Stealing entirely from Daniel Dennett. He argues that the thinking mind is actually a virtual computer simulated within the brain. He notes the way in which analysing thought led people like Turing to design their machines, and before them logicians to write their logical systems. We work well with computers because our thinking process is a computer he argues. I like this idea a lot.
For instance look how the thinking mind works. We think in steps, and there are connections between the steps. We start with data and premises and we think through using rules to get results. I want to go on a journey so I know I need some forms of transport. I gather data on what is available. I look at the possible train routes. Break the journey up into stages. Look at the possible times, durations and costs of each stage. Then I put it all back together in a few sequences of options and pick the best one. All in steps, all by processes and logic. Exactly like a machine. Animals are not so good at this, although many are surprising. Crows have the ability to think through sequentially tasks and solve puzzles with the same skill as young human children.
Yet we know most tasks we do as humans are not like this. I feel hungry. I walk to the shops. Stop at the kerb side, look for cars, cross, recognise a friend, have a chat, see the shop name, open the door, identify and pick some food I want, pay for it. These are all tasks that are extremely hard for computers. But they are also tasks that animals do equally well. These tasks we find are well replicated by neural networks and machine learning. It is not based so much on steps and combination rules (in computers it is obviously cos this is how they work) but instead works by gradually evolving arrays to collectively represent the processing. In a CPU data and instruction are loaded and the the chip performs whatever operation is hardcoded by that instruction and the machine works through the program instruction by instruction like reading a book. A neural network has its strength in distributing processing across a whole network. This is not to say parallel processing and lots of little CPUs, the problem itself has changed now into solving the most appropriate relationship between nodes.
Now I was wondering why we have the two systems. It seemed to me it works like this. We operate automatically using our neural networks for most of the time. Our neural networks learn quickly what our friend looks like, we learn quickly how to ride a bicycle. These things we just trust our brain to do and we let it get on with it. But sometimes we need to be sure that the neural networks are working correctly. Their power is also their weakness. They evolve to solutions but it means that we never quite know what our neural network will do and whether its training is complete yet. I have cycled for 35 years and have had my cycling neural networks very finely trained. I don't have tumours of brain damage that I know of so I trust that these networks work well and I cycle automatically almost all the time.
But suppose I was asked to cycle over a path with cliffs on both sides. Now I am cautious. I cannot afford to have the networks let me down. So I start to think rationally. I break the task into steps. I look at how confident I am of each step. Where are the trouble areas. Which bit are most dangerous. Which bits am I confident of. What unexpected hazards are there. I create a plan of the task in minute detail examining each part and rehearsing it. I become a project manager of the task. Even while "on the day" it is the neural networks who will be applied to each stage to get actually the job done.
Now this process of rehearsing is actually trying my networks out. I am simulating the experiences and letting the networks run through what they are going to face. But critically this process is in steps. It is not a global training program, but instead the step by step thinking of a computer.
I wonder then whether we evolved this rational computer as a way of testing and managing our neural networks. It also means that like Dennett says this is critical to the process of self-consciousness. It is all bound up in the process of self management, self development and self-regulation. That voice in our head is really just an onboard computer stepping through programs to check on our system state and performance.
Now quick check even on that. In this world of project management and performance targets have I just seen this in the brain because that is what I am taught to think by this culture. Or is that the we have evolved a world that reflects and reveals our own behaviour.
Also the addendum that it is always interesting to me that we are able to watch all these processes even the process of thinking itself. What we can see we are not. Ovid says in the story of Pygmalion in his Metamorphosis that to see clearly the eyes need some distance. Implicit to Knowing is that we are separate. The essence of Heideggar's Dasein is the "over thereness" but you could argue that the "over thereness" is a product of knowledge itself. On an aside Heideggar's says of Dasein that it has an intrinsic "mineness" too. In wonder if this too is not intrinsic but rather the social imposition of the idea of ownership. If Heideggar had ever experience Jhana he would know that absorption into the object dissolves both the "over thereness" and the "mineness." The true nature of mind is joyful and not dualistic. So we can only know about what goes inside our head because we are not what goes on inside our head! The fact that we can see our thoughts and emotions means that we are separate from these! So when we talk about brains, thoughts and all these processes we also know that this is not us. That is quite hard for modern science to grasp because it is overwhelmed by a materialistic ontology where things exist. And so the self exists. Science has always been troubled by the location of the self. It was once thought to be in the heart, but today its thought to be in the brain. Both are wrong. Self is nowhere because "where" is created by the self. Like asking the people in a painting where they are. Do they say "by a river" or do they say "on the wall." Where is not such a fundamental material question. Its relative and based upon context. There is no absolute "real" where. So this blog was simply looking at the parts of the brain that we experience in daily life and how they relate. It was not any speculation on what we ourselves are.
No comments:
Post a Comment