Friday, March 24, 2006

Artificial Intelligence / Natural Stupidity

"Artificial Intelligence." Sept. 2002

This amusing term is rife with sardonic possibilities, of course, as it presupposes that intelligence, as we humans pretend to possess it, is a 'natural' quality, akin to the thumb or color vision, rather than the truly new and different thing that is now transforming the world. It is also a tacit admission of the flawed and incomplete nature of our intelligence. We have used our intelligence to leverage our abilities by an enormous factor; over and over we have taken a simple, 'natural' ability, for example, the ability to throw a stone and knock out an enemy's eye, and enhanced it, first by putting the stone in a sling, then in a catapult, and finally into an ICBM that delivers the very fires of hell itself. We have also used our intelligence to leverage intelligence itself; first we invented language, then writing, and finally the Web, and a rudimentary Hive Brain slowly begins to assemble itself out of a myriad of morons. And yet it is obvious to individuals of average intelligence that the aggregate intelligence level is abysmally low, and is inadequate to foresee and prevent general disasters identical to those so richly studding our history, not to mention the regular collapse of whole civilizations. This could be called a 'natural' cycle for our species, like forest fires and lemming migrations, but currently the stakes are higher than ever before, with world civilization becoming ever more unified, but no better planned or regulated. We're gambling with the whole damned ball of wax, boys and girls. And, frankly, the odds are heavily stacked against us.

Hence the attraction of the concept of artificial intelligence: computers, at least in principle, never 'forget', never miscalculate, never lose their bills somewhere on their desk, and deal with information a great deal faster than we do. Our brain, already our strongest feature, becomes greatly enhanced in its cruder abilities; this enables us to solve immediate, relatively simple problems at the expense of more subtle, long-term problems. We're riding this gigantic new digital horse and getting to that opaque and deadly future far faster and more efficiently than before, but we still don't know where the hell we're going; we suspect that a fatal cliff is just ahead in the mist. If only Mr. Ed could really think, and tell us which way to go! Grandma's house is somewhere through these woods, we fondly imagine: a world of natural beauty, plentiful food, clean water, digital televisions crammed with topnotch quality entertainment, and grandmotherly comfort for everybody. Help us find the way, Micro-Ed - bring us safely home...

I'm sorry, kids. No matter how sophisticated computer calculation becomes, it will never possess courage, common sense, and a global perspective, and thus help us find the will to make all those fundamental changes that we just really, really don't want to make. All it can give us is information. We are still free to ignore, corrupt, twist and misuse the information. The next logical step would be to create an A.I. Executive, or, really, an E-Dictator, and obey its every command. Oops - then we would fight endlessly over the manipulation of the input so as to tweak the output. We're following a faint trail of breadcrumb ideas through the black and dismal forest of human ignorance, and the crumbs are being stolen by ants before we can find the way.

And so on and so forth. But imagine the scheme somehow succeeding, and the world becoming One Machine, and we all happy and stable cogs in it. Even Armageddon might be better than that, I feel.

No comments: