— A version of this essay was originally published, under the title “What Are Images Now?” in our Spring 2026 issue, “Everything’s Computer.” Order your glossy print copy or an instantly downloadable e-paper from our online shop —

Before Adam and Eve disobey God and eat from the Tree of Knowledge, they are described as naked, but not ashamed. Immediately after, they hide their nakedness, suddenly self-aware and self-conscious. Genesis 3:7: “The eyes of them both were opened, and they knew that they were naked; and they sewed fig leaves together, and made themselves aprons.” Then, they hide from God. Genesis 3:9: “And the Lord God called unto Adam, and said unto him, Where art thou? And he said, I heard thy voice in the garden, and I was afraid, because I was naked; and I hid myself.”

Knowledge, it appears, requires separation. The apple, it seems, was laced with images; most notably, that of the self.

The myth of The Fall is a fabulist articulation of what psychoanalyst Jacques Lacan called the Mirror Stage, the moments a toddler begins to recognize his reflection. Lacan conceived of the Mirror Stage as “formative of the ‘I’ function,” the “transformation that takes place in the subject when he assumes an image.” But what Lacan makes sure to highlight is that this process is a “misrecognition.” The child – an incoherent reality – finds a home for his ego in his image – a coherent fiction – resulting in a “discordance” between those two selves. A Buddhist might argue it is out of this sort of discord that all human suffering is born, either as “aversion” to what is, or “craving” for what is not – a.k.a. images of what is. (I am using the word images to include all representation, internal and external, partially because that’s how I understand its meaning in the Second Commandment, which prohibits the worship of “any graven image.")

But Lacan saw the Mirror Stage as unavoidable and necessary. The writers of Genesis seemed to agree; as many theologians and theorists have pointed out, was it not God who put the Tree there in the first place? And Eve’s desire to eat the fruit doesn’t arise until it’s explicitly labeled forbidden. God’s warning sounds a bit like the character in a horror movie who, after regaling the teenagers with titillating tales of a haunted house, says, “But don’t go in there.” God’s order has the undertone of a dare.

This is all to say that our unique relationship to images is, in large part, what makes us human. It has been our capacity to break a reality too big and dynamic to swallow whole into chewable pieces – images – and to conceive of ourselves as one of those pieces – self-image – that has allowed us to manipulate our environment and build complex societies. This fracturing of reality is, of course, a “lossy” process, but as legendary statistician George Box said, “All models are wrong. Some are useful.” Images, in a sense, were the ur technology, The Fall the ur devil’s bargain. We lost oneness, unity, and our home, but gained the world.

Now, here we are: that world is being flooded with a new strain of external images that look very much like it but bear no direct relationship to it. For two centuries, photographs offered that same quality of uncanny likeness but paired it with indexicality and a relative truth claim that – though flawed and deceptive from the start – made them, whether still or moving, immensely useful. Photographs are, after all, our ontological relatives. Very distant relatives, yes, but still born in the same world we are, the realm of light. (Broken down to its Greek roots, “photograph” means “drawing with light.”) But AI images are aliens in disguise. They’re born of another realm entirely: the black-boxed black hole of data-center darkness.

It seems the ever-growing distance between images and the referent known as reality has finally reached a width that renders the tether effectively severed. That bifurcation renders an Image World that is free to develop, grow, and function unconstrained by the material world. That absence of constraint renders images, no matter how “realistic,” less informationally congruent to everyday life. And that incongruence renders the entire medium more suited for its historically non-photographic roles: art, entertainment, ritual, etc.

In a way, the strangeness of our times could be understood as the result of reality struggling to keep up with this increasingly alien, independent Image World, our dysfunction the result of flesh and blood, photons and atoms, being and self, painfully stretching to stay connected to images we don’t yet realize have left us behind entirely. And so, fueled by dissatisfaction with material reality, humanity’s Image Drive – our ancient desire to live in and as images – has been thrown into hyperdrive. Still functioning as if our relationship to images is photographic, we’re left chasing ghosts.

Of course, much of society is built atop that photographic dynamic, so systems are scrambling to maintain it. The Associated Press, Reuters, and others are installing technologies and protocols that can verify a photograph’s nature and provenance. Dating apps are using facial-scanning technologies to prevent catfishing. Real estate and e-commerce sites are integrating AI-detection tools. But regardless of the effectiveness of these efforts, our media environment is increasingly characterized by the plausible deniability of everything. Legal scholars Bobby Chesney and Danielle Citron have dubbed this a “liars dividend”; we’re just a few years into mainstream AI, and in courts of law around the world, the burden of proof has already shifted onto images.

In short, the medium that defined and shaped much of modernity is becoming quite useless.

Useless – this is the word Oscar Wilde used to defend art from society’s utilitarian biases in his preface to The Picture of Dorian Gray (1890), concluding, “All art is quite useless.” It’s also the word Lacan uses that feels most relevant to the arrival of mainstream AI. More specifically, Lacan describes the toddler recognizing his reflection as an act of acquiring “control over the uselessness of the image.” In theory, before the Mirror Stage, the child’s reflection is useless to him because it’s meaningless, just shapes, light, and color. Then, once his brain develops enough from him to notice that, when he waves his hand, the image waves, too, he begins to identify with it, making the image meaningful and, therefore, useful.

