Workplace technology is going to eradicate inequality by increasing productivity—or it’s going to deepen inequality by making it hard to find work.

The internet is going to offer us unparalleled freedom of information and expression—or it’s going to consign us to surveillance and manipulation.

Mobile devices will liberate us from our desks and give us the freedom to travel the world—or they’re going to tie us down with an expectation of constant accessibility and the curse of digital distraction.

If we have trouble predicting the future development and impact of technology, it’s not due to a failure of imagination. Indeed, our greatest flights of imagination reveal exactly what makes technology so hard to predict.

I’m talking about science fiction, a genre that often features (or even revolves around) imagined technologies. But science fiction is also concerned with imagining possible futures, or alternative presents (and pasts).

That’s what makes science fiction so useful to those of us trying to anticipate the future of our digital world. The article “Science Fiction and the Future” quotes Arthur C. Clarke: “A critical . . . reading of science fiction is essential training for anyone wishing to look more than ten years ahead.”  And in “Does science fiction — yes, science fiction — suggest futures for news?”, Loren Ghiglione quotes author Orson Scott Card on the necessity of science fiction’s “thought experiments”: “We have to think of them so that if the worst does come, we’ll already know how to live in that universe.”

Both the idea of looking to the future, and the possibility of using fiction to do that, are relatively new. In “Has Futurism Failed?“, David Rejeski and Robert L. Olson write that:

A fundamental change in human thinking about the future began in the 18th century, as technological change accelerated to a point where its effects were easily visible in the course of a single lifetime, and terms such as progress and development entered human discourse…Speculation about the future became more common as human beings increasingly reshaped the world during the 19th and early 20th centuries, though it was seen largely as entertainment, a diversion from the often stark realities of everyday life. Yet some of that speculation proved surprisingly close to the mark.

This was the period that gave birth to the earliest examples of what contemporary reader might yet recognize as science fiction. As quoted in Science Fiction Studies, early sci-fi historian Hugo Gernsback wrote that:

Edgar Allan Poe may well be called the father of “scientifiction.” It was he who really originated the romance, cleverly weaving into and around the story, a scientific thread. Jules Verne, with his amazing romances, also cleverly interwoven with a scientific thread, came next. A little later came H.G. Wells, whose scientifiction stories, like those of his forerunners, have become famous and immortal.

It was Wells who advanced what is now a vibrant literary tradition of predicting the onward march of technology. In “The Future Through Yesterday,” Stephen J. DeCanio notes that:

Wells’s vision of future technology is rich. Wells imagined technological developments that altered the physical landscape. By 2100, people were concentrated in huge cities (the projected population of London is thirty three million) that are walled, not against any external threat but rather as a convenient means of controlling the weather…Wells also anticipated television, VCRs, and powered commercial and combat aircraft.

If Wells began the work of predicting technology through fiction, however, he also began the tradition of getting those predictions wrong. “Wells’s vision fell short of future realities,” DeCanio observes:

Fundamentally, the technologies he imagined were all merely extensions of the machinery of the late Victorian era. That is, the technologies were all essentially mechanical, macroscopic, and accessible. They were based on scientific principles whose workings could be apprehended by the unassisted human senses. Any intelligent person could master them…How this contrasts with what has actually happened! The most powerful physical technologies of the twentieth century are based on manipulation of the invisible worlds of subatomic particles and the electromagnetic spectrum.

Pick up any science-fiction book that pre-dates our current technological moment—which at this point means any book written before 2010, as well as books dating back decades—and you’re likely to encounter similar failures of imagination.

Take, for example, Edward Bellamy, who was a contemporary of Wells. In his 1888 novel Looking Backward, Bellamy envisioned the year 2000, including in that vision a tech-enabled reorganization of gendered labor. In “The Kitchen of Futures Past,” Nicholas Buchanan writes that the novel Looking Backward “captures the era’s enthusiasm for the capacity of technology to bring about utopia”—including:

the complete reorganization of housework that has taken place since the nineteenth century. Housework has been fully incorporated into the nationalized, public, paid economy, and women no longer spend their time and energies on cleaning and cooking. Instead, families eat at neighborhood cafeterias, where a professional—and financially compensated—staff devote themselves to the preparation and serving of meals such that the rest of society may devote themselves to other areas of work.

In Buchanan’s view, Bellamy’s predictive failures lay in his misunderstanding of the relationship between technology, economics, and social arrangements, which limited his capacity to imagine a world in which women enjoyed a more meaningful form of social equality. Instead, in Bellamy’s vision, women were limited in their professional options and political rights, as warranted by their limited participation in the labor force. “[B]y today’s standards, Bellamy was a person with hopelessly outdated attitudes towards the question of gender equality,” Buchanan writes, “unable to imagine his way out of his contemporary prejudices.”

