The history of predicting the future is long, but objectively undistinguished. In spite of this fact, the Mayans, Nostradamus and many lesser-known prognosticators still have their adherents. There is a ready reason for this: the undeniable attraction of a window into the future. It is literally the stuff of legend. But honestly, if the Oracle of Apollo at Delphi was good at anything, it was keeping predictions fuzzy enough to have multiple interpretations. This was Nostradamus’ trick as well. Where the Mayans went wrong, if you will, was in providing an actual date for the end of the world, which allowed for a clear-cut assessment of prediction accuracy.
End of the world predictions aside, the real business of predictions today is focused squarely on understanding the rate and direction of technological change. This it turns out can be at least as hard as predicting the end of the world because it too requires specifics. Good predictions are by definition assessable over time. Good predictions are not necessarily accurate predictions, however. A couple of classic predictions from last century make the point: Admiral Leahy’s opinion of nuclear weapons, which he gave to President Truman: “The [Atomic] bomb will never go off, and I speak as an expert in explosives;” and Thomas Watson’s quote around the same year that "I think there is a world market for maybe five computers."
Like most things modern, predictions often wear the sheen of science or, at least, the shine of rigor. In spite of the difficulties, financial and technology analysts have converted the art of prediction into big business and there are more than 500 analyst firms currently in operation today. Since NSS Labs is the only analyst operation with its own testing facilities, should our analysts stick to what they know to be true (empirical evidence from testing) or should NSS customers expect predictions to follow?