Your computer is not alive

Originally published 2002 in Atomic: Maximum Power Computing.

Last modified 03-Dec-2011.

As user-interface designers and tech-support staff have discovered to their chagrin, people don't think about computers in the right way.

Ever since "it's a computer error" became an acceptable excuse for the nine hundred and ninety-nine years worth of interest you just got charged on your mortgage, or for your airline tickets being to Beirut instead of Bayreuth, or for your car stalling when you roll down a hill in neutral, there's grown a popular perception of computers as semi-sentient.

These sorts of things are practically never actually a computer error, of course. OK, perhaps it was, if the screen just went dark. But that's likely to be more accurately describable as a "hardware failure". Most "computer errors" are really either programming or data entry errors.

When a car spears off into the bushes, it's usually not because the car did anything wrong. Cars that run off the road are usually sent there entirely by the actions of a (possibly drunk or sleeping) driver.

Similarly, computers that get things wrong are almost invariably doing exactly what they were told to, to exactly the data with which they were provided.

But that doesn't matter.

People think computers are alive.

Few people hold this belief in anything like a whole-hearted, entirely serious manner. OK, some poor individuals genuinely do consider computers to be terrifying dead-eyed monsters devoted to the destruction of humanity in general and the computer-phobe's word-processor documents in particular. But these unfortunates are very much in the minority.

Most people, instead, just sorta-kinda mean it when they say "this thing hates me" after the third preposterous error message of the day.

When one of my teeming horde of PCs fails to proceed at a crucial moment and I find myself, as a result, discovering a lot more about the cruft behind the desk than I expected, I, for one, am prone to the use of colourful Anglo-Saxon expressions which imply not only that I consider the computer to be a sentient entity, but also that I know its undistinguished parentage in detail.

Do I really mean it? When a machine's being particularly intransigent, yeah. A bit.

I can't help this. Probably, neither can you.

But if you can restrict your personification of your PC to invective only, you'll deal with the thing better.

Human beings are social animals. We try to have relationships with things. We often don't, really; a relationship is a two-way thing, and a computer may be an amazingly complex machine, but it remains a machine. Only the most irritatingly abstruse of philosophers, or the rare enthusiastic animists, insist that a PC really has a mind.

But humans are prone to see consciousness in things that don't have it, just as we see faces in things that aren't. It's the way our brains work.

Thinking that your PC actually intends to foil your plan to write the Great Australian Novel, figure out this month's operating expenses or just play a darn game hinders your ability to deal with the problem.

If you want to invoke paranormal phenomena in order to explain why your computer always hangs one to three minutes into a 3D game but is A-OK for endless boring word-processing then hey, I can't stop you. But the explanation's more likely to be something like an under-powered voltage regulator feeding juice to your video card. When your new whiz-bang graphics card's in 3D mode it draws a lot of amps, and overheats the regulator, and then runs out of volts, and faceplants. There you go.

"Well, that's obvious," I hear you say. "Why tell people this? Who's going to give up fixing problems because they believe their PC will just come up with another one?"

Heh. You'd be surprised.

And you might, in fact, find yourself in the role of the PC-personifier, no matter how ludicrous you consider the idea at the moment.

For instance, a couple of years ago I had a computer that'd crash quite often, but only when I was out of the room.

If I was sitting in front of it - even if I was just sitting there reading a book - it'd keep running.

It wasn't a screen-saver cutting in that crashed it. It wasn't the vibrations from me walking away. It wasn't the room light being turned off, or the toilet being flushed, or air-pressure changes in the room from closing or opening the door, and it probably wasn't even the loss of the comforting smell of my feet next to the PC's intake fan.

Whatever it was, the computer clearly didn't want me to see it happen.

Spooky, huh?

Lots of people have stories like this. Many of them are a great deal more convincing - and unnerving.

But what you have to remember in situations like this is that humans, including you, me, James Randi and Stephen Hawking, are not perfectly logical comprehension machines.

We misperceive. We misunderstand. We misremember.

This, coupled with our natural tendency to see faces in oatmeal and personality in tiny electronic pets, can lead us into severely misshapen ideas about what's actually going on when computers, or anything else, stop working.

People see things that aren't there and remember things that didn't happen, all the time.

The trouble is, this isn't common knowledge.

It is, however, the stock in trade of the magician - particularly the close-up magician.

A good close-up magician can cause you to seriously question the idea that people cannot in fact have paranormal powers. If you already believe in the paranormal, a good close-up magician will strongly confirm your belief.

