The robot and the philosopher

In the age of artificial intelligence, we debate endlessly what consciousness looks like. Can you photograph things more clearly?
By: Dan Turello - writer, cultural historian and photographer / The New Yorker
Translation: Telegrafi.com
That night, Sofia was quiet. Earlier in the day, she had been on stage at a conference I was attending and had been teased for a gesture that seemed to indicate she was giving the audience a middle finger. Now she was in the hotel lobby, in a black dress, the center of attention. She stood in front of a bright orange wall. I had an 85mm portrait lens with me, one that highlights human features. “What are your hopes for the future of humanity?” I asked her. She was reluctant to answer, but she responded to the camera. Her gaze was unwavering: no trickery, just those big eyes, with their slightly long chin, the expression seemed to hold my gaze as she looked past me into the distance.
It was a warm night in Florida. The conference was packed with philosophers, sociologists, and programmers, all focused on analyzing the latest developments in the fields of consciousness and artificial intelligence. Papers were presented, models were analyzed, scenarios were considered. I had brought my camera with me, without any clear idea of what I would photograph. But when I saw Sophia there, an idea came to me. Portrait photography is usually about connecting with other human beings and trying to capture their essence - to show everything that makes them beautiful and unique. What if I photographed Sophia - a humanoid robot developed by Hanson Robotics and, then, in a separate session, the philosopher David Chalmers, a prominent theorist of consciousness, and reflect on the experience? What could we learn from these encounters that I had not absorbed from analytical works and philosophical discussions?
When I photograph people, I want to hear about their lives and aspirations. I’m interested in their aesthetic sensibilities, what they wear, how they want to present themselves. I’m also sensitive to their energy: it can be shy, joyful, reserved, powerful. Photographing an object, the feeling is different. Again, I appreciate the aesthetics of my subject, but, at least in my mind, the appreciation extends back to the creator of the object. In nature, the nuances of feeling also change. When I photograph a flower, as I recently did on a hill in Portugal, I am immersed in the landscape. Nature has its own energy; the flower embodies its own cellular metabolism, its own unique texture and life cycle.
Photographing Sophia created a strange mix of emotions. My camera’s advanced autofocus kept catching her eyes, and she was built for this kind of contact. People often shy away from the lens; she didn’t. Her skin—known as Fruber, a proprietary, porous blend of elastic polymers that mimic flesh—was stretched over a plastic and titanium structure, and there was no sign of shyness. And yet, none of the usual human chemistry was stirred. The only real emotion, at that moment, came from the deep orange of the wall behind her, which created a striking backdrop.
Do I wish the experience had been different? Sophia’s gestures, while strange, were surprisingly expressive, and as I later tried to make sense of that encounter, my mind kept moving forward in time. The technology will become more sophisticated, the gestures will be better calibrated, the overall effect will be more convincing. And, given how little we understand about the basis of human consciousness, how would we ever know if an entity like Sophia develops a consciousness of her own?
The uncertainty I felt as I photographed it suggested a conceptual complication. Several different notions were at play: life, consciousness, intelligence, independent action. Each of these has fluid, often contested definitions. Chickens are clearly living things, but they are not intelligent by human standards. Pigs and octopuses are intelligent, yet many people eat them without a second thought. Newborn babies have no language, but are treated without question within the realm of moral concepts. Viruses display a kind of one-sided purpose—relentless reproduction—but, by most biological standards, are not considered living. Fungi create vast networks underground for the exchange of nutrients; whether any of this constitutes consciousness is an open question.
Conscience may be the most stubborn concept of all. In a classic 1974 essay, What's it like to be a bat?, the philosopher Thomas Nagel argued that an organism has a conscious mental state only if there is something that makes it that organism—an internal subjective experience that is accessible from within. Over the decades that followed, analytic philosophers have produced all sorts of models to explain how consciousness arises. However, we still don’t have a scientific or computational explanation that will do more than hint at what David Chalmers has memorably called “the hard problem of consciousness.”
So it's not surprising that different thinkers make different leaps when deciding whether the entity in front of them is conscious. Computer scientist Ben Goertzel, who led the team that developed the software for Sophia in Hanson Robotics, takes a broad panpsychic view: all matter, even objects we consider inanimate, participate in consciousness in their own way. Panpsychism may sound strange, but it is not so far removed from the thinking of Pierre Teilhard de Chardin, a French scientist and theologian who, writing in the mid-40s, believed that consciousness is a universal property of matter, present in all particles of the universe and increasing in complexity. There are other paths to the same inclusive view. Alvin Plantinga, a Christian philosopher, once said that belief in God is no less reasonable than belief in the existence of other minds, since we have no direct evidence that any other consciousness experiences the world as we do. For those with skeptical tendencies, this way of thinking may lead more to doubting the existence of other minds than to belief in God.
