Science fiction isn't (as a rule) about predicting the future, and science fiction writers aren't trying to predict it.

"No sensible science-fiction writer tries to predict anything," says Frederick Pohl, whose work includes the classic The Space Merchants (written with Cyril M. Kornbluth), MAN PLUS, and most recently The Last Theorem, co-authored with the late Arthur C. Clarke. "Neither do the smartest futurologists. What those people do is try to imagine every important thing that may happen (so as to do in the present things which may encourage the good ones and forestall the bad) and that's what SF writers do in their daily toil."

But many science fiction stories are set in the future, which means they need to include the future of technology (or present reasons why things haven't changed). That is, they have to extrapolate from "what/where things have been and are" to "what/where might be."

CIO invited noted science fiction authors Larry Niven, Robert Sawyer, Nancy Kress and Charles Stross to share their thoughts on technology-related predictions, including lessons learned in the "business" of imagining what the future might be like. Here's what they had to say (via e-mail).

What have you learned about predicting technology's future?

From ramscoops and brain pleasure implants ("drouds") to Romulan-class warrior cats ("Kzinti) and Earth-orbit-sized habitats (the Ringworld), few science fiction writers have given us bigger visions than Larry Niven.

Niven has written or co-authored over 50 books, including the Ringworld series, and with co-author Jerry Pournelle, The Mote In God's Eye, The Gripping Hand and Oath of Fealty. Niven has won five Hugo Awards (awarded annually by science fiction fans) and a Nebula Award (awarded annually by the Science Fiction Writers of America). His most recent books include Juggler of Worlds (with co-author Edward M. Lerner) and the upcoming Escape From Hell (with Jerry Pournelle), a sequence to their Inferno. Niven's science fiction includes a wide range of technology we don't (yet) have, from room-temperature supercomputers (Ringworld) to "stepping disks" (manhole-sized unenclosed teleportation units).

Here's what Niven has to say regarding techniques for predicting a valid future technology:

• Look for the goals humankind will never give up. Instant travel, instant education, longevity. Then try to guess when it will appear and what it will look like.

  Pay close attention to parasite control. There is always someone who wants the money for something else.

• You're obliged to predict not just the automobile but the traffic jam and the stranglehold on gas prices.

  Nobody invents anything unless there is at least the illusion of a profit.

The only science fiction movie that did this right, according to Niven (it wasn't clear whether he was referring to the last point, or all his bullet points), was the 1983 film Brainstorm, in which, according to Niven, "a valid technology was followed from its inception to its limits."

"This goes right back to the Space Race, and the movie 2001," (directed by Stanley Kubrick, written by Kubrick and Arthur C. Clarke) says Robert J. Sawyer.

Sawyer has written over 20 science fiction books, including his Neanderthal trilogy, Factoring Humanity, and Mindscan, plus numerous stories. His most recent novel is Rollback, and he has a trilogy in progress (the first volume due April 2009) in which the Web wakes up. Sawyer is one of only seven writers in history—and the only Canadian—to win all three of the world's top Science Fiction awards for best novel of the year: the Hugo, the Nebula, and the John W. Campbell Memorial Award.

"No one predicted in 1969, when the first man walked on the moon, that the last man would walk on the moon just three years later," Sawyer points out. "When Arthur C. Clarke depicted all those wonders—artificial intelligence, suspended animation, floating hotels, cities on the moon, manned interplanetary travel—in the 1968 movie 2001: A Space Odyssey, those seemed like reasonable predictions for the dawn of this century, but none of them came to pass."

"The trap we science and space buffs always fall into is thinking that everybody will want the things that we want," Sawyer explains. "They don't; they have their own agendas, and ultimately, as in everything, it's the economy, stupid. Just because you personally want something doesn't mean there's a market for it. Just because we technically could do something doesn't mean that's how others want to see their tax dollars spent."

"Future tech can only be predicted short-term—say, 10 or 15 years ahead," states Nancy Kress.

Kress is the author of 25 books, including 16 science fiction or fantasy novels, two thrillers, four story collections and three books on writing. Her most recent books are Nano Comes to Clifford Falls and Other Stories and the present-day bio-thriller Dogs. Forthcoming is Steal Across the Sky, "an SF novel of galactic crime, genetic engineering and alien atonement." Kress has won every major award in science fiction, including four Nebulas given by the Science Fiction Writers of America.

Yes, Clarke got communication satellites right, Kress says "—but he missed computers in, say, the classic Childhood's End. When writers extrapolate from existing trends, that leaves out the wild card, which is where the most interesting tech often comes from—Alexander Fleming noticing a contaminant on lab bacteria, Steve Jobs tinkering in his garage."

Talking lobsters? Post-singularity hyper-tech? Inherited parallel world-hopping? Bank robberies in virtual reality? Who else could this be but Charles Stross? In addition to dozens of articles for Computer Shopper and other computer publications on the subjects of Linux, Perl and other topics, Stross has written dozens of stories and 16 science fiction novels to date, including Saturn's Children, Halting State and his Merchant Princes series.

"Donald Rumsfeld was right," states Stross. "That is to say, his point about the known unknowns and unknown unknowns nailed the problem of predicting the future spot-on. We can point to extrapolations of current technological and social trends, but we can't extrapolate on the basis of stuff that hasn't been discovered yet. For example: In 1962 it was possible, just about, to see the future of integrated circuitry (and even, if you were very far-sighted, to glimpse Moore's Law and its implications), but the CD player was right out of the picture— solid state lasers lay at least a decade in the future."