A very similar shortcoming appears in the technological predictions of a much later novel:  George Orwell’s 1984. In “1984 and the Power of Technology,” Heinz C. Luegenbiehl writes that:

While the theoretical discussion of technology has an important role in the novel, the mention of actual technologies is cursory at best, usually being limited to one or two sentences. The major technologies dealt with are the telescreen, helicopters, the Floating Fortresses, the construction of the ministry buildings, the rocket bomb, atomic weapons, instruments of interrogation, speakwrites, novel writing machines, versificators, pneumatic tubes, and artificial insemination. Each of these developments contributes to the basic picture which is drawn in the book.

As Luegenbiehl observes,  “the technological developments discussed in the book are positively primitive. Today we can point to developments in each of the areas discussed which put Orwell’s supposed technological ‘vision’ to shame.”

Orwell’s mistake, in Lugenbiehl’s view, was an excessively narrow and utopian view of technology: He assumed that technological development would necessarily bring an end to totalitarianism because tech-driven growth would ensure that “wealth will no longer create a distinction between people, and their leisure time will allow them to begin thinking for themselves [which ultimately]…will result in the overthrow of the existing social structure.”

Instead, writing in the actual 1984—early in the computer era—Lugenbiehl observed that:

[t]he superficial comfort which technology brings with it also creates a sense of complacency in the population. Contemporary methods of data collection and storage and means of surveillance can make true privacy an impossibility if they are in the control of the powerful. It is thus not evident that technology is necessarily a barrier to the control of others.

Lugenbiehl perceived that gap in 1984, long before before Cambridge Analytica, before Facebook, before the internet was even widely accessible. It’s even clearer today. That’s to be expected, since the predictive ability of science fiction is heavily dependent on how far into the future it’s trying to reach.

“Today is the best indicator of tomorrow, but today decreases in value for points further in the future,” Joseph F. Coates and Jennifer Jarratt write in “Exploring the Future.” “All trends eventually slow down, stop, change direction radically, or reverse.”  In similar vein, David N. Samuelson writes in “Modes of Extrapolation” that “[t]he further we imagine into the future…the more technological innovations, cross-impacts, and secondary effects may interact.”

It’s these “cross-impacts and secondary effects” that illuminate not only the limitations of science fiction, but the limitations of the broader project to anticipate tech change. As Samuelson writes, “Astronomical predictions are pretty sure bets, and weather forecasting is more accurate than it gets credit for being. Predictions involving human behavior are more problematic.”

But all technology predictions are ultimately about humans just as much as they are about science. Humans determine not only which (and how) technologies get created, but also, how technologies get disseminated and used.  That means that all technology predictions are fundamentally blinkered by our current social reality, and especially, by our difficulty in understanding how our current reality biases and shapes our vision: We can’t see beyond the trajectory we ourselves are a part of. As John Huntington puts it:

If we in the present are going to think about the future in any scientific way, we have to reason from the experience of the past. For the future to be knowable there must be some pattern of continuity, some universal process, whether of change or of stagnation, which we have already perceived and which allows us to extrapolate to what will be. This process of looking ahead, as the writers themselves insist, is not visionary; its “scientific” basis, however, dooms it to be conservative, for in one way or another it must enforce some pattern from the past on the future.

When it comes to the predictive value of science fiction itself, it’s tempting to conclude that our ability to anticipate the future will always be tragically limited by our current social conditions, by our ability to extrapolate not just by decades but by centuries, and by the inherent conservatism of trying to predict future trends based on past patterns. All of these constraints apply just as much to non-fiction predictions, for our hopes and fears for real-world technology are just as limited by our ties to the present.

That’s not just a disappointment for science-fiction lovers (“Where are my flying cars?”) but also for all of us who turn to any form of futurism for clues on how technology will shape our lives in the years to come. “To be human is to ponder the future,” as Rajeski and Olson point out.  And at a moment when time appears to be moving so quickly—when technology is changing so fast that anyone over age fifteen or twenty can remember a pre-social media world that operated entirely differently—the job of anticipating the future feels more urgent than ever. We’re not just looking for clues about what life will be like in fifty or a hundred years: We need help figuring out how our lives will change over the next decade.

But even if our greatest speculative writers are unable to offer us a dependable guide to what’s just around the corner, that’s scarcely a reason to give up on the genre. As Huntington argues, “[t]hough SF often gives us a sense of facing the unknown, its true insights are generally into the known, and its primary value lies not in its ability to train us for the future, but in its ability to engage a particular set of problems to which science itself gives rise and which belong, not to the future, but to the present.”

Maybe science fiction writers can’t tell us when our robot overlords will arrive, or when and how our brains will be able to jack directly into the internet, or even how we’ll communicate in the years ahead. But in holding a mirror up to the powerful, painful struggle of living with new technology—any technology, in any time and on any planet—it can help us reflect on the biggest technological challenge of all: how to grapple with the role of technology as it stands today.