On May 13, I finally got to read my wayward science fiction story “It Is the Voice That Unnerves Me” in The Dread Machine. I had been submitting the story since the spring of 2019, and had thought many times about consigning it to the “retired” list. I knew every word, sentence and section break almost by heart, and had worried endlessly that it was stale or past due, but seeing it published gave me a chance to read it with fresh eyes.
Though her age and life situation were distant from my own, my main character, Doris, felt like someone I knew. Elderly and homebound, Doris falls under the spell of a tech gadget called a Remembrance, which she buys for her husband one Christmas. It starts as a small microphone around his neck, recording his every word. When he dies, his voice is installed in a virtual assistant, allowing Doris to keep her husband’s comforting presence by her side. Unfortunately, the device is hacked, and begins to torment Doris with her husband’s angriest words, making her wonder if her marriage was so loving after all.
As I read Doris’ story again, I thought of my grandmother and my other older relatives who had been targeted with all manner of products, from harmless mechanical cats meant to replace actual pets, to malicious schemes including impersonation and Social Security fraud.
In my story, I tried to imagine what would happen if the saccharine innocence of a playful voice assistant or a Companion Cat were exploited.
Observing me quietly at my desk—listening to me, I was sure—were the real inspirations for my story. Siri, Google, Alexa. The voice assistants that knew how to make me laugh with a well-placed “tell me a joke” command, but whose charm hid the potential to do much more damage than any phone scam. In my story, I tried to imagine what would happen if the saccharine innocence of a playful voice assistant or a Companion Cat were exploited. At the same time, it was strange to see these strong feelings rendered on the page, as if I had not just asked Google what time it was.
I finished my read-through, proudly tweeted about the story, added a link to my website, and closed out. I knew I would return to the story from time to time, perhaps include it in a collection someday, and otherwise look back on it fondly. I thought that the initial phase of publication was over.
Just over a month later, I was proven wrong.
On June 23, during an otherwise ordinary Twitter doom scroll, I saw a familiar-looking headline: “Amazon’s Alexa could soon speak in a dead relative’s voice, making some feel uneasy.” The story stopped me in my tracks, distracting me from whatever else I had been searching for. Morbidly curious, I clicked and read.
Rather than recording a repository of a person’s voice, as the Remembrance had worked in my story, Alexa would deepfake the voice based on an audio clip provided by the user. In the words of an Amazon official, “while AI can’t eliminate that pain of loss, it can definitely make their memories last.”
Feeling giddy, I looked through the reactions. They ranged from a blunt “how bout no” to more thoughtful if still skeptical considerations of the ethics. I saw several references to the 2013 episode of Black Mirror, “Be Right Back,” a spiritual predecessor of my own story which probed the idea of resurrecting loved ones in the form of AI.
I had taken a nearer-future approach to my story, picturing the first vanguard of reviving our dead relatives as clumsy, maudlin, and vulnerable to hacking.
I was astounded not only by the timing of the announcement, but also by the uncanny similarities of the circumstances. Rather than imagining a world of androids and fully sentient AI, I had taken a nearer-future approach to my story, picturing the first vanguard of reviving our dead relatives as clumsy, maudlin, and vulnerable to hacking. Something that wouldn’t happen through technology of the distant future, but via consumer devices that are on the market right now.
Of course, I indulged in some self-congratulation. My editors at The Dread Machine took full advantage of the moment. Internally I was also relieved—had the story not come out in May, it might have passed into the realm of fact. I remembered, though, my initial feeling of submission dread: that my story was not forward-looking enough. If the premise of my supposedly futuristic story had come to pass in a matter of six weeks, was I truly stretching my imagination as far or as deep as possible?
As I read article after article on the Amazon announcement, I also discovered the astounding array of issues I had not found room to fully explore in my 4,300-word story. How would someone consent to their voice being used after their death? Even if they did, would there be a limit to how long their consent lasted? Could a dead loved one’s voice last longer in a computer speaker than their real voice did in life? What would happen if Grandma, while resting in peace, started ordering Tide Pods on subscription, or was weaponized to ask about overdue bills or back taxes?
As with my other aimless spins through social media, I had been left feeling uncertain, and with more questions than when I’d started.
In the popular imagination, science fiction is seen as a predictive force, with authors from years past graded on how well they anticipated technologies.
In the popular imagination, science fiction is seen as a predictive force, with authors from years past graded on how well they anticipated technologies like cellphones, robots, self-driving cars and video chat. There is an entire genre of articles that invites readers to marvel at how accurately books, TV shows and movies saw these technologies coming—or scoff at how badly they got it wrong.
Many examples of accurate predictions are classic works, which have had time for their worlds to come to pass, but more recent examples abound as well. Lincoln Michel’s 2021 novel The Body Scout, which is about, among other things, the ominous prospect of pharmaceutical companies sponsoring our sports teams and bio-engineering their players, was vindicated after a minor league baseball team renamed itself the Wild Health Genomes after their medical-clinic sponsor.
Predictions are of course about more than just gadgets and details. The rise of Donald Trump, the broader threat of global fascism, the specter of COVID-19 and the rollback of reproductive rights made prophets of many dystopian science fiction writers, foremost among them Margaret Atwood, who has no problem saying “I told you so.” TV and media franchises, such as Atwood’s The Handmaid’s Tale or the series based on Philip K. Dick’s The Man in the High Castle (bankrolled, it should be noted, by Amazon) trade in the coin of prediction and counterfactual, all implicitly saying “this is the way the world might be, but should not be.”
I figured authors were bound to get something right simply by the law of averages.