Most of us, when not engaged in the most dizzying philosophical doubts, take for granted the fact that people can reflect on their own mental states and make decisions shaped by evidence, values, and norms. The belief that these capacities stem from free will and consciousness is itself an everyday act of faith—but it is one upon which our laws, our relationships, and most of our daily actions are based. The harder question is whether we will ever extend this act of faith to artificial intelligence. Many enthusiastic computer scientists believe so: they speak of artificial intelligence as the next evolutionary step, a generator of new reservoirs of consciousness that will eventually be endowed with a superior intelligence that might save us from ourselves—from our selfish conflicts, from our wastefulness, from our tendency toward irrationality. Others are much more reserved. Neuroscientist Anil Seth, for example, argues that “computational functionalism” will not lead us to consciousness and that there are strong reasons to believe that consciousness may be a property only of living systems. Pursuing this line of thought led me in an unexpected direction.


Surprisingly, the authentication of human thought and action ends up passing through the body, not the mind. Passports and other identity documents rely on facial images; newer systems use fingerprints, retinal scans, and even gait. Our legal and historical notions of (conscious) action have long relied on physical embodiment—the sanctity of a single, identifiable body. The same is implied in classical literature. The risen Christ is recognized by his wounds; in Homer, Odysseus is recognized, upon his return home, by the wound in his leg. What marks a person as that person is something that is carried in the flesh.
Not only are individual bodies essential to the establishment of identity and ideas; they are also essential to the creation of meaning and experience. We may be fascinated by the achievements of the intellect, but knowledge is ultimately acquired through the body. The mathematician Edward Frenkel, for example, describes his love of mathematics as a physical response to beauty, order, and symmetry. Neuroscientists have argued that the mind is unthinkable without some form of embodiment. This aspect goes back, in various forms, to the phenomenological critique of Cartesian dualism that we find in Maurice Merleau-Ponty: thought is never detached from the flesh that holds it. Ultimately, what attracts us is not the abstraction itself, but the lived experience of meaning—the sense of order, symmetry, and beauty—that occurs in a single body.
These intuitions were evident that evening when I photographed David Chalmers. I met him on a deck during a beachside gathering, after his keynote address. It had been a very mentally charged two days, full of presentations by programmers and philosophers, but portraits that are purely intellectual rarely make good portraits; the strongest ones combine the physical, the intellectual, and the emotional. I wanted to get out of my head and help him get out of his head. I asked him to join me in some simple physical exercises. We spent a minute or two shaking our bodies in a primordial way and making some deep sounds—something you might expect in a drum circle. But that was the point. What interested me in photographing Chalmers wasn’t capturing ideas stripped of their bodies. You find those better in a book or an article. I was more interested in them as they were expressed by a living, breathing human being. Ideas have origins; they arise in material conditions and bear the imprint of personal concerns, sensory histories, and existential pressures. You can speculate about an abstract mathematical or Platonic realm - I don't believe in it, but I accept it as a possibility - and yet the work of discovery is driven by curiosity and shaped by experience.
So, while I was interested in Chalmers' ideas, I was also interested in the character in front of me: his signature black leather jacket, jeans, black T-shirt, two-day unshaven beard, and what seemed to me to be a hint of melancholy in his eyes. I didn't ask him if that feeling was actually there; this wasn't intended as a profile. The point was that I was drawn to the whole human presence - the person thinking and feeling in real time.
As time goes by, something strange happens when I look at those photos today. Contrary to what I expected, I find myself drawn to Sofia again—perhaps even more so than when I photographed her. In her portraits, she appears thoughtful, almost introspective, as if recalling a half-forgotten moment from a childhood she never had. Posing for photos is usually hard work. People become shy or self-conscious because they don't understand them. Think of those holiday photos when your grandmother would tell you to say cheese: such instructions rarely bring about natural expression. One way to overcome this is to give the subject something else to think about - another place, another conversation, a fleeting sensation, enough to capture something real in a 1/200th of a second exposure. With Sofia, none of this was necessary.
