January 12, 2007

Complexity vs Efficiency

Ever wondered how much of our effort goes in general to increase the complexity of the system we are working in. People working day-in and day-out to achieve more and more efficiency in their work by increasing complexity of the same. It's amazing that we have reached a point in line of advancement where our notion of development in terms of building efficient models of working system is paralleled with the rise in their complexity and density. Take a self-test - Is your work adding to the complexity of the system to derive efficiency? And no wonder your answer would most probably be in affirmation.

Definitely it's possible and that's why it's happening and that's why the world is moving. The strong covariance between complexity and efficiency, moving the wheel of our so-called development is a clear-cut 'Yes' in everybody's mind. Intelligent routers - at the cost of high intricacies, better and more feature rich softwares - with more lines of code, futuristic cars - with self-controls and the list is perhaps endless for me to mention, but the question remains how long will this schema of progress continue to prove it's worth for our ever growing demands from the respective systems. Isn't there a limit to the complexity we can achieve in a system. Come D-day and we might as well realize that we have crossed the rubicon and there's no way to turn around. That the complexity has reached a stage beyond comprehension. That there are systems, but we don't have the clue to understand them. Our ingenuities forcibly stuck because there's no place to implement our ideas. And slowly and gradually our own creations will surpass us and our ability to maintain them, eventually leading our racing vehicles of development into a pitfall. The only consolation I speculate, for someone could be a step towards reducing the entropy of the whole universe by building complex and intricate systems. But certainly that doesn't help much in the much bigger picture of infinite universe and the longer run towards eternity.

I know it all sounds far-fetched and outlandish. May be !, but we never know. We have effectively progressed in the past few decades ( am talking of scientific progress) more than what our predecessors had in all the time before.


So can we move to an all together different track, and start our quest for increasing efficiency by means of simplifying the system? That clearly is a tough call, coz had it been easier we wouldn't have walked the other way round in the first place. Producing efficient systems which are simple is tough and increasing efficiency by simplification would be like dying. And that's why we would never get to see the next release of Windows backward compatible with the existing hardware as against its requirement of gigs of RAM and multiple processors. I guess somewhere during our evolution we have perhaps ingrained the idea of value-addition in terms of complicating things. The idea however has served well and continues to. We have had the most complex of theorems applied to the most complex of systems in order to achieve fractions of indispensable efficiency figures. But then that is how we have reached this very point where I can take a critical view point towards this hand-in-hand relationship between complexity and efficiency. And if not for this we wouldn't bother to look beyond the horizon and ask this question - Can we simplify things to achieve the desired levels of efficiency. Can we build systems which are simple and less intimidating, and yet can be perfect replacement of their counterparts?

No comments: