1. Neuroscience and the 'computational brain'

I am back again on this topic but this time, science (more specifically, neuroscience) has brought in fresh winds of insight.

In my first entry, I will be writing on how the brain is much more flexible and yet flawed as a computational toolkit.

The brain is a toolkit. As predicted by the toolkit analogy, larger neo-cortexes can be seen in animals in more complex social environments and larger hippocampus can be seen in animals with a need for more complex spatial memory. Some animals can literally ‘grow’ larger brains in seasons when they need more complex problem-solving tools and then ‘shed’ it later when it is not needed. Why? This is because brain maintenance is costly! Our huge brain takes up 25% of our daily calorie intake.

Learning is an important feature of a toolkit. Without this ability, the organism may not be as valid as changes occur. But learning is often only thought of in a linear fashion. Some behaviors are so important that learning is and should not be an issue as they are hard-wired into our brains. This then reduces the chances of us making costly mistakes that would jeopardise our and our specie’s well-being. Learning is not always productive and its costs may outweigh its benefits. In situations where the target of learning is not constant, then costs and benefits set in and learning is not important because probrability will win over learned (conditioned responses). Learning can also occur in creatures without brains. Slime mould without brains show complex problem-solving behavior that we often ascribe to a mind.

The brain has the ability to compute on an abstract level. Neurons in our brains fire upon recognition of concepts. The word ‘bird’ or sounds/pictures of ‘birds’ will cause the same neurons to fire.

Like processor units, brains continously compute an analysis of the outside world so that the body can react accordingly. But this computation process is not perfect. This is due to the limited abilities of any organism's sensory tools and so brains often take in semi-complete signs (as suggestions) from the environment and has to fill in the gaps (computes what is expected).