Of all the cool and creepy pieces of hyped-up software to have emerged since the Web went mainstream, the coolest and creepiest have got to be intelligent agents. And, according to the press releases jamming my E-mail server, they’re here: autonomous pieces of programming trained to race around cyberspace doing our bidding while we’re out watching football, making love or doing whatever else we’ve decided not to let our software do for us.
The idea is simple. Intelligent agents trek through cyberspace and fetch stuff. Unlike Web search engines that scan existing data bases, agents are free-roaming pieces of code that, once you’ve given them a command, roam out across the Internet itself, searching for information that satisfies your stated need.
An agent learns new things and alters its own commands based on its growing knowledge of its master. You provide the initial instructions, but once it leaves your control – like a child setting off for school – it’s on its own.
Conventional wisdom – as conjured by MIT’s Media Lab and a book by IBM programmers called Constructing Intelligent Agents with Java
– holds that agents will prove to be the Internet’s most compelling tools.
By allowing us to set our desires on autopilot, agent-assisted on-line commerce and entertainment could yield profits unimagined outside the pages of Wired or the promotional literature of Silicon Valley’s venture-capital firms.
In this case, however, I suspect that such conventional wisdom may perhaps give way to wisdom that’s a lot older and even more fundamentally conventional.
Back in 1967, Marshall McLuhan, the father of all media theorists, explained to a bewildered Barbara Walters that the discomfort associated with electronic media stems from the fact that they transport the people who use them.
“On the telephone,” he said, it is you
who gets sent, not the message. That’s why the medium is the message. It’s because it sends you, and not just what you’re saying.”
With intelligent agents, however, that rule appears to have been broken, or at least bent. Are we being sent somewhere, or not? In this new context, McLuhan’s sentiment harks back to a North American-native belief that photographing a man steals his soul – that somehow some essence is lost, or at least transferred forever beyond its bearer’s control.
Similarly, the age of electronic media steadily facilitates humanity’s deepening immersion into inorganic cyberspace.
I’ve always argued that each new device or network through which we express our essence, however electronic, serves as an extension of our will, even our spirit. But agents may extend us a bit further than we’ll find we like.
If the phone call or visit to a Web page sends us somewhere, then what does an autonomous agent do? Is it an extension of who we are, or some sort of offspring representing our interests?
When we’re on the phone or in a chat room, we are in direct interaction with the person on the other end. When we send an agent to do our bidding, we are at least one and perhaps many steps removed from the effects of our actions.
Something is transported beyond our sight and command. As McLuhan said, it’s a piece of us
out there.
Agents’ very power derives from their ability to alter their own programming based on increasing familiarity with our desires. Even the most primitive first-generation agents amply illustrate this potentially unsettling principle. One film-finder program asks you to rate a series of movies. Thereafter your agent will spit out only the titles of movies it thinks you’ll like. Over time, as it keeps track of more films you rate, the software learns your tastes so well it begins to anticipate them better then you’d be able to yourself. Your photo has been taken.
This software’s potential is, frankly, staggering. An academic agent will do months’ worth of research in a matter of minutes. A newsgathering agent will provide a daily, hourly or more frequent distillation of information precisely tailored to your needs.
A shopping agent will get you the best price for a given product to be found anywhere in cyberspace.
Their promoters, as best as I can glean from their recent flurry of hype, expect agents will someday rule the on-line world. It would certainly be a boon to business, but where does it leave us humans?
Those of us who have fallen in love with this new frontier understand that it’s a give-and-take relationship. The more of ourselves we give to cyberspace, the more human it will become for us.
But anyone who has sat before a glowing monitor long enough has experienced, at least momentarily, the same instinctive revulsion that draws the native back from the camera: the fear of losing a piece of oneself to the darkness on the Other Side. It seems to me that intelligent agents only exacerbate this perception. They don’t truly extend spirit; they suck it dry.
The engines of capitalism, exerting themselves through technology, tend to pull us in whatever behavioural direction offers the fattest bottom line. But we’re also spiritual mammals caught in the thrall of ancestral memory and the desire to move toward greater levels of awareness and connection to one another.
I suspect that, for many of us, confronted by software whose raison d’etre
is to absorb and embody our mental essence while leaving us behind, the cautious tribesman will win out over the rapacious venture capitalist.
After all, as history’s first modernist said, what shall it profit a man to gain the world if he should lose his soul?
Douglas Rushkoff is a New York-based social theorist and the author of several books, including Media Virus! Hidden Agendas in Popular Culture and Cyberia: Life in the Trenches of Hyperspace, a bestselling portrait of 1990s cyberculture. His first novel, Ecstasy Club, was published by HarperCollins last summer.