What's It Like to Be? Only Human Art Can Tell Us.
It’s not that AI art looks or sounds or feels soulless. It’s that it is soulless. And when it comes to art, the soul (i.e., consciousness) is all that matters.
Is there something it’s like to be a bat?
That’s the (somewhat confusing) question philosopher Thomas Nagel famously asked in his 1974 paper, What Is It Like to Be a Bat?
The paper explores the question of what exactly defines consciousness. When is a creature not just alive but actually aware of itself? Which creatures are conscious of their own existence in an experiential way, and which ones are essentially biological machines, a collection of instincts, impulses, and behaviors with no subjective experience?
Nagel’s thoughts on the subject would go on to form the basis for the neuroscience community’s equivalent of the Turing test. Its merits, applications, and limitations have been widely debated in the philosophical and scientific communities for five decades.
Nagel’s position goes like this:
A thing should be considered conscious if there’s something it’s like to be that thing. If you were that creature, could you be described as having a subjective experience of your inner world and the world around you?
This is best understood by thinking in terms of human beings, the only species on earth that we know with absolute certainty meets Nagel’s criteria.
Actually, that’s being a bit too generous. Each of us only knows for certain of one individual who experiences consciousness: ourselves. I know that I’m conscious, because there is something it’s like to be me. I know what it’s like, and that it’s like anything at all, because I am me. If you could somehow occupy my mind and body, you would experience what it’s like yourself. But you can’t.
The assumption that others of my same species are also conscious is certainly a safe assumption based on what we know about neurobiology and observed behavior. But it’s not an absolute certainty. We’re incapable of embodying the subjective experience of another person, so we can’t know for sure.
Basically: It’s vanishingly unlikely that I, Taylor Berrett, am the only conscious human being to have ever lived. But not impossible.
When it comes to animals, the conscious-or-not debate is fairly straightforward when you’re talking about intelligent creatures, then becomes murkier as you go lower down the intellectual totem pole.
Apes? Almost certainly conscious. Whales and dolphins? Surely.
Dogs and cats? The general consensus is yes.
What about lizards? Frogs? Fish and crustaceans?
Opinions begin to split. And what about insects? Dust mites?
Is there something it’s like to be a dust mite? Or are they just a collection of signals. Move forward. Move backward. Eat. Reproduce.
Many lower-level animals have pain receptors, but that doesn’t mean they feel pain. The feeling of pain is a subjective experience. Those receptors might just be how their brain says “something harmful is happening, move away from it.”
Where we draw the consciousness line — and how fine or fuzzy that line should be — is a problem scientists and philosophers have grappled with for centuries.
Now, before we go too far down the rabbit hole (or dust mite hole) of who is and isn’t conscious, let’s return closer to home. Let’s talk about a specific subjective experience many of us probably-conscious human beings have had.
You’re walking through an art museum, taking in beautiful classic works by the big-name artists of history. I call them the Beyonces, the greats who are frequently identified by a single name. Monet. Kahlo. Picasso. Le Brun.
Then you enter the modern art exhibit. You find yourself confronted with something like an enormous blank canvas with a single red dot in the middle. Or maybe a gray line. Or a pair of black squares side-by-side. Or maybe it’s even more avant-garde; maybe one of the pieces is just some cat urine on a paper towel, suspended from the ceiling by a cable.
Unless you’re extraordinarily open-minded, you might instinctively think to yourself:
“How is this art?”
The question of what art is has been debated, analyzed, over-debated and over-analyzed for so long that it’s become a meta-joke, a tired discussion thought to have no real value. We imagine a group of intellectual artsy types, probably wearing black berets, sitting around and pontificating on what art really is and what it’s for.
Point is, it’s a discussion that’s come to be seen as very academic without much concrete value.
Except something new has happened. Now a series of electrical impulses inside a computer can instantaneously scrape the entirety of all digitally available art or music and spit out… something. Upon request. In seconds.
What is that thing it’s made? Is it art?
Or, to put it another way: Does something that looks or sounds like the art people make, but wasn’t made by a person, art?
Suddenly, that “What is art?” question becomes a lot less academic and a lot more material. Even existential.
The answer to the question “What is art?” may also answer the question “Is art dead?”
Okay, now let’s get back to the animals.
Despite all of our advances in neuroscience, biology, and brain imaging, we still don’t know what it’s “like” to be a bat, or an insect, or even whether it’s like anything at all. We don’t even really know what it’s like to be a dog, despite loving them as a species — and despite the fact that we, you know, created them to be our friends.
