Today, I’m dusting off two quickie-but-goodie items from my accumulated stash of articles. They’re two different perspectives on the same question, a question that lies at the very heart of this blog: Can we ever know what the future holds?
The Folly of Prediction
David Pogue, writing in last month’s Scientific American magazine, says future prediction is for fools. To be sure, it’s a lighthearted piece, but his point is well taken: It’s virtually impossible to make technology predictions without eventually looking like an idiot. He cites, as an example, the assertion that “there’s a world market for maybe five computers,” apocryphally said by the chairman of IBM in 1943. Oops.
It’s all good fun to laugh at Thomas Watson in retrospect, but this bodes really poorly for me! If the guy whose name now graces an advanced quasi-intelligent computer could get it so wrong, what chance do I have? I am but a simple blogger.
Fortunately, Pogue does throw me a small bone, namely the observation that it’s safer to predict something will happen, than that it won’t. If nothing else, he suggests, I can just add “yet” to any prediction that hasn’t come true, and I won’t be wrong. Hear that, readers? Henceforth, be it known that there’s an unwritten “yet” at the end of every wild speculation I post. Lawsuit averted!
The Prediction of Folly
The other item is this baffling idea featured in the same magazine, one issue earlier. Sociology researcher Dirk Helbing, in Zurich, wants to build a comprehensive computer model of everything in the entire world. His proposed system would integrate streaming data about politics, climate, economics, sociology, psychology, agriculture, technology, geology and more.
As a climate model might guess at next year’s hurricanes, Helbing’s “FuturICT” could predict the next financial crisis, or political revolt. Or you could input a new environmental policy, and then observe its cascading effects on land use, the local economy, the spread of disease, the careers of various politicians, and more.
Color me skeptical. Just to get started, Helbing needs an advanced computer system; vast amounts of highly detailed data; and intricate, novel algorithms that can turn the data into predictions.
Even if you take those as given, I’m pretty sure this “super-model” is theoretically unworkable. A single random event anywhere in the world could easily throw any of the predictions into disarray. The simplest possible system that contains every relevant variable in the world already exists. It’s the world itself, and I don’t think it can be simulated.
The Future of This Space
Look at me, not half a page later and I’m already breaking my new resolution about negative predictions! Maybe I’d be safer if I predicted Helbing’s success… but no matter how many “yet”s I add, if it’s impossible in principle, then it’s just plain impossible.
Well, shoot. Prediction sure doesn’t seem to have much of a future, does it? It’s enough to make a blogger wonder whether he should hang up his future goggles and go start the internet’s umpteen-millionth beer blog instead. Not to worry, though, in the end I’ve decided I’m not going anywhere. For one thing, I’m content to describe a nebulous future, a patchwork of various possibilities. Big ideas with far-reaching potential are more interesting than detailed predictions.
Plus, there’s only so much I could say about English brown ales before I meandered into the moral implications of artificial intelligence. Folly or not, the future’s in my blood – why fight it?