In earlier think-pieces I discussed a very normative, predictable, conservative (in the sense of unadventurous) version of the likely shape of the next century.
Of course, it's not going to be like that.
I have, in general, very little time for Donald Rumsfeld; but he's very occasionally right about something, and in February 2002, in the run-up to the invasion of Iraq, he made a rather remarkable speech for a contemporary politician; one in which he attempted to distinguish between categories of uncertainty:
[T]here are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns — there are things we do not know we don't know.Contorted though his language might be, that's a pretty good guide to the future.
Here's my recipe for building a near-future world (in the context of writing an SF novel).
Start with a horizon 10 years out:
85% known knowns
10% known unknowns
5% unknown unknowns
For every additional decade, knock 7% off the "known knowns" column and add 5% to "known unknowns" and 2% to "unknown unknowns". Very approximately. Warning: warranty expires when "known knowns" drops below 67%, i.e. around 2052 if we do this exercise right now (in 2012).
The devil, of course, is in the details.
Let's take a simple example: cellphones. We know what the state of the art is right now. We also know roughly what ARM and Intel are planning for the next 2-3 years. We know that mature segments of the consumer electronics market tend to split between a major incumbent with 80%, a minor incumbent with 10-15%, and a bunch of also-rans. (This happens in other market sectors. For many years the automobile sector in the USA split roughly 80:15 between regular cars and pick-up trucks. It's now been disrupted with the arrival of SUVs, blurring the car/pick-up boundary, but we're in a time of change anyway as electronics also impact automobiles.) We can expect smartphones to therefore settle into an 80/15/5 split over the next decade, with a major incumbent (say, Android—on present form), a minor "different" incumbent (iOS—although the place I've assigned to Android may end up with Apple, and vice versa), and also-rans. Which incumbent ends up where is therefore a known unknown.
The "unknown unknown" for phones over the next decade is that a Carrington event wipes out all our high-tech infrastructure and we starve, or maybe a sudden breakthrough gives us pseudo-instantaneous quantum entanglement instead of radio, or Windows 9 Mobile takes the world by rapturous storm and the Apple Taliban ditch their iphones and switch to Windows with shrieks of glee.
(I ought to add a fourth category of unknown called the "implausible unknown" — developments not compatible with the laws of nature as currently understood, or overturning major scientific paradigms. Tachyons, alien invaders, or telepathy all fall into this basket, and if you dumpster-dive it for ideas in fiction you are, at best, writing science fantasy.)
Again, two other "unknown unknowns" bit SF authors in previous decades. Prior to 1980 the portrayal of personal computing devices in SF was noticeable by its absence. There were a couple of books and stories that had them, but by and large they were invisible. On the other hand, prior to 1990 virtually all SF set in the near to distant future presupposed that manned space travel was going to be relatively easy and commonplace: a picture that doesn't look anything like as inevitable today.
Two important aspects of the unknown/known unknowns approach are that the further into the future we peer, the more the unknown unknowns stack up; and also, as new evidence comes to light, stuff may need to be shuffled between columns.
The first point should be obvious. To anyone writing an SF novel prior to 1987, it seemed perfectly obvious that the USSR was stable in the medium to long term. Boy did we get sandbagged by that one! And now, with 20/20 hindsight, the signs are obvious that dictatorial systems in general are unstable once the dictators are not hereditary monarchs with an incentive to pass a working system on to their children and a broad base of popular support. During the 1990s and early 00s, with very little fuss outside the continent in question, South America transformed from a continent dominated by quasi-fascist dictatorships to an almost complete sweep of democracies. In 2011, the Arab world erupted in democracy demonstrations and revolutions. The short-term process of the Arab Spring is still playing out: but one long-term consequence is clear—any diplomat who bases their foreign policy assumption on the "known known" that "the Arab street doesn't care about democracy" is sticking their head in the sand.
So in consequence of the collapse of the Warsaw Pact and the USSR, the South American transition, and the Arab Spring, I think it's reasonable to move "democratic revolutions in dictatorships" from the unknown-unknown column (where it resided in 1984) to the known-unknown column (we know it's going to happen; the only questions are where and when).
Another principle of near-future world-building is that human civilizations are fractal. That is: all human cultures have in common the fact that they are organizations of humans. In the absence of a major change in the nature of humanity (which falls under the heading of "unknown unknowns"— we can't rule out strongly superhuman AI or intelligence augmentation, but we don't know when it's going to happen or even if it is going to happen) we have the same Lego bricks to build civilizations with, so certain design patterns keep recurring. Moreover, the bricks decay (they die of old age; and their bureaucratic institutions—an attempt to build patterns that outlive the decay of individual bricks—succumb to capture by special interests or outlive their societal context). Knowledge is lost, and in the absence of knowledge gained by personal experience, the same damn mistakes keep being repeated over a time scale approximating a human lifespan.
We live longer these days, but I would feel fairly confident, in an SF novel that posits no breakthroughs in intelligence amplification or longevity, and no huge rupture in the way we allocate resources, of predicting some sort of major economic collapse/disruption circa 2090-2110. Because those of us who lived through and understood the ongoing crisis since 2008 will be dead of old age by then, and bright young things who've never even heard of the Great Depression, much less the Long Depression of 1873-96, will assume nothing like the horror of 2007-2015 can happen to them. Recurrent depressions are a known-unknown corollary of industrial capitalism. An unknown-unknown would be a singularity event in which vastly intelligent, benevolent entities give us everything we want for the asking, or in which we live 200-600 years (so those of us who remember the last depression are in a position to prevent the next one), or someone comes up with a virus that modifies the way we think, making New Soviet Man a reality. Or something else that breaks the self-similar patterning of industrial capitalism so that the Penrose tiles human societies are built from are replaced by new patterns.
So, to summarize: we have a complex mess of human societies that are, nevertheless, full of recurring non-identical variations on a core theme, because the common object the societies are all built out of are human beings. Barring changes in the nature of human beings, we can therefore expect chunks of history to be built out of the same components. We also have a future where, the further out we probe, the more unknown unknowns we run into; but in the short term, the main factors that shape what the world looks like are familiar. We can expect the world of 2022 to look similar to the world of 2012, insofar as many of the same cars will still be on the roads, fashion continues to iterate around a bunch of attractor themes scattered over the past century, many of today's large corporations will still exist (although some will have collapsed), and so on. There will be some surprises (maybe there'll be a hotel in space, or a Chinese Moon base) but overall it will be recognizable. But by the time we push the boat out to 2032, the unknown-unknowns will be building up. Signs of climate stress and overpopulation will be more visible, we may have driverless cars, there may be major disruptive effects arising from the development of direct brain interfaces or something else that today is a research and development curiosity. And by 2052, the unknown unknowns will have driven the world to be a very different place from anything I can predict today.