1. A recent lunch conversation skittered around awhile before landing, not atypically these days, on how we should all be preparing for the AI apocalypse.
My husband tends to imagine the impact of artificial intelligence in terms of employment, a la Kevin Drum’s “You Will Lose Your Job to a Robot.” But our friends described a more dire scenario, which they’d absorbed from a 2017 Vanity Fair article profiling Elon Musk. In this view humanity’s only hope in the face of the inevitable AI revolution is to forge a path to Mars. And even then, they noted, our chances aren’t great.
I’m putting this glibly because it all feels glib to me—like Y2K with a slicker public relations team. But the forms AI might take, and how those different forms might affect how people really live, still seems worthwhile to ponder. If AI is to thrive, and if we are to thrive alongside it, what will that relationship look like? For answers, my go-to isn’t Musk or even Stephen Hawking (whose grim outlook was also quoted in Vanity Fair). It’s science fiction.
2. “Science Fiction Makes You Stupid,” the headline reads. Well then.
The post’s author, Chris Gavaler, admits that the blog heading is flashier than his actual findings, which he published with co-author Dan R. Johnson under the more sober title, “The Genre Effect: A Science Fiction (vs. Realism) Manipulation Decreases Inference Effort, Reading Comprehension, and Perceptions of Literary Merit.”
Still, news that science fiction decreases mental effort and understanding, even if it does not go quite so far as to induce stupidity, is dismaying to those of us who find it a useful tool to think with (all the more so if we teach it to medical students).
That news also runs counter to recent arguments made by Tod Chambers and Phillip Barrish, respectively, that speculative fiction helps medical humanities scholars expand and enrich their work. Chambers observes that the discipline of literature and medicine relies too heavily on (and indeed, often underestimates) realistic fiction. Showing how three science fiction texts play with the cannibalism taboo to destabilize our confidence in the human/non-human species divide, Chambers concludes that we should “embrace the power of speculative fiction to challenge fundamental assumptions in bioethics.”[i] Barrish’s “Health Policy in Dystopia” analyzes future dystopias that depict a health care system—not medical technology itself, but the effects of medicine’s distribution and access—as either a primary symptom or even a cause of the dystopia. (The 2013 science fiction/action film Elysium is one of his examples.) “I am drawn to the genre of speculative fiction because of how it encourages readers, first, to develop an analytic understanding of the social structures of a given fictive world,” Barrish writes.[ii] Paying attention to those structures may generate insights about health care as a system rather than as, say, a series of clinical encounters or interpersonal vignettes.
Among skeptics of the Gavaler and Johnson study I would also expect to find Livia Arndal Woods, Gabi Schaffzin, and the authors of this piece in the Lancet as well as more razzle-dazzle voices like that of The Medical Futurist.
To dispute the study’s conclusion, though, we need to learn how its authors reached it.
Gavaler and Johnson conducted the study by gathering about 150 volunteer readers and dividing them into two groups. One group received a 1000-word passage of “narrative realism”; the other readers were given one of “science fiction.”[iii] The two passages, however, were nearly identical except for “setting words” like airlock for door and space station mess hall for diner. (And if you’re already objecting that mere architecture does not a science fiction story make, you’re in good company: several such critiques appear in the blog post’s comment section, where Gavaler engages gamely with them.) The authors then asked the readers in each group to gauge their story’s quality as well as the mental work they had put in to reading it (or “inference effort”). Finally, they tested the readers’ comprehension of their assigned passage.
Gavaler and Johnson found that readers faced with the airlock version scored it lower in terms of quality than readers of the door version. And to the point about “Mak[ing] You Stupid,” the airlock readers themselves scored lower on tests of comprehension and empathetic imagination (or “theory of mind”) after reading it. As Gavaler puts it: “The Science Fiction readers appear to have expected an overall simpler story to comprehend, an expectation that overrode the actual qualities of the story itself. The science fiction setting triggered poorer overall reading.”[iv]
In other words, people who think less of genre fiction will think less with genre fiction.
Let us not dismiss these poorer readers too quickly. After all, parodies such as Paul Simms’s “Eight Science Fiction Stories” would not be so impishly funny if they weren’t hitting such recognizable beats:
The gene-splicers had tinkered with the DNA, producing a race of warriors who craved just two things: the thrill of battle and the taste of their own feet. They hungered for battle. They literally ate their own feet. None survived to reproduce, and within a few short years they were all gone.
The Gene-Splicers chalked it up to experience, and decided to try harder the next time.
. . .
He’d had a real name at one time, but even he’d forgotten it. On the Net, he was known as Captain Fantastic, the Brown Dirt Cowboy—or CFTBDC69, in Net-handle speak.
He plugged the jack into the shiny port in the back of his neck and pressed Enter.
Twenty-six hundred baud of digital packetry surged through his system, and once the nausea—and the euphoria—wore off, he came to and ordered three polo shirts from jcrew.com without even touching his keyboard.
“If this is the future,” he said to himself, “me likey.”[v]
Powerful but short-sighted bioengineers; genetically enhanced “races” programmed for battle; the exchange of a natural identity and bodily integrity for entrance into a sordid, commercial virtual reality—these are all recognizable tropes indeed. If, as Gavaler and Johnson’s readers apparently suspect, science fiction has little else to offer us, then it will certainly not help medical students consider the implications of various biotechnologies, or the policy decisions that would make those technologies available to some but not to others, or how to weigh the benefits of these advances against the costs and ecological impact of their production. It would also prove little use to my AI apocalypse investigation.
3. I am not one of those who thinks less of genre fiction.
In my next post, I discuss the features and qualities of texts that make good on science fiction’s promise, both for students of medicine and for the not-Mars-bound.
[iii] The experiment also included a second version of each passage, where characters’ interior feelings and thoughts were spelled out: “Jim’s face is heating up now too, but not with embarrassment. He’s getting angry.” The additional explanation was meant to reduce readers’ “theory of mind” effort—how much work they had to put in to understand why the characters acted the way they did.
[iv] Chris Gavaler, “Science Fiction Makes You Stupid,” Patron Saint of Superheroes blog (October 30, 2017). https://thepatronsaintofsuperheroes.wordpress.com/2017/10/30/science-fiction-makes-you-stupid/.
[v] Paul Simms, “Eight Science Fiction Stories,” Shouts & Murmurs, New Yorker (September 14, 2015). https://www.newyorker.com/magazine/2015/09/14/eight-short-science-fiction-stories.