In his latest novella Ted Chiang tells us where we’re going. Sorry, there will be no ready-made robot butlers there.
Photography by Andrew Waits for City Arts.
TED Chiang has no children. The forty-two-year-old Seattle native does have four dependents living with him and his girlfriend in a Bellevue home, but they are all cats.
“On the spectrum of pets, cats are relatively undemanding,” the author says, taking a long pause to consider his choice of animal. “The time it takes to raise a useful cat is not that long. The time it takes to raise a useful dog is longer. If you have a border collie, you are going to have to put some serious effort into that dog. But that’s nowhere near the time it takes to raise a useful human being.”
Chiang and I are sitting at a booth in the Experience Music Project and Science Fiction Museum café on a quiet June afternoon. At the other end of the misshapen building is a collection of artifacts from science fiction’s past: Buck Rogers’ XZ-38 disintegrator pistol, Darth Vader’s helmet from The Empire Strikes Back, the pistol wielded by Harrison Ford in Blade Runner, as well as first-edition books by the authors Chiang devoured as a child: Arthur C. Clarke, Isaac Asimov, Ray Bradbury. In front of me is the man who, some have said, is the future of the genre, a quiet, yet deeply inquisitive author who has little patience for embellishment in his writing and no stomach for science fiction that fails, in his exacting mind, to do its job.
Chiang’s musings about pets have come up as we discuss his latest work, The Lifecycle of Software Objects, which tells the story of two people named Ana and Derek and their difficulties in raising three synthetic pets called digients that turn out to be more than they bargained for. The book will be released this month to much anticipation from the science fiction community. The publisher, Subterranean Press, is sold out of its first special-edition run and is already planning future print runs. To be clear, this is not a blockbuster. Chiang is not expecting to field offers from Hollywood, go on book tour or even to lighten his workload as a freelance technical writer. As a writer specializing in short works of fiction, he is under no illusions. “Nobody makes money writing short fiction,” he assures me.
And even if somebody could, it probably would not be Chiang. He just doesn’t write much of it. Since publishing his first short story in the now-defunct science fiction magazine Omni in 1990, Chiang has authored a paltry ten works of short fiction and one novelette. Those few works, though, have earned Chiang science fiction’s greatest prize, the Hugo Award, twice. Lifecycle, a novella, is his longest work to date. It was probably, he says, also his hardest work to write, likely contributing to the gray hairs that streak his long black hair, the only telltale sign of the baby-faced author’s age.
“Lifecycle took a very long time,” he says, taking a moment to sip the mango-pineapple juice in front of him. “I wrote a draft of it three or four years ago, and I had to put it aside. Then I worked on a different story. Then I went back to it and wrote almost an entirely different draft and then I had to set that aside for a while. Then I finally returned to it, so it’s been in gestation for quite a while.”
Chiang had intended the story to be a work of short fiction but found that the ideas he was attempting to grapple with demanded a larger stage. Initially inspired by Alan Turing’s work from the mid-twentieth century regarding artificial intelligence, Chiang soon discovered that he could not take shortcuts. In fact, he would realize that, in science fiction, shortcuts are the problem.
“In science fiction, you see a lot of artificial intelligence stories,” Chiang says. “Stories where people have robot servants, or HAL 9000, or something like that, which are super-smart, they’re totally obedient, they can even be funny or loyal or any of those things. I felt that there is this ... ”
Here Chiang pauses, his brow furrowed. After holding that pose for a full ten seconds, he continues.
“ … error or an assumption in all these stories that if you are able to create artificial intelligence, when you turn it on, it will instantly be your super-smart, super-loyal, perfect robot butler.”
Chiang utters these last two words with great precision, hammering the plosive “b”s. Then he pauses. His face doesn’t reveal it, but you get a sense that Chiang enjoys this phrase, “robot butler.” He continues.
“I think that is an error in thinking about the nature of intelligence,” he says. “I don’t think you can get something that behaves that way the moment you turn it on. So these stories are ignoring, or they’re glossing over, the fact that you would need to spend years to train an intelligent robot how to be a useful servant, in the same way that it takes years to turn a human being into a useful employee, a useful butler, a useful slave. It takes a lot of training and time. By glossing over that period, those years of training, they are ignoring a lot of significant questions, because that is a big investment, and I don’t think these stories posit a good model of people spending years training these robot butlers, or robot servants, or robot slaves.”
When talking with Chiang, it is easy to forget that you are talking about science fiction. He easily slips from discussing the role fiction plays in the exploration of complicated ideas to the idea of science fiction as a model for the application of those ideas. And then he’s back to talking about HAL 9000, the star of the Space Odyssey series of books and films.
“Computers are often described as learning,” he says, again pausing for a long moment. “But it is not at all clear what their starting point was. With HAL 9000 in 2010, they talk about the very first lesson that HAL has with his trainer, and HAL goes, ‘Good morning, Dr. Chandra. I’m ready for my first lesson today.’ But it’s clear that he is already an exceedingly polite and subservient butler. On the first day of class, no child says, ‘Good morning, Dr. Chandra, I am ready for my first class.’ No!”
