The Heart and the Chip: Our Bright Future with Robots by Daniela Rus and Gregory Mone; W. W. Norton, 272 pp., $29.99
When Daniela Rus tells people what she does for a living—designing artificially intelligent robots at the Massachusetts Institute of Technology—“they generally have one of two reactions,” she writes. “Some get anxious and … want to know when the robots are taking over the world. The second group asks when their car will drive them to work.” (Her answers? “Never, and not for a long time, respectively.”)
Those two emotions, fear and eager wonder, course uneasily throughout Rus’s new book, The Heart and the Chip (co-written with Gregory Mone). Parts of the book read like a fleshed-out PowerPoint presentation, full of listicle-style sections. Other parts, however, are far livelier, as Rus lays out all the incredible things robots can do nowadays—explore Mars, snake through rubble in disaster zones, deliver drugs to remote villages via “electronic stork” drones—as well as what awaits us in the future. Rus provides plenty of personal examples from her MIT lab, an experience akin to peeking inside Q’s lair in the James Bond franchise. But as wild as some of the ideas sound, the crucial components already exist.
Consider a tiny surgical robot that Rus describes. It would come encased in an ice capsule, which a patient would swallow. When the ice casing melts in the stomach, the robot would unfurl like origami and get to work. If a magnet is attached to the robot, surgeons could also use an external “wand” to guide it around. These tiny machines could repair organ damage without the need for messy incisions, or apply medicine directly to wounds, vastly increasing the potency. Best of all, the robot would be made of stiff sausage casing, so once it completed its job, it would dissolve harmlessly. Rus’s lab has already built a prototype.
The presence of a surgeon in that scenario isn’t incidental. Rus does not picture robots replacing human beings as much as robots and human beings working together—making us stronger, smarter versions of ourselves. “Robots can be designed to filter the dull, repetitive tasks out of our lives,” she writes, “and allow us to focus on higher-level work and interactions.”
Of course, despite Rus’s assurances, people do fear that robots and AI will steal their jobs, or at best demote them to underlings. To counter this dystopian idea, Rus contends that technology usually creates jobs: between 1980 and 2015, for example, computers eliminated 3.5 million jobs but generated 19 million new ones. Perhaps so, but were these good jobs? Quality should count as much as quantity. And for every inspiring new application of robotics in the book, Rus includes another idea that made me cringe. On one page, she describes a smart glove that could help an elderly stroke victim regain the use of her hands and write a birthday card to a grandchild. Wonderful. Then she talks about adapting the glove for children, to take control of their hands when learning to write and circumvent the hard work of mastering this skill. Dreadful. In spite of her warnings about the need to properly train robots and AI systems, she seemingly forgets that human muscles and nerves need training as well, and that failing and flailing are integral parts of learning. Short-circuiting that process has a cost. Such lessons apply to adults, too. As much as I’d love to turn all yard work over to feral Roombas and never rake leaves or shovel snow again, such chores do instill discipline and pride. They’re also forms of physical activity, something most of us need more of. Indeed, it would be one thing if people offloaded tedious chores to robots and spent their free time hiking mountains or running marathons—but lying on the couch eating potato chips seems likelier. We’ve all seen the human blobs of Wall-E.
Rus seems even more blithe about other dangers. Take her approach to the famous trolley problem in psychology. In short, people are asked how they’d respond to a runaway trolley on a train track, and specifically whether they’d throw a switch to divert the trolley away from, say, three old people and allow it to strike one young person instead. Do three lives count more than one? Or should the one’s youth—all those unfulfilled years—take precedence? People differ starkly in their responses. The dilemma has obvious applications in self-driving cars, where AI brains will need to make such decisions. So what’s Rus’s take? That we don’t need to bother with such hard questions. We’ll simply invent advanced sensors for robocars, allowing them to spot danger and hit the brakes in plenty of time to save everyone. Whether the brakes or sensors might fail gets no mention. This is all the more aggravating because, in the very next paragraph, she stresses the need “to think thoroughly and creatively about all the different things that could go wrong” when robots interact with the real world—despite her having just failed to do that.
Unfortunately, this was not an isolated lapse. In one section, she cites the myth of Icarus as an inspirational tale, writing that she always enjoyed the part about Daedalus building the wings, and preferred to ignore the bit about Icarus falling and dying. She also dismisses Disney’s Sorcerer’s Apprentice—a potential cautionary tale about bots running amok—by saying, “[T]he idea that the brooms kept working through the night while Mickey slept is more of a programming glitch than a real concern.” In both cases, she misses the point entirely. Rus later calls on her colleagues to consult science-fiction writers to help think through all the things that could go wrong with advanced, AI-powered robots. But I was left wondering whether Rus has thought such things through herself.
Rus calls herself a “robot mother,” a telling phrase. The Heart and the Chip soars at many points, and she has accomplished a remarkable amount in her lab. But mothers can’t always view their offspring objectively. It’s unfair to expect one book to solve the gigantic problems that robots and AI could introduce, and given Rus’s immense talent, society is probably better off just letting her play Q. But however clear her vision for our robot-human future—one in which “human expertise sets the scenario. Then robotic precision throws the bullseyes”—I found myself nervous by book’s end whether we’ll really hit the mark.
