Chris Gulker column

The New Year is a transition point, a good opportunity to review priorities, evaluate successes and failures, and, of course, make resolutions. My plan was to write this column about hi-tech New Year's resolutions, but that was before I got on a plane for Florida.

There's an American blues tune with the refrain, "if it wasn't for bad luck, I wouldn't have any luck at all". These words perfectly describe my relationship with the city of Miami.

This New Year finds me arriving for (what else?) a technology conference in Miami, a place where I have been robbed, caught in a riot and (honest) struck by lightning. And that was just during my first three visits.

So it was no great shock when, shortly after I got in from a late flight, my hotel's fire alarms went off, sending me and a few hundred others into the hot and muggy gloom of Miami Beach's Collins Avenue. Oddly enough, the experience taught me a little something about technology in the process .

My New Year's resolution is to look hard at how technology could genuinely improve my life. I want to weed out the buzz-words and the hype, and concentrate on things that really make a difference to the Gulker bottom line.

My other resolution is to try more new technologies, even (especially) the ones I'm sceptical of. By "try", I mean, "really live with them", at least for a while.

Gary Starkwether, who invented the laser printer when he was at Xerox's Palo Alto Research Center, tries things like this all the time. He once scanned in every piece of printed information he received and lived only by what he learned by reading from his computer. Gary gives a fascinating presentation based on this experience, in which he notes that, in a sense, you and I are richer than ancient Egypt's Pharaohs. It cost Pharaoh $400 to publish a papyrus document, where you and I can create e-mail for fractions of a cent.

I went through a similar, if less consuming, phase a couple of years ago, doing things like reading whole novels in electronic form on a laptop computer (not as bad as it sounds, but clearly not a compelling experience).

Learn from gritty experience, that's my 1998 goal.

Which gets me back to my hotel in Miami: the fire alarms are blasting weird electronic whoops, strobe lights are flashing and a recorded voice begins to blare "proceed to the staircases, do not use the elevators" at ear-splitting volume.

Resigned to my current Miami fate, and stepping into the hotel hallway, I see concerned-looking guests emerging from their rooms. A woman, clutching a small child, steps into the hallway and asks a sweating, extremely worried- looking security guard where to go. "Quick, the elevator," he says. They rush to it, and, true to the recording, it's not working.

The "do not use the elevators" recording comes back on, and the mother, flashing an evil look at the security guard, turns and rushes for the stairs, grim-faced and determined, followed by the security guard. Having watched the film Titanic a few days ago, I'm content to let women, children and clueless security guards go first, and follow.

The "fire" turned out to be an ice machine that went supercritical, shorted out and triggered a smoke detector. The worst aspect of the evening, other than the unplanned constitutional, may in fact have been that the beach- front dive that I ducked into while the Miami fire department cleared the smoke left the sauteed onions off my dish of chicken plantain with black beans and rice.

Getting back to the security guard, here was an instance where technology may, in fact, have worked well. The automation that stopped the elevators from working when the alarms went off kept three people from making a potentially dangerous mistake.

The alarm system automation overcame the shortcomings - whether it was training or judgment - of the security guard's decision-making process. Elevator shafts can presumably turn into chimneys during a fire, the elevators can stall mid-floor, and, anyway, experts warn us to steer clear of them.

Harking to my resolution-mandated scepticism, however, makes me wonder whether there aren't situations where the automation would misdirect people. What if there were fires at the base of the hotel's stairwells? Might not the elevator be the best escape in that admittedly unlikely situation? And, in the event of an ice-machine short circuit, wouldn't it be better to skip the whole exercise?

This, of course, goes to a broader issue. Can we trust software to make decisions that people would normally make? Computers can, after all, perfectly record and replay algorithms - such as what to do in case of fire - that we humans often have problems with.

Many science fiction movies have been made on this subject. Indeed, the film Titanic, not to mention the event of that ship's loss, is an object lesson in the foolish assumptions humans sometimes make about technology.

Few nowadays would tempt the gods so far as to say they had built an unsinkable ship, yet many computer and software companies routinely make marketing claims that are only slightly less egregious. "User friendly" is one such claim.

The human brain can still process vastly more variables than any computer, regardless of its speed. IBM's Big Blue can perhaps take a few games of chess from Gary Kasparov, but in the event of conflagration, my money is on Kasparov getting down the stairs before Big Blue does.

cg@gulker.com

ww2.gulker.com/news