Bill Joy, one of the founders of Sun Microsystems, recently wrote in Wired magazine that he's concerned about the potential for powerful 21st century technologies to make the human race an endangered species. Joy can hardly be considered a Luddite; he has a reputation as a thoughtful and reasonable technologist, and both he and Sun have contributed a great deal to the success of computer networks and the Internet. So why is he concerned?

Joy's concern is for the future - some 30 years hence, by most estimates - when advances in genetic engineering, robotics and nanotechnology will provide the ability for self-replicating programs to run amok. He's concerned with the fragility of software, and senses that scientists and technologists don't "understand the consequences of (their) inventions while in the rapture of discovery and innovation;" he also feels that "we tend to overestimate our design capabilities."

I appreciate Joy's concerns, and also the fact that someone of his stature has taken a hard look at technology with something other than marketspeak in mind. After all, it's not like technology has a faultless track record. The nuclear power industry, for example, has spent untold billions of dollars building fantastically complex power generating plants that use the heat generated by nuclear fission to boil water - kind of like using cannonballs to ring doorbells or chainsaws to cut butter. These plants also generate material for use in nuclear weapons, as well as highly poisonous waste that will remain toxic for thousands of years and for which we have no comprehensive storage plan. I doubt if there's ever been a more convoluted, technology-run-amok method for boiling water; Rube Goldberg would be envious.

But let's get back to the future, as they say. Joy's concern is that technology is advancing rapidly without a plan, with no control, and no brakes. Joy suggests that we limit the pursuit of certain kinds of knowledge, but I question the viability of such an approach in an age of religious fanaticism and political terrorism. Either we need some kind of bureaucracy of checks and balances to better control technology, or we need to better comprehend the issues of complexity in technology and resultant interdependencies.

We tend to underestimate the interdependencies of complex systems, which limits our ability to control them or accurately predict what they will do. According to Webster, the word complex "suggests the unavoidable result of a necessary combining and does not imply a fault or failure," which I think spells out fairly well what Joy is getting at. I don't know about you, but that word "unavoidable" makes me a little uneasy.

The engineering mindset seems to have something of a predisposition for complexity. I suggest that this predisposition is an extension of our fascination with logic, and that the current state of computer software design is a manifestation of our tendency to go a bit overboard in this regard. Logic has its limits, and it doesn't always reflect the real world or provide the best answers. Could logic alone lead us to an understanding of quantum mechanics, or the paintings of Picasso, or the economics of professional sports in America today? Or, for that matter, clicking the Start button to shut down a computer?

Also, Western society seems to have evolved with the misguided notion that the world and the roles we play in it are all separate and discrete, as if we exist in vacuums and the professions and systems we belong to have minimal interaction with each other. Engineering, economics, computer technology, politics and law - to name just a few - are but subsystems of the world we inhabit, and sometimes it seems that blinders are required to excel in many of them. When our systems fail us, it's often because we perceive them to be separate from each other - and the world doesn't work that way.

Short of developing and enforcing an Occam's Razor-like simplicity standard for scientists and engineers, or becoming homicidal terrorists like the Unabomber, what can we do? We can't really ban complexity. And we are capable of building certain complex systems that provide great benefits to mankind, such as jet planes and computer networks. We'll probably always need to deal with unanticipated problems with these systems, such as mechanical failures and bozos with backhoes, but I think being truly vigilant about knowledge-enabled mass destruction will require us to rethink the ways in which we interact with the world around us. I also think a dose of humility wouldn't hurt; we, too, are but a subsystem of the world we live in, and there are things that are beyond our control.

Who would have thought the liberation of knowledge could pose such formidable challenges? We seem to have an amazing propensity for tying fantastically complex - there's that word again - knots in our own shoelaces, so that it becomes a challenge to walk down the hall to the bathroom. Sometimes one can't help but wonder why we don't all hop around like kangaroos - but then I think that's maybe what we are doing.

© Copyright 2000 by Stephen Wacker. Contact Stephen Wacker regarding use of this copyrighted material.

Stephen Wacker writes about technology, culture and society. His career as an information technology professional has focused primarily on communications and the Internet. Mr. Wacker also writes about contemporary popular music and is an accomplished songwriter and guitarist.