We make guesses about dogs’ subjective experience based on what we know about their brain function and what we observe in their behavior. We look at their brains, and we see a spot in the brain that humans have, a spot that seems connected to conscious thought, and so we think our dogs must have it, too. When we see their tails wag whenever they see us, we feel certain. But we’re not. And even if we could be certain that they are conscious, we have no way of knowing exactly what their consciousness feels like. Let’s assume there’s something it’s like to be a dog. But we still have absolutely no idea what it’s like.
Even across individuals you would assume there must be a wide degree of variation. What it’s like to be a Schnauzer living with a family of six people would be very different from what it’s like to be a Great Pyrenees spending months alone, protecting sheep for a rancher in the mountains of Europe.
Now just imagine if dogs had language. Imagine if they knew how to express themselves with words, or even with images. Imagine what they could tell us about how they experience the world. We could learn more about dogs’ subjective experience in a single afternoon than decades of neuroscience have told us.
Human beings do have language. We’re the only known creatures that do. We’re able to express ourselves.
Think about that phrase: express yourself. You’re not just expressing an idea, or a fact, or a thought, or a memory you had. No matter what you’re talking about, no matter how basic, the way you talk about it is an expression of your very self.
To do that, you have to have a self.
Point is: we don’t have to wonder what it’s like to be a person, because we can tell each other.
How do we tell each other? Through self-expression. What’s another word for self-expression?
Art.
When it comes to being people, we know what it’s “like” to be a person generally, but we don’t know what it’s like on an individual level. We know what goes on within the circuitry of the brain, to a very limited extent. We may know that a certain area is for humor, another for spatial determination, another for short-term memory. But none of those things tell us what being a person is like on an experiential level.
So how do we find out? The people around us tell us. And in their telling us, we learn about what it’s like for us. That’s what art is for.
It also tells us what art is. It’s an individual (or group of individuals) making an intentional attempt to say something about what it’s like to be them. Even art that’s not introspective, that speaks to messages about society, class, humanity, whatever — it’s all an expression of someone’s perspective on those topics. That perspective is the result of their fingerprint-like, completely unique subjective experience influenced by everything from genetics and upbringing to what they had for lunch that day or whether they had a good night’s sleep or whether they’ve had a life of loneliness and despair or love and pleasure.
Imagine two strangers, standing side by side in a park on a sunny day. Their objective reality is almost identical. Same view of trees, benches, grass, sky, clouds, people, etc. Same smells, same sounds, same soft, damp earth beneath their feet.
But their subjective experience of that park is influenced by a whole host of things that are independent of the objective reality of that place and moment.
One of the strangers looks and sees a park that reminds her of a park she used to play at as a teenager, the same one she was playing at when her mother received the phone call that her father had died in a car crash on his way home from work. This memory casts the entire scene in a glaring, jagged light. The sun doesn’t shine, it blares down harshly. The sky is more pale-gray than blue, when she thinks about it, and she feels an odd yet familiar sense of bitterness rising up from somewhere in her toward the other strangers at the park that day. Look at them, she thinks, with their lives devoid of any real tragedy, they have no idea how lucky they are to be able to unambiguously enjoy this sunny day, this park, the experience of being alive. And yet, at the same time, she feels a strange sense of love and gratitude welling up in her chest, for her mother, who that day in that park became a single mother, and raised her and her brother as best she could, completely alone.
The second stranger, standing beside her, brings an entirely different perspective. Born and raised in a densely populated inner city, natural spaces like these almost overwhelm her with their simple beauty. How incredible, she thinks, to be here in this place, breathing clean air, standing on bare earth, the breeze rustling leaves in the trees, the sky stretched out above me, almost entirely unobstructed. Sure, she feels a twinge of sadness at the close friends she left behind when she moved to the suburbs, the wild late nights coming home drunk on subway trains, the feeling that everything important in the world was happening right there, right at that moment, right in that place that felt like the center of the universe. Still, she’s at peace with her choices, because they brought her to this beautiful moment and others like them.
Two people, having precisely the same objective experience yet entirely different subjective experiences.
That is why human art matters. Art made by a person, born out of their unique consciousness, communicated with deliberate intent, imbued with their unique perspective.
