Betting lives on computers

A popular plot device in science fiction has technology that’s built to defend us turn on us instead. For some of us, the implausibility of that concept makes the story less entertaining. A computer that’s programmed for defense can’t “change its mind” and behave in a manner contrary to its programming. It won’t rewrite its code because of a glitch. And what magical power source do these computers find after they’ve gone evil on us? Computers are stupid hunks of tangled wire capable only of doing exactly what they’re told, and when we unplug them, they’re not just dumb. They’re dead.

Savvy movie audiences and novel readers choke on these logical gnats, but they let pass the camel-size implausibility of the plot line: Computers will never do squat to protect us. They can’t strike or retaliate; they can’t spy; they can’t detect or police; they can’t advise or predict or diagnose. Even the most cynical among us sometimes forget that computers can’t analyze. We’ve twisted the meaning of that word to remove the requirement of reason.

This would be a windy social commentary but for the incalculable cost of the technology-centered delusions indulged by the majority and fed to them by those who know better. We draw a lot of comfort from the belief that when bad things happen, we keep them from happening again by reprogramming the computers and the people that failed to spot trouble the first time.

Sept. 11, 2001, reminded us that our vast network of satellites and radars can’t spot an airplane in broad daylight if its radio is switched off. Current headlines tell us that the best trained, best equipped, and most dedicated high-tech military can’t pick off one guy with a mortar or a backpack full of land mines. When CNN reports of laser-guided missiles wandering off target and teenagers cracking into billion-dollar computer networks, the audience rolls its collective eyes.

Many day traders lost everything by trusting computers with many times their net worth, nursing the fantasy that they were in control by deciding to hold when the computer advised them to buy. Same gnat, same camel, but here the camel stepped all over the economy. Was the recession caused by stupid people making big decisions or by smart people who foolishly believed that computers could do half, or a quarter, or even 1 percent of their strategizing for them?

I’m not being fatalistic; the simple cure is for IT to say “computers can’t do that.” Stock market safeguards, missile shields, the elimination of spam, effective digital content protection, automated airport screening, proactive physical security systems, Santa Claus, and weather prediction have one trait in common: They’re all riding camels.

But reality isn’t that horrible; technology will make us safer if we trust it to handle the things it can do on its own. In my view, the only thing computers could do perfectly, better than any two humans ever could, is communicate. To get 8 bits from one place to another, a computer can leap from fiber to copper, from digital to analog, from short-haul wireless to cellular, from shortwave Morse code to smoke signals to pony express. Computers won’t keep the stock market from crashing, but they can take “the lines are down” or “she’s not answering” out of the equation when people need to connect.