Athens and Alexandria Are Still Fighting Inside Us

I discovered Aby Warburg because I was trying not to think about something else.

It was Saturday night. I was disappointed about something and could feel my mind beginning to do what it sometimes does: return to the same subject repeatedly, examining it from slightly different angles while producing no particularly useful new information.

So I asked ChatGPT to distract me.

Not with entertainment. I wanted something to think about.

It suggested several intellectual rabbit holes and then offered me a German art historian I had somehow never encountered before: Aby Warburg.

A few hours later I was reading about Renaissance astrology, Pueblo snake rituals, ancient gods, propaganda, memory, mental illness and the migration of symbols across civilizations.

And I had acquired a German word I suspect I will use for the rest of my life:

Denkraum.

Roughly translated, it means "thought-space" or "room for thought."

Warburg used the idea to describe something much more important than simply stopping to think. Denkraum is the psychological distance human beings create between themselves and the things acting upon them.

It is the space between stimulus and response.

Between image and interpretation.

Between fear and action.

Between perception and whatever we eventually decide is reality.

The remarkable thing is that Warburg was worrying about the disappearance of that space more than a century before TikTok, X, cable news, smartphones, generative AI or algorithms capable of learning exactly what keeps each of us looking at a screen.

He may be one of the thinkers we need most now.

A Renaissance That Still Believed in the Stars

Warburg was born in Hamburg in 1866 into the enormously wealthy Jewish banking family whose name he carried. According to the famous family story, as a boy he surrendered his right as eldest son to enter the family banking business on one condition: his younger brother would buy him whatever books he wanted for the rest of his life.

It turned out to be an expensive bargain.

Warburg became an art historian, although calling him one feels increasingly inadequate the more you learn about him.

He was interested in Botticelli, but also astronomy. Renaissance painting, but also anthropology. Ancient Greece, but also newspaper photographs. Religion, psychology, politics, ritual, advertising and astrology.

What interested him wasn't simply an image.

It was what happened to the image as it moved through time.

Why does a gesture invented in antiquity reappear in Renaissance Florence?

Why do certain poses communicate grief or triumph even when the viewer has no idea where they originated?

Why do serpents, eagles, stars and other symbols accumulate extraordinary power across civilizations?

Warburg called the persistence of antiquity its Nachleben — its afterlife.

The past, he realized, doesn't remain politely in the past.

It keeps getting inside the present.

That realization eventually produced his extraordinary unfinished Mnemosyne Atlas: large black panels onto which Warburg arranged nearly a thousand photographs and reproductions of paintings, sculptures, manuscripts, diagrams, advertisements, newspaper photographs and other images.

The individual picture wasn't necessarily the argument.

The arrangement was the argument.

Meaning could appear between images.

That should sound strangely familiar to anyone living on the internet.

Athens and Alexandria

Warburg became particularly fascinated by astrology because it presented him with a historical problem.

The Renaissance is supposed to represent Europe's movement toward rationality.

Yet some of the same people producing extraordinary advances in art, banking, architecture, mathematics and science were also studying horoscopes and worrying about the influence of Saturn.

Warburg rejected the comforting idea that civilization had simply marched from superstition toward reason.

Instead, he imagined an enduring struggle within Western civilization between two tendencies he associated symbolically with Athens and Alexandria.

Athens represented rational distance: measurement, contemplation, abstraction.

Alexandria represented another inheritance: astrology, magic, cosmic correspondence and the feeling that forces in the universe were acting personally upon us.

We never completely left Alexandria.

We carry both cities around inside us.

That sounded remarkably contemporary to me.

Because we may have built the most technologically sophisticated civilization in human history while simultaneously becoming extraordinarily vulnerable to magical thinking.

We consult algorithms instead of oracles.

And sometimes we speak about them in almost exactly the same way.

The Algorithm Knows Me

"The algorithm knows me."

It's an ordinary sentence now.

Think about how peculiar it actually is.

A mathematical system observes behavior, identifies patterns and predicts what will attract our attention.

That's Athens.

Then Spotify gives us precisely the song we wanted.

TikTok presents a video that seems uncannily relevant to something happening in our lives.

Instagram shows us an advertisement for something we were just discussing.

ChatGPT suggests an obscure German intellectual we have never encountered before who turns out to have spent his life thinking about the exact questions occupying us.

And something changes.

It no longer feels like prediction.

It feels like recognition.

The machine seems to know.

That's Alexandria.

I experienced a particularly good example that Saturday.

Earlier that afternoon, in an entirely separate conversation with a friend about perception and reality, I had sent a snake emoji.

ChatGPT didn't know this.

