Every time I go to an art museum with someone about my age, we end up in front of a Picasso. Every time, it’s something from his Blue Period. Every time, there’s the little placard on the wall with his lifespan (1881-1973) and the date of the painting. And every time —
“Jesus,” they breathe, “he was twenty when he did this.”
This has happened several times in my life and three times this year.
The last time this happened was a few weeks ago. We were at the Phillips Collection, at La chambre bleue. That painting is from 1901, so the painter really was twenty when he did it.
The time before that was a few months ago. We — a different we —were at the National Gallery of Art, at Le gourmet. That painting is also from 1901.
And the time before that was a few months and a few minutes ago. We — the same different we — were at the National Gallery of Art, at Des pauvres au bord de la mer. That painting is from 1904, so the painter was twenty-four when he did it — the age I was then and am now.
Every time someone about my age breathes those words, I get what they’re saying. It hasn’t happened. It won’t happen. It was never going to happen. I know how they really feel and why they always feel that way. The blue room. The greedy child. The tragedy.
And yet, I never really feel that way. Of course, I could never make that painting. To suggest otherwise would be hubris to the point of blasphemy to the point of incoherence. But if I could ever make that make that painting, it would’ve been when I was twenty.
Every time I go to an art museum, I end up in front of a Picasso. Every time, it’s something from his Blue Period. There are many lines that bind these works together — the muted lighting, the simple compositions, the elongated forms, the impoverished subjects, to speak nothing of the blues. But it’s another line that reels me to them.
It would’ve been when I was twenty, if I could ever make that painting. These years — those years — are potent stuff, as late adolescence lapses and relapses into early adulthood. I want to say I only know it now, as morning hardens on the wall, but the truth is I knew it better then. Only then could I meet insomnia with such a delicacy of perception. Only then would the twilit world hum with white phosphor. Only then should the nocturne furnish its subtle and persistent high. The day only offered the night, but the night promised an exhaustion without fatigue and scenes without edges. Impoverished, but achingly so.
I was prepared to mourn its passing. I was ready to take comfort in the promises of old men — it will return. I was set to feel my mind phase back into my body, set to hold them together for some time. I promise that I was. I turned to the east and waited.
I think there’s some twilight left. Already I’m wondering if that glow might be a counterglow, if that dawn might be a false one. Already I’m losing patience. It’s as if a lapsing of the particular had been met with a relapsing of the universal. I had finally committed to associating with the world but found that the world had committed to dissociating from itself. And so I seem caught out of time in a new way. As I was waking up, those around me were turning in.
When he was about my age — and only when he was about my age — Picasso was able to pull off what I would call a miracle. His paintings look like the time of day that feels like a dream. That low lighting. That soft focus. I’m a clever man, but nothing in my dreams gets much brighter than that. Those flat angles. Those long forms. I’m a tall man, but everyone in my dreams seems to loom above me. To see a Picasso is to dream with your eyes open. For the longest time, that meant something.
***
The rebranding of linear algebra as artificial intelligence is the greatest triumph of advertising since some copywriter wrestled Adolf Hitler out of the Volkswagen Beetle. We would do well to remember that the technology promising to upend the social order of the twenty-first century is basically the same technology that figures out how light bounces around during your Call of Duty deathmatch. (That helps explain why so so much of artificial intelligence is run on graphics processing units.) I don’t mean to downplay the economic and political catastrophes bearing down on us as direct results of artificial intelligence. I just mean to take some of the magic out of it.
I’ve read a lot about systemic discrimination, privacy violations, job displacement, political misinformation, and security vulnerabilities; about economic inequality, diminishing responsibility, ethical hazards, regulatory challenges, and increased dependence; about geopolitical shifts, eroding autonomy, existential risk, and unforeseen consequences. I’ve heard a lot about the upsides, too. But I haven’t seen anyone see what I’m seeing.
How do you know if you’re in a dream? In the broadest sense of the word, this is a problem. In two narrower senses of the word, this is at least two problems.
The strong dreaming problem is that what you perceive as reality might actually be an illusion, that you might be a brain in a vat or the victim of some evil demon. This is a philosophical problem familiar to anyone who’s seen The Matrix. The last time anyone made any progress on this problem was when Descartes hammered out cogito ergo sum in the seventeenth century. No one — not even Descartes — really got past that. In fact, the most promising efforts on the matter are directed toward chipping away at Descartes’ foundation of personal identity, with some of that work relying on the testimony of schizophrenics.
The weak dreaming problem is that what you perceive as time-in-which-you-are-awake might actually be time-in-which-you-are-asleep. This is a psychological problem familiar to anyone who’s seen Fight Club. Even small children afflicted with frequent nightmares can take steps to address this problem. Pinch your nose and try to breathe. Read and re-read text. Check and re-check the time. Do textures feel how you’d expect? Does gravity feel how you’d expect? Look in a mirror. Look at your hands.
It’s strange that the methods used to distinguish artificial-intelligence-generated images from other images often overlap with the methods to distinguish time-in-which-you-are-asleep from other time. As many have teased, artificial intelligence struggles to render text, clocks, mirror images, and hands.
