Our brains the Next Gen Supercomputers

Moore's law an empirical observation stating that the (in 1965) number of transistors in a chip would double every 2 years. Basically, computing power available at the cheapest cost would go on increasing exponentially.

But there is a limit to this imposed by the emitted heat etc. So does it mean that there are no alternatives to increase computing power? New upcoming technologies like multiprocessor architectures and nano transistors are constantly pushing the limits of scalability. But all these efforts are like maintainence work on a vintage car!

Is there a better solution to our ever increasing computing demands, our need to cut on power and to miniaturise... Yes the answer could lie right on our top - 'our brains'. Yes, everyone agrees that our brains have immense raw computational power with its billions of neurons and heavily connected neural network. It arguably is the best computer around. But is there a way to use it? My answer is yes.. and how goes like this..

Imagine a simple computation like 2 times x where I say x is 5. Before the brain gets to the computation part what goes on would be something like semantical processing to understand the inputs and determine the steps needed to acheive the result. It would first associate 2 with being a numeric entity and then the tougher part would be to build a logical map of the unknown x; assigning it the properties of an abstract entity and then using the brain's contextual ability gained through experience and learning to define the abstractness into a "tangible" quantity or logical parameter. Once that is done 'simple' - but not so simple computation would be done for the result. The brain with its super- developed network of trained neurons would get into the act of multiplying the two. All this for a simple 2 * 5 which a computer could do billions of times a second! The why look on our brain as a supercomputer?

Simply because if the same process(exactly the same as the brain) be carried out in the computer would require groundbreaking techniques to be able to simulate the brain in this regard. Instead lets take the example of a ball thrown from one side of a field to the other and you running to catch it. Your eyes view the trajectory of the ball gauge its speed, calcualte the speed you require to run at to catch it, estimate when to open your hands and when to close them and volla the ball sticks in your hand. The same cannot even be imagined to be emulated by our machines. We are even today way backward in getting them close to doing something similar. Imagine getting the machine to solve the complex differential equations to calculate the trajectory, determine the running speed, time the extension of the hands and their closing and opening. A mathematician would be testimony to the fact that the equations involved are just too complex to compute manually or get the computer to do so. But the brain does it!

So the key finally lies in trying to simulate a given differential equation, a linear equation, a complex formula, a forecasting problem or anything for that matter into a real world scenario. Something like a ball to be catched. Given that the brain does this the best than anything around. We get our solution in the fastest and most efficient manner. Basically, we translate the problem from a pure mathematical equation to a real world situation like juggling balls, catching a ball, opening a door, walking, running, crossing a road just about any general physical activity involved. Each of these activities may entail a separate set of mathematics to be dealt with. Formalise and create the problem situation and feed the corresponding impulses into the brain. The brain treats it as just another event and does the needful. We have our output in no time since most of these actions are impulsive and involve lot of involuntary actions they are quick. Another machine similar to the translator called the inverter which basically does the opposite job that is to convert the neural impulses sent from the brain back to those mathematical equations which are nothing but solutions to our problem.

Although its a little far fetched the idea is certainly thought provoking and considering the advantages demands serious attention. Absolutely no dearth of resources, no power, no wastage, no bio hazards, immense power and scalability, its hard to think of any cons. Moreover one could use the brain of a lesser developed species such as even rats for example since every species knows how to move around, dodge, and fend its food. It's brain surely has all the required neural networks in place to tackle the best of what can be thrown at them!