For a while, I didn’t set much store by the more detail-focused predictions. Over a long career of writing science fiction, I figured authors were bound to get something right simply by the law of averages. For every prescient observation about the rise of mobile devices, there may have been a dozen premature hovercars or Segways. Science fiction stories can sometimes inspire inventions, stretching the boundary of the word “prediction” as life imitates art. I was more inclined to focus on an author’s vision, spending high school English class arguing over whether George Orwell or Aldous Huxley had a more compelling depiction of the way we would eventually live.
My recent experience has made me reconsider.
While I write science fiction and fantasy, my roots are in literary realism, and one of the trends I have most enjoyed watching in fiction over the past decade has been the steady erosion of barriers between what we think of as “literary” and “genre” fiction. I appreciate the fact that speculative fiction is increasingly infused with the realist’s eye for detail, social structure, the small observations and interactions that make up a life just as much as the broader setting outside a character’s window. I hope today’s speculative fiction writers open a broader conversation with their realist counterparts, not just the classics, but also the contemporary authors with a masterful sense of how to, in John Updike’s words, “give the mundane its beautiful due”—writers like Greg Jackson, Deesha Philyaw, and Clare Sestanovich.
Science fiction has long privileged the ambitious, operatic vision, and should never lose its sense of wonder, or grandeur, or horror. But a vision is nothing without the particulars. What would Brave New World be without Fordism and soma holidays, and what would Nineteen Eighty-Four be without the many lasting neologisms of Newspeak?
In the much-less-studied world I created for my main character Doris, the broader vision was a society where the elderly and infirm are not only victimized by scams, but where the scammers have adapted and deployed more sophisticated methods, capitalizing more directly on the heartbreak and grief bound up in a long-lived life. I am comforted to think that my story would not have worked as well without the Remembrance, the critical pin holding the vision in place.
I believe science fiction has a role to play in predicting the future, not just concerning the universal, but also the particular.
I believe science fiction has a role to play in predicting the future, not just concerning the universal, but also the particular. There are works of popular futurism that specifically set out with the goal of prediction and projection, and fiction with this same single-minded aim risks losing its focus on the trusted literary engines of character, plot, language, form and conflict. That doesn’t mean authors shouldn’t take a shot at predicting the next big invention—or sinister personal technology feature.
What will happen, though, when devices like virtual assistants and many other technologies predicted by science fiction become old-fashioned? Even if a writer’s vision of the future still rings true, will those dated details weigh it down?
In a speech delivered to the 2012 winners of the Whiting Award, novelist Jeffrey Eugenides instructed the honorees to “write posthumously,” a piece of advice gathered thirdhand from essayist Christopher Hitchens and Nobel laureate author Nadine Gordimer. Eugenides argued that “to follow literary fashion, to write for money, to censor your true feelings and thoughts or adopt ideas because they’re popular requires a writer to suppress the very promptings that got him or her writing in the first place.”
Reading Eugenides’ words at the time they came out, I was a first-time writing student and deeply diffident about my work. The notion of avoiding “literary fashion” had a strong impression on me during a formative period of my writing career. To me, Eugenides’ remarks posed a critical question: should fiction be timely or timeless? While his speech was more about making a career as a writer than the finer points of craft, I have always tied it back to the question of universal versus particular—to add detail and specificity is to set one in time and space, to write as someone living rather than dead.
I soon realized that the more I tried to write posthumously, the more it pointed up a critical flaw in my own writing. I focused on vision, forward momentum and plot at the expense of everything else. My stories were hollow, unobservant, oblivious to the texture of life.
Eugenides’ advice was not universally well-received, with some finding his words precious and hypocritical. In a mixed evaluation, author Todd Hasak-Lowy pointed out for The Millions that “there’s a powerful counterargument that challenges Eugenides’s advice as both simplistic and naïve.” Writing for The Atlantic, Noah Berlatsky made his opposition more direct: “[I]f you’re a writer, your time and place will shape you too…[y]our parents, or someone, taught you the language you’re using, and once you’ve begun in such a derivative manner, it seems silly to be embarrassed to go on with it.”
Whether they were set a year, a hundred years, or many millennia in the future, I was drawn to stories that placed their characters in a well-defined milieu.
The more stories I wrote, the less I found myself agreeing with Eugenides. I enjoyed stories, even those from decades and centuries past, that enveloped a reader in contemporary details, that felt very much “of their time.” I found that this pleasure translated to science fiction. Whether they were set a year, a hundred years, or many millennia in the future, I was drawn to stories that placed their characters in a well-defined milieu, that felt timely despite vast temporal distances. These were the stories to me that felt most universal.
Looking back at my writing process through this lens, “It Is the Voice That Unnerves Me” and Doris’ detailed near-future world represented a pushback against the advice that I absorbed in my early days as a writer. I wanted Doris to feel timely, and her exploitation by the Remembrance to feel urgent and present. The fact that reality was so quick to mirror her world felt like a validation of that choice.
Amazon’s efforts to create an AI virtual assistant that mimics real people will surely not be the last attempt to restore or extend the life of our deceased loved ones. Compared to the fresh horrors of the future, a speaker that talks in the voice of a cherished grandparent might eventually seem quaint or even charming.
It is up to science fiction writers not to just imagine this future, but to imagine its specific manifestations, the particular ways in which it will collide with, color, and influence a character’s life. I will continue to write with an eye toward this goal. As eerie, gratifying and unsettling as this experience has been, it has given me a chance to appreciate what draws me to speculative fiction, to reevaluate why I write, and to challenge myself to envision the future even more boldly. At the very least, six weeks between prediction and reality seems like a low bar to clear for next time.