I would argue, though, that from the perspective of a newborn baby, his reflection is actually neither useless nor useful, as it has no more or less significance than anything else. The image’s “uselessness,” then, exists only for observing adults. Mapping this slightly amended trajectory onto the myth of The Fall: 1) A prelapsarian state – life in the Garden – is imageless and unified; images are neither useless nor useful. 2) A postlapsarian state – the “human condition” – is image-saturated and fragmented; images are useful. 3) …

Third States – the dangling carrots of psycho-spiritual freedom promised by the proverbial analyst, prophet, or philosopher. Consider the theological hero’s journey, its structure usually something like this: 1) A prophet is born into image-saturated society just like the rest of us – think: the Buddha starting out a prince, or Muhammad a merchant. 2) The prophet grows disillusioned and leaves society – think: Jesus or Moses’s retreats to the desert, though a cave, jungle, or mountain will do, any arena suitable for imageless asceticism. 3) The prophet returns to society, but can now see it clearly because he can see through its layer of appearances, images now functioning more like a window than, say, like a photo.

There are many formulations of this tri-part sequence, many echoes of it in both religious and secular spheres: 1) Thesis, 2) Antithesis, 3) Synthesis; 1) Structure, 2) Liminality, 3) Communitas. 1) Life, 2) Death, 3) Resurrection. But usually these Third States are characterized by a newfound, integrated freedom, a capacity to live among society’s images without being controlled by them. It is in these elusive Third States that images are – theoretically, for the first time in one’s life – both ubiquitous and useless.

What does that mean for our current state? Our hyper-mediated, hybridized reality where life can feel a lot like the Internet Live, an IRL URL? Third States, appropriately, require third ways, neither less nor more, neither backward nor forward, but an outside-the-frame approach. That would make the path not one of fewer images, nor iconoclasm or luddism to get us back to the garden; not more images, nor iconolatry or technicism to bring us to some rapturous future. Those are political modes, business as usual.

Given the unlikelihood of a society-wide forty days and forty nights of imageless asceticism – essentially, the opposite of what occurred under Covid lockdown – maybe the way to freedom from, but still with, images would require some sort of mass-neutering of the medium, a defanging that left images still able to bite, but unable to hold, a shift that rendered them a little weakened, isolated, and – back to Wilde – useless. Maybe the closest thing the doomscrolling laity could get to some sort of practical, everyday realization of Enlightenment, Nirvana, Moksha, Janna, Olam Ha-Ba, or the consummation of the Kingdom of Heaven, would not be the end of images, but the end of the reign of images. More to the point, maybe it would require a development not so terribly unlike that which AI seems to be instantiating.

From what I can tell, it has been revolutions in communications mediums, like the internet, not technologies that populate those mediums, like LLMs, that lead to mass death and suffering. AI is often compared to nuclear weapons as a way to comprehend its existential dangers and how it might be regulated. That analog may be useful in a counterintuitive sense: approximately two hundred thousand people have been killed by nuclear weapons in history; 70-85 million people died by other means in the rest of WWII, a conflict inextricably tied to the advent of the radio.

The rise of that generation’s cohort of authoritarians, the spread and fervor of their combustive ideologies, was fueled by an access to citizens’ living rooms that was new in level and kind. In his speech “Radio as the Eighth Great Power,” Nazi Minister of Propaganda Joseph Goebbels said, “It would not have been possible for us to take power or to use it in the ways we have without the radio.” The party distributed Volksempfänger (“The People’s Receivers”), increasing radio ownership from 20% to 75% in just a few years – a quieter operation than that conducted over Hiroshima and Nagasaki, but exponentially deadlier.

Four centuries earlier, when Gutenberg's press and its processes reached ubiquity in Europe, the continent fell into a protracted period of chaos and conflict, beginning with the German Peasants’ War in 1524. Humanity’s first-ever experiment with widespread pamphleteering didn’t go well: 12 million dead over a century of fighting, and famine so bad that cannibalism was excused. It was in this period that the word propaganda was coined, from propagare, meaning, “To spread.”

All to say, I don’t worry about AI. I worry about the internet, the medium without which AI neither exists nor poses an existential threat. And though I loathe accelerationisms or teleologies that leave the realm of thought exercise, and though there’s little I fear more than a potentially painful transition out of the reign of images, coexisting with my anxieties is a feeling I suspect has consciously or unconsciously nagged most people: the sense that, though mainstream AI has just arrived, we actually outgrew our mode of relating to images when the internet arrived; that. if today’s situation feels frighteningly new, it’s because AI pushed us to the necessary breaking point of an unsustainable period of semiotic limbo.

The hope, then, is that AI could force our senses of self, our anchors of meaning, and our notions of truth, like crabs whose shells so clearly no longer fit, out of our mirrors and screens. That is, the hope is that the New Uselessness might allow us to at least partially divest from the stable, finite, and knowable realm of images, and come to rest on the unstable, infinite, and unknowable ground of reality.

Just before the famous bombastic final line of his preface, Wilde writes, “We can forgive a man for making a useful thing as long as he does not admire it. The only excuse for making a useless thing is that one admires it intensely.” AI is, of course, not just threatening to make images quite useless but human beings – made in “the image of God” – useless, too. A terrifying prospect, unless one uses the word as Wilde does.