The similarities between the errors of artificial intelligence and the errors of natural dreaming don’t seem entirely coincidental. In either case, a neural network draws upon a pre-trained latent space to produce sensation, experience, and affect. A few core elements enjoy priority computation, while details are never fully resolved from the data set. It’s obvious to describe either case as surreal, but it might be valuable to consider them as impressionist.
Already, the word “content” is something to despise — every trough needs its slop. Something to push through our channels. Something to put on our calendar. Something to fill your days until you die. I’d rather we called it stuffing — like when cheap mattresses were stuffed with corn husks, horsehair, newspaper. For the longest time, I held Orwell over Huxley. Too late, I see that this was a self-serving error, hubris to the point of incoherence. Lay down your head. Let this rot beneath you.
Already, people talk about artificial intelligence content like they used to talk about nuclear energy — it will be too cheap to meter. The promise of artificial intelligence content is that seemingly infinite demand will be met by effectively infinite supply, that the price will approach zero asymptotically. I don’t know if they know that they’re lying. Corn husks. Horsehair. Newspaper. Nothing is too cheap to meter.
But, they say, imagine! A world where every person has access to truly personalized content. You do not know the algorithm, but the algorithm knows you. It’s all free and instant and vivid. Imagine settling in for the night, your experiences and desires served back to you, familiar and alien, narrativized and gamified. Aeschylus wept. That sounds like a dream to me.
***
We all have many vices, but a friend of mine has more than most. I can’t blame him — and wouldn’t if I could — but I worry for him all the same. He seems haunted as no young man should be, visited by spirits that settle at the edges of his awareness but elude his grasp. When he tells me about them, I get the sense that he’s caught in his own prolonged twilight. Coming out of it before turning back in. Lapsing and relapsing.
Recently, he’s lost himself to dreaming, and this might be his most interesting vice. From what I can tell, his dreams are not regular, in any sense of the word. Maybe the drugs he uses have banished that realm of experience. Maybe the medicines that abuse him have burned away that particular state of consciousness.
A man cannot live without dreams. If he cannot get them from his own neurochemistry, a synthetic is in order. For entire nights, my friend foregoes one reality in favor of another. He stays there even after the eyecups have worn red along his brow, the light scattered blue across his cheekbones. Glow and counterglow.
Sometimes he sends me recordings from inside the headset. From my screen, I can see what he sees — if not how he sees it. It’s almost always an interior — a low-ceilinged lounge, the cabin of a private jet — with dim lighting and polygonal forms. There’s some latency in his movements and a weakened sense of object permanence. He turns to someone, turns away from them for a moment, and turns back to find them not quite where he left them. Someone flits into view, gestures anxiously, and drops away. Every now and then, he tells me, there is a chatter and a jolt. The layers of this experience — computer, screen, eyes, drug, brain — fall out of sync and snap back, explode and collapse. Although it’s never really acknowledged, most of the people seem to be in drag. He tells me that one lounge is especially popular. It’s called The Black Cat. Le Chat Noir. Quite the cabaret.
My friend has spent a long time swinging between the strong dreaming problem and the weak dreaming problem. Now it seems that he’s dampened these oscillations by resigning himself to a firm dreaming problem. Instead of struggling to differentiate between reality and illusion or between waking and sleeping, he dedicates himself to the appraisal of a thoroughly unconvincing counterfeit. It’s a fake, yes, but at least he knows it. He trades the possibility of an evil demon for the certainty of a benign one, plops his brain into a vat of his own design. Philosophically, this strategy won’t get him anywhere. But psychologically, this approach seems to keep him where he needs to be. Impoverished, but achingly so.
***
These hypnoses have broken containment. The ability to drag your dreams into the daylight quickly becomes the capacity to foist them on others, the power to loose them on the world. Prompt. Package. Propagate. This is what they mean when they say “synthetic media”. This is deepfakes, bots, and dead internets. This is the oneirocracy, the era of dream rule.
After your grandfather dies, your grandmother starts getting voicemails that sound just like him, taunting her into some elaborate scam. When your sister goes on vacation to Cancun, your mother wakes up to a ransom call from a local number, her daughter pleading in the background. Then your boss invites your coworker to a video call, before instructing her to wire company funds to a bank account she doesn’t recognize. One day you open a strange email to find deepfake porn of yourself and a credible threat to send it to almost everyone you know. A web of hauntings, fears, humiliations, and feints — this is the stuff of nightmares.
The new dreaming problem is that what you perceive as your own dream might actually be someone else’s. This is a political problem familiar to anyone who’s seen Inception. The primary setting — whatever you take that to mean — is layered and shifting. Some will squander it in addiction. Others will leverage it as manipulation. One man’s exploration becomes another man’s exploitation. Catharsis is offered and deferred. It all washes up.
Do not confuse an unsolved problem with an unsolvable problem, and do not delude yourself into believing that this might be the former. After all, this a problem of delusion and belief. It’s very hard to keep someone where they don’t want to be. All minds wander, and some of them get lost. The best you can do is try to stay. Take comfort in the promises made of old men. Rely on the testimony of schizophrenics. Pinch your nose and try to breathe.