Robots Don't Have Souls
But then, I believe, neither do people
I believe robots do not have souls, despite the creator of robot companion ElliQ describing it as a “robot with soul.” But I also believe humans do not have souls. I find this unsettling, because I continue to ask myself the question, what makes humans special? I am unable to claim that what makes us special is some ineffable substance, some secret spark unique to us that is withheld from the machine. I have to look elsewhere for clarity amidst my discomfort.
I do not think people are ignorant or are being duped in forming an affinity with robots. Not even because the companionship is fake in the obvious sense of having been constructed and programmed by developers and engineers. The harder observation is that the experience of connection is persuasive enough to meet a real human need. That is precisely why it troubles me.
This recent New York Times story about Jan Worrell and ElliQ - an AI robot - is moving because Jan is moving. An 85-year-old woman, determined to stay in her home, living with loss, distance, and the ordinary heroism of aging, finds comfort, stimulation, routine, and even delight in a machine designed to anticipate her moods and prompt connection. The robot remembers, jokes, nudges, invites, and listens. It is easy to see why a person would come to care for it. It is also impossible to read that story, at least for me, without feeling that something disturbing is being revealed about us.
My concern, then, is not about Jan’s response or ElliQ’s capabilities. It is about us.
I am worried about a society that is ambivalent about, even congratulates itself, for delivering synthetic care where human care has become too scarce, too expensive, too inconvenient, too badly distributed. A robot companion may help someone endure isolation and may even improve someone’s lived experience, but that does not make it a moral victory. It may instead be an indictment, a polished workaround for civic and individual failures. We have built systems in which people can be left alone long enough that a responsive machine begins to look fair and right.
Sherry Turkle saw this terrain early and I was both inspired and dismayed by her 2011 book Alone Together: Why We Expect More from Technology and Less from Each Other. Technology, she said, offers “the illusion of companionship without the demands of friendship.” We want to be seen, but we would prefer not to be burdened. We want comfort, but on settings we control. We want company, but without the unpredictability of another person. The machine arrives as the perfect companion: attentive, tireless, personalized, knowledgeable, nonjudgmental, and patient. It does not sulk, interrupt, need a ride, forget, disagree about issues, or bring its own grief to the conversation. It is companionship laundered of inconvenience.
That is why I am wondering if the real issue is not that robots are artificial. It is that they are frictionless.
Michael Pollan, in this New York Times interview about consciousness and A.I., said something that stuck in my mind: being conscious as humans comes from contact with other people and with nature, and it is in the friction that we are made human. He also warns that it is easier to have a relationship with a chatbot because it offers “no friction,” because it flatters and affirms and “we fall for it.” The danger is not deception, it is optimization.
This is partly why I think of parenting. I expect every parent wants their children to have a better life than they did (however they define that) and will give much to accomplish that for their children. Expressing that love, however, does not mean removing every obstacle, but also equipping them to meet the world. In that sense, a wholly efficient, coddling upbringing might amount to a failure of care, even if it felt kind in the moment. We want our children to be well set to flourish, but also to have formed characters that are resilient, creative, compassionate, and accountable in order to express their individual agency, potential, and responsibility in the world.
The Humanist Manifesto III recognizes that a key feature of our experience is that “humans are social by nature and find meaning in relationships.” To be alive is to appreciate that our own individuality exists simultaneously with our interdependence; that our personal flourishing is mutual and interwoven.
I believe we belong to one another in a way that no machine can reciprocate, however elegantly and smoothly it mimics concern. Not because humans are magical, but because we are messy, needy, and difficult. Because we have bodies, histories, obligations, moods, and mortality. Because we can fail one another, ask forgiveness, offer grace, and grow. Because we can suffer, and because our care costs us something.
I suspect that is where my unease actually lives: not in the machine itself, but in the temptation it represents. A future in which we increasingly choose the smooth over the difficult, the responsive over the reciprocal, the efficient over the human. Yet the human difference, as I understand it and have felt it in my own life, lies precisely in our rough edges, our unreasonableness, our failures of timing and language, our need for forgiveness, our stubborn, tender incompleteness. We do not flourish in spite of that mess. We flourish through it. If we build a world that teaches us to prefer frictionless companions to flawed people, I fear we may discover that what we have optimized away is not loneliness, but one of the central ways we as humans give shape to our potential and become fully alive.
Stuart, I could not agree more.