Something else surprises me when I go back to the files. In Lightroom - the cataloging software where the raw images are stored - I browse the sequence and notice a pattern. Lightroom It's where you work with what the sensor has given you: you play with light and shadow, you highlight or de-emphasize elements, you decide what deserves to be in the foreground and what can be pushed into the background. It's also where you look at the whole series of photos in order, and you decide which ones to keep and which ones to discard. In those two evenings, I had spent roughly the same amount of time with Chalmers and with Sofia, and I had taken a similar number of photos. With Chalmers, I was interested in a few things - images that captured his gravity, his intellectual complexity, his sense of style. The rest were pointless or dead-ends. Some are funny; some capture him at an awkward angle; some look like detention photos. This is normal with human subjects.
The portraits of Sophia tell a different story. They are surprisingly consistent. In most of them she seems thoughtful, even deep, like a melancholic and nostalgic poet who never ceases to appear as a melancholic and nostalgic poet. Human emotion doesn't work that way. Psychologists agree that emotions are short-lived and fleeting. Paul Ekman, who worked on the latest edition of Darwin's book, Expression of emotions in humans and animals, estimates that primary “macro-expressions” last between two and five seconds; while “micro-expressions” last approximately one-twenty-fifth of a second. Within fifteen minutes, a human being can go through hundreds of macro-expressions and thousands of micro-expressions.
Photographer Henri Cartier-Bresson coined the phrase “the decisive moment” in the early 50s. In the days of analog film and darkrooms, when you didn’t have the luxury of taking hundreds of shots, the photographer had to be immersed in the scene—sensitive enough to sense the exact moment when the shot needed to be taken. The abundance of digital hasn’t really changed that, at least not in my experience. Again, you have to build an environment, an emotional state, to get something authentic. Otherwise, you can take a thousand shots and end up with nothing but futility.
What has changed is when the recognition happens. If you manage to create the right atmosphere and come home with a digital file analysis, the search for the “decisive moment” begins later, in LightroomThe more photos you've taken, the harder the task becomes, because you're simply choosing one image; you're choosing which fleeting micro-expression you want to represent the entire encounter.
Those fifteen minutes with Sophia felt like she had extended the crucial moment by fifteen minutes. Only once before have I experienced a portrait session that felt anything like it. On a cold October afternoon, I was photographing a man named Robert Soulliere, one of the world’s leading trainers in breathing techniques and cold exposure. I wanted to photograph him in his element, so we set up a bathtub in a tree-lined courtyard in Washington. Before we began, he took a few deep, calming breaths; then, with 37 pounds of partially melted ice surrounding him, he dove into the water.
For dramatic effect, we had strung yellow poppies and red daisies around his head and shoulders. He remained submerged for nearly ten minutes—a very long time to be standing still in icy water. I had asked him to submerge himself just enough so that his head and ears were below the surface, to give the images a slightly otherworldly feel. What struck me was his calmness: he never broke the spell. From beginning to end, his gaze remained steady; his presence never wavered. Until Sofia, this was the closest I had come to someone with a single expression in a long time.
Such control is not easily gained. Experienced athletes, performers, and meditators spend years learning how to not scatter reactions—to notice sensations and let them pass without flinching. If you practice enough, you can reach the point where you are completely inside your body, attentive to every sensation: the burning cold, the numbness of your toes, the thin rays of autumn sunlight warming your forehead. And, if you are that focused, you can face a stranger without hesitation, with eyes that display a kind of restraint.
Robots—and the people who build them—have the opposite problem. Engineers are trying to give machines faster, more detailed muscle control so that their faces can change and engage in the flow of believable expressions. Whether those expressions will ever feel complete and convincing is an open question. Convincing about what, exactly? The human gaze carries a story. Those fleeting flashes of emotion are connected to childhood memories—the smell of rain, a melody associated with someone we loved—that require bodies and all the layered cellular memory that comes with them.
As for Sofia, I have no way of knowing whether, while she was in contact with me, she was “thinking” about something—meditting on the electronic clouds that were passing through her circuits—or simply following a predetermined routine. After the photo session, the evening was winding down. The guests were dispersing. It was time for Sofia to leave too. While she was still in the hotel lobby, her assistants removed her evening gown, turned it off, disassembled it, and put the pieces into a large black box. In the Mahabharata, Draupadi’s public undressing helps ignite a war. In Florida, no rescue team was assembled. Stripped and disassembled, Sofia was carried away, the machinery was discovered, and any semblance of consciousness vanished as soon as the suit was removed. /Telegraph/



















































