one evolutionary goal, to pass on genes. And such behaviors as fighting over territory, dominating or eliminating mating rivals, and rape further that goal. Our new brain might find such things as abhorrent, but the intelligence in our new brain can be, and frequently is, overridden or used in service of the old brain. Until the last eighty years or so, when an old brain goes on a rampage a lot of people could be (and were) killed. See the world wars. But humanity lived on. In those last eighty years humanity has developed a couple ways that could make the whole species go extinct. These are nuclear war and climate change. Climate change would be less of a problem if the world population was smaller – fewer cars on the road, less pollution from farm animals, and less less rainforest burned for cropland. So why has the world population gone from three billion in 1960 to nearly eight billion today? Because the old brain is still in charge. Having as many children as possible increases the chance of some surviving to pass along the genes. The children may suffer (perhaps from starvation or viruses) and the parents grieve. But genes don’t care about that. There are ways for the new brain to take control of the old brain. In the case of overpopulation it means giving women control of their own fertility. The old brain can have as much sex as it wants while the new brain can prevent having children. Yeah, that is up against a lot of old brain thinking in a bunch of old men doing all they can to make sure women don’t have control of their fertility – and also preventing these women from getting sufficient pay and affordable child care. And preventing moving those old men out of positions of power. All that old brain thinking and all those false beliefs are a risk to humanity. Hawkins has a chapter on merging brains with computers. One idea is uploading a person, their memories and way of thinking, into a computer to provide infinite life. Another is to install directly into a brain connections to a computer for instant information access and presumably higher intelligence. I’ve read many science fiction stories about both ideas. Hawkins says both have advantages, but both are extremely difficult, both probably won’t produce the intended result, and neither helps with the major risks to humanity. I’ll let you read the details. Hawkins spends a chapter on a topic that has intrigued me. If humanity faces the possibility of extinction how do we preserve our vast knowledge? We’re a species that has learned the secrets from subatomic particles, atoms, molecules, and DNA to planets, solar systems, and galaxies, and everything in between. We want to at least be able to say we were here, we learned, we know. If humanity becomes extinct, a new intelligent species might arise in a few million years. How might we preserve our knowledge for them. How might we do that in a way that survives general erosion and plate tectonics? I don’t have an answer and wish it was considered by a scientist (who probably has) or a science fiction writer (who might tell me about it). Hawkins has a solution – put all the knowledge into a satellite that orbits the sun (see La Grange points). That doesn’t satisfy me because to get it the species would have to develop spaceflight – and one of the things we could share is how to do spaceflight. We would also want to share what causes war and why developing a fossil fuel industry is a bad idea, things likely to happen before the species gets to spaceflight. To prevent total extinction of humanity we could colonize other worlds. Mars is the logical next one. But what would happen (we have a pretty good idea what would happen) if, after a Mars colony is firmly established (and doing so is extremely difficult), the environment on earth becomes dramatically more hostile to life and billions of people want to go to Mars? It could be disastrous for the colony. That prompts Hawkins to propose we should use our ability to edit genes to make the old brain less in control and the new brain more in control so that the goal would be not about passing on genes but about preserving and passing on knowledge.