Title: A Thousand Brains: A New Theory of Intelligence
Author: Jeff Hawkins
Published: 2021
Genre: Non-fiction, Biology, AI
As the foreword says – this is a book about how the brain works. Although it is not just about the working of the human brain. The book ends with questions on intelligence and how the collective human knowledge can be preserved for the future.
There are some fundamental questions
What are we?How did we get here?What is our destiny?What makes us intelligent?How did our species become intelligent?What is the destiny of intelligence and knowledge?
The book is a step in answering these questions. The answers are not perfect. As the author, Dr. Hawkins says, we are still exploring these answers.
Dr. Hawkins presents his theory of how the brain works. There is an ongoing battle between the two parts of the brain – the reptilian (old) and the mammalian (new) which have grown on top of the older ones. Older parts control primitive behaviour while newer ones create more sophisticated ones.
The world we perceive is a simulation of the real world. All thoughts and perceptions are relative to the brain's model of the world, not the physical world outside the skull. The very act of thinking is a form of movement. The brain's neocortex creates a predictive model of the world based on information acquired throughout. Any deviation from this prediction is immediately noticed. The model is based on predictions, perceptions and actions. The neocortex learns a rich and detailed model of the world which it uses to constantly predict what it's next sensory input will be.
All our thoughts and actions are results of activity of neurons. Knowledge is distributed in the brain and not stored at any particular neuron. Brain has 150,000 cortical columns and each column is a learning machine.
The book then turns to the question of Artificial Intelligence (AI). The author asks the basic question “Where is the I in AI?" Today's AI can only do the thing which it is trained to do, unlike humans who can perform multiple tasks. e.g. a chess playing computer is great at playing chess, but doesn’t know if chess is a game? The search is for creating Artificial General Intelligence - how to make a machine learn everyday tasks like a human. Intelligence can't be programmed into a software. It can be given a model which needs to be learned and takes time.
Will AI become a threat to its creators, i.e. humans. Life is based on a simple idea. Genes make as many copies of themselves as possible. And brains act as a facilitator for this. Hence, anything which can self-replicate is dangerous e.g. a virus. A meme is something that replicates and evolves much like a gene but through culture. Our biggest distinguishing features, language helps spread beliefs both true and false. It also helps in expansion of knowledge beyond what is directly observed.
Then we come to the section on Merging Brains and Machines. (seems to be the concept of the Amazon series Upload!) Uploading a brain into computer even if possible would create two beings which will have then distinct lives. Creating a direct connection with a computer seems a better and more feasible option. But why would humans do that. The answers being preservation of our collective knowledge, of intelligent life, signal to other intelligent lives in the universe.
These are the intriguing notes at which the book ends. The book is an amalgam of ideas – attempting to understand how the brain works, is there any artificial intelligence and if one can be created and finally to ideas for preserving human knowledge.
Overall, an interesting read into a subject in which humans have made lots of progress and yet we are still a long way from really knowing it. Certainly a subject which makes you think and use your brain a bit more!
Previously on BookMarks: Lies
No comments:
Post a Comment