Saturday, 7 July 2012

Dennett's views on computers and the brain

Daniel Dennett

I have read Dennett's article A Perfect and Beautiful Machine with interest and I feel his whole argument is based on a false premise. Any mathematician knows that there are usually ways of building alternative mathematical models – and you need to select a model appropriate to the task in hand. Because Turing's model explains the success of the stored program computer Dennett makes the assumption that it must be the appropriate model for understanding the working of the brain. But let us look at the model Turing started with, as described by Dennett
He took human computers as his model. There they sat at their desks, doing one simple step after another, checking their work, writing down the intermediate results instead of relying on their memories, consulting their recipes as often as they needed, turning what at first might appear a daunting task into a routine they could almost do in their sleep.”
Come on. Lets be realistic. Several hundred generations ago we were all hunter gatherers and what is being describe has absolutely nothing to do with the way the brain evolved up till then, and there is no evidence that it has changed significantly in the last 10,000 years or so. A room full of human computers is a very artificial situation as only some people (not everyone would be good at it) could carry out out a highly repetitive and deadly boring task. To do it they have to behave like a zombie (because they could almost do it in their sleep) to a recipe (because they cannot be trusted to use their own initiative, so a recipe “creator” had to tell them what to do). The task is beyond their brains normal capacity (they write down information because they cannot trust their memories). In addition the overall effect of their labour is to carry out a well-defined and highly formal task which is related to some very narrow aspect of the culture of their society. Their activity is totally directed – while the whole point of evolution is that it is blind to the direction it is going. Put in this way it is hard to see why anyone should think that Turing's “human computer model” has anything to do with the environment in which the human brain evolved.

If we are really honest we must admit that the task that faced Turing was the design of systems which could quickly and efficiently carry out highly repetitive computational tasks, and he was employed to do this because humans could not carry out such tasks in a fast and reliable manner, even when given extensive training and very precise instructions. The Turing model is actually an extremely useful and powerful model of what humans cannot naturally do!!!  

No comments:

Post a Comment