If we acknowledge that humans have limits, and build technology around human capabilities, then we've got a chance

Click to follow
The Independent Culture
Simple things can get complicated very quickly. We in hi-tech are used to complicated things, but maybe it's not just us. Recent readings, and experience, helped me to appreciate this.

It's Sunday morning, a fine and easy day stretches ahead - all I have to do is finish the column you're reading, on the topic of complexity. Piece of cake!

Then my wife learns she has to be in New York on Monday, leaving me with the shopping, dishes and laundry while she scrambles to pack and make the plane.

My step-son calls and wants to take me up on the offer to teach him HTML.

My father-in-law calls and asks me to help him trim a tree.

Our sewer gets blocked.

It begins to rain hard.

Suddenly, Sunday's shot and the column's going to have to wedge itself in among market trips, laundry folding, code lessons, plumbers, tree-limb hacking and a soggy, frantic me.

Apparently, if you believe people like Cambridge's Stephen Hawking and Princeton's Freeman Dyson, the whole universe works this way. Everything, from the lowest clod of earth to the grandest galaxy, is comprised of just a handful of particles and a couple of forces that bind them together. Simplicity itself, according to Hawking's lecture "Life in the Universe" - available on a richly illustrated CD.

A couple of primordial particles bind together to make a simple atom, which, by-and-by, clicks into an elementary molecule. One thing leads to another, and before you know it, you've got the Spice Girls, 64-bit microprocessors, operating systems, nuclear non-proliferation treaties and a whole planet full of complicated things.

What, you may wonder, was the Creator thinking? Would you create the firmament in seven days and nights if you thought the likes of the Spice Girls and Windows 95 were the inevitable outcome?

Well, I won't presume to speak for the Creator - she might not like that. But complexity, and its oft-unintended consequences, is definitely worth a look for people in the technology biz. After all, our lives start out complicated, not to mention where they wind up.

Freeman Dyson, the mathematician and writer, has a way of making difficult things easy to understand. In this month's Wired magazine, he explains the n-squared law by pointing out that the value of a network, like the Londond Underground, goes up by the square of the number of routes. The more connections and lines, the easier it is to get around.

But he also says that as n gets very large, it doesn't really get much easier to get around. Tube passengers have to deal with other limitations - like the need to get to work on time and the number of hours in a day - that keep them from benefiting past a certain point, no matter how large n, and the system, become.

London Underground patrons often wish the powers that be would add trains and lines to ease their commutes. But, if you've navigated Tokyo's mammoth and ever-growing subway system, you would know it's probably hopeless - no matter how many lines and trains are added, it's still a crush during rush hour.

Math and common sense both seem to say the problem can be solved, but, in practice, it just ain't so. Indeed, sometimes the best-intended actions have precisely the opposite result.

This past year, two of America's largest railroads merged. The theory that a single system, trimmed of redundant lines and effort, would be far more efficient was quickly disproved. In a space of months, rail traffic in the US became so bad that the federal government was called to step in and fix the gridlock.

It was clear that massive jams at a certain Midwestern switching yard were the culprit, but no one could come up with a fix. Any schedule change at that yard had a domino effect on hundreds of other lines - small changes here had large and unpredictable effects elsewhere. The cures were always worse than the disease.

The system had become too complex for any human, or group of humans, easily to figure out and fix. Moving a locomotive and a string of freight cars from point A to point B seems like a simple enough a concept, but moving thousands of trains efficiently over shared tracks and junctions nationally turned out to be another matter. Even a hundred years' experience running large rail networks couldn't keep the system from plunging into chaos.

The Internet's topology make the US rail system look simple by comparison. Yet, it almost always works. So, what's the deal here?

The deal is that the Internet's architecture relies on a relatively simple and humanly comprehensible concept and set of rules. Unlike trains that must stay fixed together, the Internet's "cars", called packets, are free to make their journey any which way they can. Our computers take care of putting them back together in the right order to make a meaningful e-mail message or Web page.

If we acknowledge that humans have limits, and build technology around human capabilities, then we've got a chance.

An innocent hydrogen atom, born in a moment of cosmic simplicity, turns into a rock star almost before you know it. That being the case, it's no surprise how quickly the affairs of billions of humans get out of hand.

The moral? Keep it simple (or, at least, human) stupid! At the very least, be prepared for Sundays not to work out the way you planned.

cg@gulker.com

Comments