Chiang slaps the sides of his cupped hands down on the table and smiles in disbelief.
“The first day of class, kids are crying and screaming, and it takes another fifteen to twenty years for them to sit still in class and say, ‘I am ready for my lesson.’ It is stated that the computers learn; they start out knowing a lot more than a person starts out knowing. You say that the [computers] are learning, but you are assuming that they come pre-filled with all of this experience. Where does that come from?”
In Lifecycle, Chiang does away with this trope of the genre. Taking inspiration from the theories posited by Turing, the author constructs a narrative built around a model of artificial intelligence more akin to child rearing than computer programming. Over the course of years, the novella’s human characters go from owning clumsy, charming virtual pets capable of learning new tricks, to developing real-world bodies for those pets, to starting schools for them. Ultimately the characters reach the point of contemplating independence for what have, by the story’s conclusion, become their children. It is a moving portrait of life, filled with frustration, failure, horror, comedy, misunderstanding and, most important, love.
This is not your grandfather’s science fiction. And that, Chiang’s supporters say, is exactly what science fiction needs right now.
“I think, oddly enough, science fiction is in trouble.”
I am on the phone with Gavin J. Grant. Along with noted speculative fiction author Kelly Link, Grant runs Small Beer Press, a small publishing house in Easthampton, Massachusetts. In October, the press will republish a 2002 collection of Chiang’s previous works, Stories of Your Life and Others, a body of work that, Grant admits, he has been coveting since it was published by Tor Books, the house that has been recognized as the best science fiction publisher by industry organ Locus Magazine every year since 1988.
“Science fiction is one of the biggest threads in popular entertainment,” Grant says. “There are lots of big movies, lots of TV shows based on science fiction premises, but science fiction is having this real trouble looking into the future and imagining what will happen. To look a decade or a hundred years in the future is very difficult, and there are a couple of outs that writers have been using. One, they say it’s too difficult to look into the future so we don’t have to, we can just write fantasy. Or they just look in the past and write steampunk and things like that. Don’t get me wrong, I like these things, but I think one of the things Ted does is to extrapolate rigorously and somewhat harshly, in a way that people can recognize from the life they’re living.“
Grant has not read Lifecycle, but he might as well be talking about it. There is no polish on the novella’s world. Chiang’s characters live in a place that is very much like our own, and the mechanisms that make the story possible are all present in our everyday lives. Lifecycle’s Data Earth, where Chiang’s humans escape from their monotonous lives and where the digients grow and learn, is a fictional version of Second Life, the virtual world where real people escape from their monotonous lives. The tech sector as envisioned in the book is strikingly similar to the one we’re familiar with in real life, featuring the same instability of start-ups and breathless pace of development. Even the seemingly unattainable technology that could create a synthetic being with the ability to learn the way a child does is no longer far out of reach for us.
A 2008 study called Parsing the Turing Test claims that the theories that Turing voiced more than fifty years ago have been proven. The study describes a computer program titled, cloyingly, HAL that, after fifteen months of training, developed verbal skills that an “unsuspecting professional psycholinguist” could not distinguish from the verbal patterns of a fifteen-month-old child. “Daddy work bye bye,” says the real world’s HAL. “This sun bright bright bright,” says Lifecycle’s digient pet, Jax.
Chiang is not clairvoyant. The only powers he possesses are those of observation and imagination. With observation he can shine a light on the transforming world we live in now; with his imagination, he shows us the world we could end up living in.
This is the extrapolation that Grant is referring to when he talks about Chiang. It is the extrapolation that makes science fiction important. And it’s the extrapolation that popular culture strips from the genre – which is why, Chiang thinks, science fiction, as imagined by the masses, has become a shadow of itself.
Back at the museum cafeteria, Chiang’s composure is slipping slightly. He is talking about the current perception of science fiction and what it has come to represent.
“People have said, ‘District 9 isn’t really science fiction, it’s about apartheid,’” he says, referring to the 2009 alien invader film set in Johannesburg, South Africa. “To science fiction fans, that’s really insulting, because the implicit subtext is that it can’t be good if it’s science fiction. People want to avoid the taint of science fiction, because they associate it with something bad.”
Chiang begins rapping on the table, his fingernails clacking along with each syllable he utters.
“They say, ‘This is not science fiction. It’s about the effect of technology on human beings,’ or, ‘It’s about philosophical questions on the nature of reality.’ But that is exactly what science fiction is about.”
At this point, Chiang leans forward and looks me in the eye.
“I want to say that my work is science fiction,” he says, his hands now still. “It absolutely is science fiction, and, as a result, it is about questions of technology and philosophy and what it means to be human and our role in the universe.”
Chiang leans back and smiles slightly. I turn off the tape recorder and ask if he wants to take a tour of the Science Fiction Museum. “No,” he says. “I have looked around before. It’s a fine place. But I do think that they could do some really great things with it if they had some more space.” •