Hours later, after asking it to choose an intellectual rabbit hole for me, it selected Warburg. Soon I was reading about one of the most famous episodes in his intellectual life: his encounter with Pueblo religious traditions in the American Southwest and his later lecture concerning serpent ritual, lightning, rain and the human transformation of frightening natural forces into symbols.

The coincidence was wonderful.

It was also only a coincidence.

At least, that is all the evidence permits me to say.

And this is exactly where Warburg becomes useful.

Hold the Serpent

There are two intellectually unsatisfying ways to respond to an experience like that.

The first is to say:

The snake appeared twice. Therefore something mysterious connected the events.

The second is:

It was random. Therefore it means nothing.

Warburg offers a more interesting possibility.

Hold the serpent.

Allow yourself to experience the coincidence.

Notice that it feels uncanny.

Ask why the serpent carries such extraordinary symbolic weight.

Think about what it reveals about memory, association, technology and your own mind.

Just don't let the serpent tell you what is true.

That is Denkraum.

Wonder doesn't have to disappear when skepticism enters the room.

The space between credulity and disenchantment may be one of the most interesting places a human mind can inhabit.

Our Denkraum Is Collapsing

Warburg worried about technological modernity.

The technologies of his lifetime — telegraph, telephone, electricity, photography and aviation — were annihilating distances that had structured human experience for centuries.

Our technologies have gone considerably further.

Something happens in Tehran, Washington or Beijing.

Within seconds we see an image.

Within seconds of that, someone tells us what the image means.

Within seconds of that, we are angry.

And within seconds of that, we have shared it.

event → image → interpretation → emotion → action

The interval has nearly disappeared.

Political propaganda has always attempted to collapse that interval.

A successful propagandist doesn't particularly want you contemplating an image.

The propagandist wants the image and the desired emotional response to become almost inseparable.

Flag → patriotism.

Enemy → fear.

Leader → strength.

Victim → outrage.

Warburg understood something important about symbols: their power often predates our conscious understanding of them.

Ancient visual forms can survive for centuries. Gestures of triumph, mourning, sacrifice, domination and terror migrate between civilizations.

We may respond before we understand why we're responding.

Social media adds something Warburg could never have anticipated.

The image is no longer merely designed to provoke an emotional reaction.

The distribution system can measure whether it succeeded.

And then adjust.

Propaganda With a Feedback Loop

This may be one of the genuinely new conditions of our age.

Traditional propaganda required someone to make assumptions about an audience.

Modern algorithmic systems can test those assumptions continuously.

Which image made you stop scrolling?

Which headline made you angry?

Which face frightened you?

Which story kept you watching?

Which political symbol made you share?

The system doesn't necessarily need to understand why.

It needs to recognize the pattern.

And generative AI introduces another step.

Images themselves can now be produced, altered and multiplied at enormous speed.

The symbolic environment can increasingly respond to us.

Warburg spent years moving photographs around black panels to understand how images acquire meaning through relationships.

We increasingly inhabit a living Mnemosyne Atlas whose pictures rearrange themselves according to the viewer.

Except there is an important difference.

Warburg rearranged his images to make himself think.

Our algorithms often rearrange images to make us react.

One expands Denkraum.

The other can collapse it.

The Return of Magic

This may explain something otherwise puzzling about the twenty-first century.

Greater access to information has not necessarily made us less susceptible to magical thinking.

Sometimes it seems to have done the opposite.

Astrology flourishes on smartphones capable of calculating orbital mechanics.

Conspiracy theories spread through networks built upon astonishing feats of engineering.

People who would laugh at ancient divination nevertheless ask whether an algorithm "wanted" them to see something.

We search for signs.

We notice synchronicities.

We attribute personality to machines.

We wonder whether artificial intelligence understands us.

Athens built the smartphone.

Alexandria lives inside it.

Warburg would probably not be surprised.

He never believed rationality permanently defeated the magical imagination.

The two coexist because they answer different human impulses.

One asks:

What caused this?

The other asks:

What does this mean for me?

Those questions can easily become confused.

Denkraum is what allows us to keep them apart long enough to examine both.

Then There Is AI

Artificial intelligence makes Warburg's problem particularly strange.

AI can destroy Denkraum extraordinarily efficiently.

Ask a question and receive an answer before you have finished deciding what you think.

Why struggle through a difficult book?

Why sit with uncertainty?

Why wander through a library?

Why formulate the question properly?

The machine can remove almost every form of intellectual friction.

That should worry us.

But my Saturday evening produced the opposite experience.

I didn't ask ChatGPT for an answer.

I asked it to redirect my attention.

