Predictions that Work

If you're ever up for a laugh, dig up futurist predictions for the 1930s or 50s, and see what they thought the future would be like. Yet for every atomic rocket car, there are some predictions that were eerily spot on. What are the characteristic of these predictions, vis a vis the ones that don't pan out?

The best predictions of the 21st century were those relating to the computer revolution. Moore's Law, JCR Licklider's papers about the internet (or Intergalactic Computer Network as he referred to it,) and Douglas Englebart's theories about human-computer interfaces. Their ideas were first formulated in the early 1960s, but took until the mid 80s to reach a critical mass of popular use. 20 years out is pretty good for a prediction.

Now, compared to the atomic flying car, to pick one particularly egregious failed futurism, these technologies possessed clear benefits even at early stages of their development, fulfilled a unique need, and had capital costs low enough to be purchased by individuals. As the cost and complexity of a system rises, the number of stakeholder increases, and the odds of a single person derailing the entire endeavor rises. An incomplete nuclear power plant offers little to no benefit. An atomic flying car has to actually work before anyone will buy it. Meanwhile, computers are useful as bulk data processors, and phone lines for communication, long before they become truly interactive.

Of course, the biggest advantage that Licklider and Englebart had was that they made their future. First they envisioned it, then they built it. Scientists have historically been only slightly better at futurism than their lay peers, and no better at the social aspects, but they are the drivers of innovation.