2007 : WHAT ARE YOU OPTIMISTIC ABOUT?

[ print ]

Author, Machines Who Think, The Universal Machine, Bounded Rationality, This Could Be Important; Coauthor (with Edward Feigenbaum), The Fifth Generation
Writer; Author, Machines That Think

Understanding What Really Happens To Humans In Groups

At seventeen I saw that contemporary literature—I studied it then, and hoped eventually to be part of it—is an abyss of despair.  No surprise: it reflects the unspeakable circumstances of the 20th century. Even so, it's no good thing to be seventeen and without hope. Luckily, chance brought me together with some scientists, and I discovered that in science, optimism was, and is, abundant. 

Since then, I've spent much of my life trying to persuade my friends in the humanities that optimism on behalf of the human condition is a plausible point of view. It isn't the only point of view—the 20th century's horrors are a fact, and most of them happened through human agency.  But the full life can support several points of view, often simultaneously, and my personal inclination is toward optimism, however qualified it must be.

For a long time my optimism centered on computing in general, and what kinds of benefits it might bring us. Events have shown I entertained far too modest an optimism—I'm embarrassed to say that the impact of the Internet, in particular the World Wide Web, eluded me completely at first. A few years ago, I returned to artificial intelligence, which I'd written about early on, and then gone away from. Press narratives were uncritical about the field's death throes, and I expected to write an elegy. Instead, I found a revelation. Artificial intelligence is not only robustly healthy, building on its very significant gains since I first wrote about it, but the field's present ambitions burst with, well, vitality.

Lately I've been examining a new aspect of computing, the modeling of human behavior in groups, small and large, beginning from the bottom up, playing out dynamically, as only computer models allow. Years ago, in a casual dinner conversation with a social scientist, I wondered aloud if what prevented us from understanding what really happens to humans in groups is that we haven't found the code.  I meant to make a vague comparison between the genetic code and something hypothetical that encoded human behavior. Instead of laughing, she solemnly agreed. 

Such a code is not yet on the horizon, but thanks to some marvelous new work by very gifted social scientists, its intimations are teasing us.   I'm optimistic that it will eventually be found. When it is, it will be a scientific triumph. It will open not just the future to our understanding, but also the past. It will be a human triumph.

What will it mean to have such a code? For one thing we can plan more intelligently. Want to wage a war? Call in the experts to run a few scenarios for you, laid out in bottom-up detail, humans and their interactions with each other and the terrain they're going to fight it out on. 

Watch silicon agents melt away to fight you another day; watch them reach out for help elsewhere. Once you watch the model run its course, maybe you don't want to fight that particular war after all. Want to predict the possible spread of a disease? Good, the silicon model will tell you how many will fall, and where you can intervene to pinch off contagion effectively, where it's a waste of effort. Want to figure out the ebb and flow of urban crime waves? And then how to prevent them? Play it out in silicon first. Why do humans cooperate, at least as much as they compete? Compare identical silicon societies, same people, same resources, but vary the amount of cooperation, the amount of competition.  Which one collapses? Which one survives?  Which one thrives? Where's the tipping point?

Perhaps as interesting, we'll be able to reach backward in time. How, really, did Mesopotamia become a desert when once it had supported a network of rich societies?  How much of that collapse was climate change, how much human folly? Build a model of early modern Europe and show what really caused the European Renaissance. Compute in detail how Great Britain came to rule the waves—and then didn't any more.

We assume we've solved some of these problems, though historians dispute one another ferociously, as do epidemiologists, as do economists, sometimes over details, sometimes over emphasis, sometimes over fundamental assumptions. Here comes a chance to nail it down, and these techniques offer us insights we couldn't get any other way. Finding the code I once thought was only hypothetical will revolutionize our view of who we are, how we got that way, and who we might become, the same way cracking the genetic code revolutionized biology.

But of course I'm chronically too modest in my hopes, so you can comfortably hope for more.