I did that “two strangers” exercise with students in one of my classes this week, asking one volunteer from each class to come stand beside me and face the class. I discussed how despite our identical view (in this case, the rest of the class and the performance space where we meet), our perspective — our experience of that moment — was entirely different. Me, a new teacher in a new country trying to use my humble status as the leader of this group of twenty young people for some positive end.
The student beside me— well, I don’t know. I can’t know. Until they tell me.
And so, all these questions of “Will AI replace human art and artists?” miss the point. They come from an extremely capitalist-centric, product-centric mindset which we’ll need to address before we can make any meaningful progress on safely integrating AI into our lives — a subject I’ll dive into in an upcoming newsletter.
Someday something might replace the need for human-made art, but it won’t be AI. The products of AI generative — they’re products made from products. All of their data comes from secondary sources.
But art is made from inside an as-yet-unknowable mystery box: the subjective experience of a human mind. Our desire to both express those mysteries and have them expressed to us won’t go away, until science enables us to go beyond interpreting a person’s subjective experience based on external evidence, or electrical signals on a computer screen. Until we can truly occupy another person’s experience — a monumental scientific achievement we are, frankly, nowhere close to reaching — we’ll continue to need art made by people, shared with people.
Even so-called derivative art — say, a ripoff of an existing pop hit written by a lesser songwriter — brings with it some potentially beautiful baggage. Any great artist will tell you that their unique personal style was, at least initially, born out of their failure to effectively copy their heroes.
When you tell me what it’s like to be you, I also understand a bit more about what it’s like to be me, and what it’s like for us all to be ourselves here, together, as a human race.
That’s what art is for. That’s not going anywhere anytime soon. And if it does, it won’t be AI that does it.
AI is progressing quickly. Every day, it does something that was impossible the day before. But these advances have all been a matter of degree. Exponential growth in one area doesn’t point to eventual breakthrough in a completely different area.
The AI that’s blowing all of our minds right now is generative AI. It takes your prompt, or request, and uses everything it’s learned based on its inputs — what its creators feed it from the internet, mostly — to try to give you what you asked for. It’s gotten really good at it.
But it isn’t conscious. There’s nothing it’s like to be Chat GPT. It isn’t expressing itself. It’s curating. If you peek behind the Wizard of Oz curtain, it’s Google with a language modeler and a drawing tablet. It’s taking all of its inputs and combining them into something “new,” in the sense that the product its created doesn’t exist in that exact arrangement anywhere else.
When you ask AI to create a piece of art for you, it will present you with something that looks like art, but isn’t.
If you’re wondering whether that’s a distinction without a difference, consider a famous thought experiment known as the Chinese Room.
Imagine that you’re in a room behind a locked door. Every couple of hours, an unseen person slides a handwritten note below the door, written entirely in Chinese. You don’t speak Chinese. But inside the room is an entire library filled with information about how to respond when presented with specific Chinese symbols. So you take the note, and you consult the books, and then you write a response and slip it back under the door.
To the person outside the room, the mystery person inside the room would appear to know Chinese. After all, they’re responding perfectly to every Chinese phrase you give them. But in reality, you don’t know anything. You’re just searching through a database to find the expected answer to a prompt.
That’s AI. It will continue to get more powerful, and it will continue to create more and more complex, genuinely unique products — visual designs, stories, even complete songs.
But they won’t be art, because art is self-expression, and self-expression requires a self. That’s not out there in the database. It doesn’t exist in any library. What AI-generated products can tell us a whole lot about what the collection of human made art looks like, and not one bit about even a single person’s subjective experience of being a human being. It will continue to be the equivalent of studying dog brains on tables to understand what it’s like to be a dog. It will only ever get us a small part of the way.
It’s not that AI products can’t look good or sound nice, it’s that they don’t come from the perspective of human experience, only its products. That direct expression, from brain to creation, is something which (so far) only humans can do. In that regard, AI has to be classified below a dust mite in our hierarchy. We know for certain that it doesn’t experience anything — we made it. When it comes to dust mites, we can’t be sure.
And even the most complex, visually striking AI-generated art must rank far below the urine-soaked paper towel we mentioned earlier. It’s not that AI art looks or sounds or feels soulless. It’s that it is soulless. And when it comes to art, the soul (I.e., consciousness) is all that matters.
I enjoyed the piece, although I myself reach a quite different conclusion!
I wrote a response to this opinion here: https://extramediumplease.substack.com/p/dont-let-ai-art-distract-you-from