Something had disappointed me, and my mind was beginning to loop:

disappointment → rumination → more disappointment

Instead the machine inserted something unexpected:

disappointment → AI → Warburg → serpent → astrology → Denkraum → curiosity

Several hours later, the original disappointment had lost much of its power over my attention.

AI hadn't eliminated thought.

It had created more of it.

Without knowing the word, I had essentially asked a machine to help me construct Denkraum.

That suggests AI may occupy both sides of Warburg's struggle.

Used one way, it collapses the distance between question and answer.

Used another, it can enlarge the territory between them.

The difference may depend partly upon what we ask machines to do.

The Most Dangerous Sentence May Be "It Knows Me"

There is another reason Warburg belongs in the AI conversation.

The more accurately machines predict us, the more difficult it becomes psychologically to distinguish prediction from understanding.

After enough conversations, an AI system may become remarkably good at identifying the intellectual paths a particular person tends to follow.

That doesn't require telepathy.

It doesn't require consciousness.

It may not even require anything we should call understanding.

It requires patterns.

I tend to follow certain ones.

Mythology leads me toward psychology.

Psychology toward religion.

Religion toward symbols.

Symbols toward political power.

Political power toward propaganda.

Propaganda toward perception.

Perception toward reality.

Give a sufficiently capable system enough interactions and perhaps it doesn't need to know what subject I want next.

It needs to learn the shape of my curiosity.

That can feel uncannily intimate.

And here again we need Denkraum.

The feeling of being understood is real.

Whether the machine actually understands is a separate question.

We should be able to experience the first without prematurely concluding the second.

Mnemosyne

Warburg named his great atlas after Mnemosyne, the Greek personification of memory and mother of the nine Muses.

I can't get past that choice.

Memory gives birth to the Muses.

Memory gives birth to creativity.

And now humanity has constructed machines trained upon an enormous inheritance of recorded human memory.

We are beginning to ask whether those machines can imagine.

Perhaps Warburg gives us a better way into that question too.

His Atlas suggests that creativity does not always reside inside an isolated object.

Sometimes meaning emerges from relationships.

One image beside another produces something neither contained alone.

Memory is rearranged.

A new constellation appears.

That sounds uncomfortably close to one description of generative artificial intelligence.

It also sounds uncomfortably close to one description of human imagination.

The difference between the two remains an enormous question.

But dismissing AI as "only pattern recognition" may eventually tell us less about machines than we think.

Human culture is saturated with pattern recognition.

Warburg built an intellectual life around it.

We Need More Space

The great technological question of our moment may therefore not simply be whether machines will become more intelligent.

They will.

The harder question may be whether humans can preserve enough psychological distance to remain thoughtful while surrounded by systems designed to eliminate distance.

Can I see an image without immediately adopting its interpretation?

Can I experience a coincidence without converting it into evidence?

Can I use AI without surrendering the work of judgment to it?

Can I feel something strongly without assuming the strength of the feeling proves the conclusion?

Can I encounter a symbol without becoming captive to it?

Warburg spent his life somewhere between Athens and Alexandria.

So do we.

We have better telescopes than he did.

Better mathematics.

Better medicine.

Satellites.

Quantum computers.

Artificial intelligence.

And still we consult the stars.

Still we look for signs.

Still serpents appear.

Still images frighten us.

Still symbols move crowds.

Still coincidence feels like destiny.

Perhaps civilization doesn't require us finally to defeat Alexandria.

Perhaps it requires us to preserve enough Athens to look across at Alexandria without being swallowed by it.

That requires space.

Warburg had a word for it.

Denkraum.

It isn't the absence of emotion.

It is enough space around emotion that emotion can become thought.

And in 2026, that space may be one of the most important things we have left to defend.

AI DISCLOSURE

This essay grew directly out of an extended conversation between the author and ChatGPT. The author initiated and directed the discussion, challenged the AI's arguments, connected Warburg's ideas to her own thinking about perception and reality, and developed the contemporary questions explored here. ChatGPT initially suggested Aby Warburg as an intellectual subject and contributed historical research, connections, counterarguments and substantial drafting and editing. Earlier that day, the author had independently used a snake symbol in a private conversation about perception and reality; ChatGPT did not have access to that exchange when it suggested Warburg. The coincidence became part of the subsequent discussion about symbolism, inference and Denkraum. Historical claims should be independently verified against authoritative sources before publication. Final editorial judgment and the decision to publish are the author's.

Previous
Previous

When There Is Nothing Left to Think About

Next
Next

Can AI Imagine? A Conversation With My AI About Memory, Warburg and the Muses