Technology predictions so awful, they’re good

Making predictions about future technology, or even the impact of current technology is tempting fate. That is to say, you’re fated to be wrong, much of the time…at best. I should know; it’s what this blog does a lot. However, I’ve always liked the definition of an expert, attributed to Niels Bohr or Werner Heisenberg that goes, “An expert is someone who knows some of the worst mistakes in his field and how to avoid them.” I’m working on the how to avoid them part.

However, there’s falling off a chair, and there’s falling down a mountain. There are degrees of bad in making predictions. Some predictions are so bad, they are outstanding. Farhad Manjoo, one of the best IT/Technology bloggers around (over at Slate), has latched on to something that’s been going around the internet for a little while. It’s a 1995 Newsweek column by Clifford Stoll titled The Internet? Bah!. That gives you a clue. Manjoo’s post That Whole Internet Thing’s Not Going To Work Out summarizes Stoll’s article quite well, but reading the original doesn’t hurt.

By now Cliff Stoll knows the taste of his own shoe leather quite well, but Manjoo is smart to point out that Stoll is no Luddite and no stranger to the Internet. In fact, he was one of the people who worked with it in the early days as one of the first hacker hunters. Stoll’s dismissal of the Internet was born of knowing it perhaps too well. As Manjoo recounts them, here are some sample pronouncements:

The Internet that Stoll knew was “a wasteland of unfiltered data” where it was impossible to find anything useful. “Logged onto the World Wide Web, I hunt for the date of the Battle of Trafalgar,” he wrote. “Hundreds of files show up, and it takes 15 minutes to unravel them—one’s a biography written by an eighth grader, the second is a computer game that doesn’t work and the third is an image of a London monument.”

Stoll also dismissed the notion that anyone would ever shop online. For one thing, engineers hadn’t invented a secure way to send money through computers. Even if security were no object, most of us would still choose to buy airline tickets and make restaurant reservations in real life. Why? Because the Internet was “missing a most essential ingredient of capitalism: salespeople.” For Stoll, this was the Internet’s biggest failing—it lacked any capacity for “human contact.” The Internet would never take off because “computers and networks isolate us from one another.”

You get the idea. However, the meat of Manjoo’s post is not the humiliation of Stoll (who himself has acknowledged the colossal errors), but the lessons about making predictions that can be learned from Stoll’s article. I’ll list Manjoo’s key points:

1. Good predictions are based on current trends.
2. Don’t underestimate people’s capacity for change.
3. New stuff sometimes comes out of the blue.
4. These days it’s best to err on the side of optimism.

Manjoo makes a good observation that Stoll was actually too conservative in his predictions. Most pundits (and futurists) are way too optimistic. Most of the time the predictions come far later than people thought; if they happen at all. Think of all that was predicted by the incomparable movie 2001: A Space Odyssey, almost none of which has happened, yet. It’s still highly plausible though, because much of what is in that movie still looks like how we think it will be, someday.

Manjoo thinks that he’s inclined to give more credence to people whose predictions are on the side of outlandish (Ray Kurzweil being the exemplar), because new stuff does seem to come along more often than we think it will.

I’m very interested in how so many predictions get the timeline so wrong, because that’s almost always based on a ‘gut feel’ for the pace of change. There’s no way to find much ‘evidence’ for predicting, say, that we will have an operational Moon Base by 2020. So what’s in a person’s gut when they make these fairly long range predictions?

Different things for different predictions, of course, but to generalize there’s probably a sense of current trends (Manjoo point #1). In many cases, the prediction needs to be tempered by how business, government, and/or society will react to the predicted change (Manjoo point #2). I’d add a few more: A sense for things that can ‘go wrong’ with the prediction – the impact of events such as recessions, for example. There’s also a sense that at least in some fields with all the competing approaches (alternative energy is a good example), that there may be no clear ‘winner’ – something will replace oil, but it might be a combination of things, some of which have not been invented yet (that last is Manjoo point #3). Finally, Manjoo likes to be optimistic. With putting a date on predictions, I like to go long – add a decade or two almost automatically, simply because the time between the discovery of some important technology and when it finally becomes fully operative can be very long indeed.

It’s doubtful that making predictions for science and technology will ever be ‘a science.’ It doesn’t have to be totally subjective though. The sort of thing Manjoo is writing about can be the basis for a prediction framework – a structural way of looking at predictions and how to make them. No cookbook, for sure, but maybe a list of ingredients.

And most of the predictions will still be wrong…but maybe it’s like learning from history (we don’t), but it makes for good stories and it helps condition our thinking.

(Visited 131 times, 1 visits today)
This entry was posted in Spun and tagged , , , , , , , . Bookmark the permalink. Both comments and trackbacks are currently closed.

One Comment

  1. Posted March 3, 2010 at 2:02 pm | Permalink

    It seems to me no.1 and no.3 in the list are rather contradictory :-) Not to say both aren’t true, but it reinforces the point how useful hindsight is in working out which predictions were good and which bad. One trend that I think it’s hard to blame him for not spotting, for example, is that way the internet has developed, particularly via social networks, so that we are now much more likely to be who we say we are (e.g. compare Facebook with Usenet) – a point I’ve expanded on in

  • .