There exist in the world more than a few people who can, for instance, leave you with the memory of a close-up magic performance in which you stared at a corked wine bottle whose cork, as you looked at it, suddenly dematerialised from the bottle and instantaneously rematerialised in the magician's hand.

Some of these sorts of tricks are done with ingenious, and often expensive, gimmicks. But others are done with nothing but carefully applied headology.

In the above situation, the magician, in brief, distracts you in a way you are unlikely to remember, uncorks the bottle in a possibly tricky but by no means spooky way, and thus leaves you remembering something that didn't happen.

Since close-up magicians can't perform for more than a handful of people at a time, most people haven't had this experience. Most people think magicians are all tuxedoed rabbit-yankers at least some of whose props are required to be too big to fit in a garage. This sort of big-audience magician won't teach you the same lesson.

People who watch on television a David Copperfield performance in which a Learjet or the Statue of Liberty disappear will, most likely, at least accept that it is possible that the people introduced as "independent witnesses" may, simply, be shills. And if you're in the audience watching a panda come out of the box into which the Beautiful Assistant was locked a moment ago, your mind may well turn to thoughts of mirrors, trapdoors, and extensive flexibility training for The Lovely Fifi, rather than dimensional portals or polymorph spells.

Close-up magic is different. It slaps you in the face with the inaccuracy of your own cognition. There does not appear to be any room for trickery when your wallet, which you know you never took out of your pocket since you put it there this afternoon, suddenly contains the playing card you picked, and signed.

That cognitive slap will do you good. Ask a policeman who's taken statements from people after a crime happened. Gee, it'd be nice if so many people weren't all so totally certain about contradictory things, and instead recognised that it's easy to believe what isn't so.

This is why I think everybody in the world would do well to, once in their life, take at least a mild hallucinogen. Not because I'm a fan of the transcendent connection with the universe, man, but because a good deliberate hallucination or three will, if it doesn't turn you into a religious nut, alert you to the fact that your perception of the world is very highly coloured by a very large amount of mental processing.

"You" are not your brain. You are something that your brain does. And there's lots of stuff that your brain does of which you, the consciousness that talks to people, are not aware.

To find out what's really going on, you have to be scientific. That means using techniques that prevent your faulty thinking from colouring your conclusions as much as it usually does.

Accordingly, I took notes on what my apparently spooky computer actually did, and when. I tallied when it crashed during use, and when I found it dead on coming back into the room. I did a bit of simple analysis.

The result?

Only a slightly higher number of for-no-obvious-reason crashes while I wasn't there than I'd expect, if there were a steady number of crashes per hour.

There were lots of crashes while I was there; I'd just not remembered them. It was a simple case of selective thinking and confirmation bias.

This was a PC running Windows 98, for heaven's sake. The thing crashed all the time as a matter of course. It'd be spooky if a heavily-used Win98 box didn't fall in a quivering heap at least once a day.

OK, the computer's for-no-obvious-reason crashes did happen somewhat more often when I was outside the room than you'd expect, if you were just working from the relative hour counts. But the thing about random crashes is that they're random, and can only be expected to smooth out into an even distribution as the sample size - the number of events at which you're looking - becomes large.

With a relatively small sample size, there's nothing remarkable about finding event clusters which, if they reflected a general trend, would indicate truly non-random behaviour to a high degree of certainty. In the short term, as disaffected day traders will tell you, any darn thing can happen.

Taking assiduous notes when an information system fails is, in reality, only worth doing if it's your job to do so. Or if you reckon you've got a chance of successfully suing the knuckle-draggers that created the deuced thing.

But recognising the fallibility of your comprehension of pretty much any somewhat complex situation will serve you well. At the very least, it'll stop the tech support people from filing your complaint in the PEBKAC pile because what you say is happening clearly, actually, isn't.

If you simply must persist in the belief that your PC is sentient and behaves like a supercilious waiter who may or may not be spitting in your soup, allow me to suggest a better way of dealing with it.

Have some dead and dismantled hardware lying around. I recommend hard drive platters; they're quite attractive, and look good hanging on the wall. Get three platters of different sizes and you can achieve a flying-ducks effect.

Buy a big screwdriver. Flat-head. Yes, I know the screws in the computer are all Philips and Pozidriv and Torx. That's not what the big driver's for. It's for prying and pounding.

You see, if your computer must have an opinion about you, I think it should view you as a technological Vlad Tepes.

If you're going to have delusions, you might as well make them fun ones.