[ print ]

Psychologist, MIT; Internet Culture Researcher; Author, Alone Together

I will see the development of robots that people will want to spend time with. Not just a little time, time in which the robots serve as amusements, but enough time and with enough interactivity that the robots will be experienced as companions, each closer to a someone than a something. I think of this as the robotic moment.

Sociable technologies first came on the mass market with the 1997 Tamagotchi, a creature on a small video screen that did not offer to take care of you, but asked you to take care of it. The Tamagotchi needed to be fed and amused. It needed its owners to clean up after its digital messes. Tamagotchis demonstrated that in digital sociability, nurturance is a "killer app." We nurture what we love but we love what we nurture. In the early days of artificial intelligence, the emphasis had been on building artifacts that impressed with their knowledge and understanding. When AI goes sociable, the game changes: the "relational" artifacts that followed the Tamagotchis inspired feelings of connection because they push on people's "Darwinian" buttons: they asked us to teach them, they made eye contact, they tracked our motions, they remembered our names. For people, these are the markers of sentience, they signal us, rightly or wrongly, that there is "somebody home."

Sociable technologies came onstage as toys, but in the future, they will be presented as potential nannies, teachers, therapists, life coaches, and caretakers for the elderly. First, they will be put forward as "better than nothing." (It is better to have a robot as a diet coach than just to read a diet book. If your mother is in a nursing home, it is better to leave her interacting with a robot that knows her habits and interests than staring at a television screen.) But over time, robots will be presented as "better than something," that is, preferable to an available human being, or in some cases, to a living pet. They will be promoted as having powers – of memory, attention, and patience – that people lack. Even now, when people learn that I work with robots, they tell me stories of human disappointment: they talk of cheating husbands, wives who fake orgasms, children who take drugs. They despair about human opacity: "We never know how another person really feels; people put on a good face. Robots would be safer." As much as a story of clever engineering, our evolving attachments to technology speaks to feelings of unrequited love.

In the halls of a large psychology conference, a graduate student takes me aside to ask for more information on the state of research about relational machines. She confides that she would trade in her boyfriend "for a sophisticated Japanese robot" if the robot would produce what she termed "caring behavior." She tells me that she relies on a "feeling of civility in the house." She does not want to be alone. She says: "If the robot could provide the environment, I would be happy to help produce the illusion that there is somebody really with me." What she is looking for, she tells me, is a "no-risk relationship" that will stave off loneliness; a responsive robot, even if it is just exhibiting scripted behavior, seems better to her than a demanding boyfriend. I ask her if she is joking. She tells me she is not.

It seemed like no time at all that a reporter for Scientific American calls to interview me about a book on robot love by computer scientist David Levy. In Love and Sex with Robots Levy argues that robots will teach us to be better friends and lovers because we will be able to practice on them, relationally and physically. Beyond this, they can substitute where people fail us. Levy proposes, among other things, the virtues of marriage to robots. He argues that robots are "other," but in many ways, better. No cheating. No heartbreak.

I tell the reporter that I am not enthusiastic about Levy's suggestions: to me, the fact that we are discussing marriage to robots is a window onto human disappointments. The reporter asks if my opposition to people marrying robots doesn't put me in the same camp as those who had for so long stood in the way of marriage for lesbians and gay men. I try to explain that just because I don't think that people should marry machines doesn't mean that any mix of adult people with other adult people isn't fair territory. He accuses me of species chauvinism and restates his objection: Isn't this the kind of talk that homophobes once used, not considering gays and lesbians as "real" people? Machines are "real" enough to bring special pleasures to relationships, pleasures that need to be honored in their own right.

The argument in Love and Sex is exotic, but we are being prepared for the robotic moment every day. Consider Joanie, seven, who has been given a robot dog. She can't have a real dog because of her allergies, but the robot's appeal goes further. It is not just better than nothing but better than something. Joanie's robot, known as an Aibo, is a dog that can be made to measure. Joanie says, "It would be nice to be able to keep Aibo at a puppy stage for people who like to have puppies."

It is a very big step from Joanie admiring a "forever young" Aibo to David Levy and his robot lover. But they share the fantasy that while we may begin with substituting a robot if a person is not available we will move on to specifically choose malleable artificial companions. If the robot is a pet, it might always stay a puppy because that's how you like it. If the robot is a lover, you might always be the center of its universe because that's how you like it.

But what will happen if we get what we say we want? If our pets always stay puppy cute; if our lovers always said the sweetest things? If you only know cute and cuddly, you don't learn about maturation, growth, change, and responsibility. If you only know an accommodating partner, you end up knowing neither the partner nor yourself.

The robotic moment will bring us to the question we must ask of every technology: does it serve our human purposes, a question that causes us to reconsider what these purposes are. When we connect with the robots of the future, we will tell and they will remember. But have they listened? Have we been "heard" in a way that matters? Will we no longer care?