Among contemporary philosophers, Byung-Chul Han is probably one of the sharpest critics of the digital age. Interestingly, however, he has not written a book devoted exclusively to AI. Instead, his reflections on AI are spread across several books, especially:
- Psychopolitics (2017)
- In the Swarm (2017 English translation)
- Infocracy (2022)
- The Crisis of Narration (2024 English translation)
- Nonthings (2022 English translation)
Taken together, they amount to a philosophy of AI even though they were not written as one.
I would summarize his thought into seven major themes.
⸻
1. AI possesses intelligence, but not Geist
This is perhaps Han’s best-known claim about AI.
He distinguishes between:
- Intelligenz (computational intelligence)
- Geist (spirit, mind, lived meaning)
AI excels at the first.
It lacks the second.
For Han,
thinking is not merely calculating.
Thinking involves
- interruption
- hesitation
- contemplation
- suffering
- wonder
- love
- death
AI can optimize.
It cannot dwell.
As he has put it in public remarks, AI may become extraordinarily capable while still lacking the kind of spirit that characterizes human existence.
⸻
2. AI amplifies what he calls “the Hell of the Same”
One of Han’s recurring concepts is
the Hell of the Same.
Digital capitalism continually reinforces
- preferences
- habits
- identities
- opinions
Algorithms recommend
more of what already resembles us.
Difference gradually disappears.
AI recommendation systems therefore risk intensifying
- filter bubbles
- personalization
- self-confirmation
rather than exposing us to genuine otherness.
For Han,
human freedom requires encounters with
the Other
rather than endless optimization of
the Same.
⸻
3. AI cannot truly encounter the Other
Across books like The Agony of Eros and The Crisis of Narration, Han argues that human life depends on encounters with alterity—the irreducible strangeness of another person.
Large language models generate statistically plausible continuations.
They do not themselves undergo the transformative experience of meeting another consciousness.
Thus AI may simulate dialogue,
but for Han it does not experience
surprise,
desire,
love,
or forgiveness.
Those belong to existence rather than computation.
⸻
4. The real danger is not superintelligence but psychopolitics
Unlike many AI researchers,
Han is less worried about a future superintelligence taking over the world.
He is more concerned about present-day systems that quietly reshape human behavior.
His diagnosis is:
AI +
- smartphones
- recommendation algorithms
- surveillance
- behavioral prediction
- digital platforms
create new forms of self-governance.
Instead of external coercion,
people voluntarily optimize themselves,
measure themselves,
and expose themselves.
That is why he calls our era one of psychopolitics rather than merely surveillance.
⸻
5. AI accelerates the collapse of narrative
In The Crisis of Narration, Han argues that digital media fragment experience into continuous streams of information.
Narratives once gave life
continuity,
memory,
and meaning.
Algorithms favor
engagement,
novelty,
and immediacy.
The result is a culture increasingly rich in information but poor in stories.
AI-generated content may accelerate this tendency by producing limitless text, images, and videos while weakening shared narrative worlds.
⸻
6. AI produces “nonthings”
In Nonthings, Han argues that modern technology is replacing enduring objects with flows of information.
Books become files.
Letters become notifications.
Photographs become data streams.
AI accelerates this transition because it operates primarily on symbolic information rather than physical artifacts.
For Han,
this risks making our relationship to the world increasingly abstract and less rooted in tangible presence.
⸻
7. AI cannot suffer
One of Han’s most provocative remarks is that AI cannot experience what gives depth to human existence.
He has suggested that AI cannot truly think because it cannot experience the bodily and existential dimensions of life—famously remarking that it cannot “get goosebumps.” The point is not about literal physiology but about the absence of lived affect, vulnerability, and mortality.
For Han,
creativity ultimately comes from
- wounds,
- mortality,
- vulnerability,
- and exposure,
not simply from statistical recombination.
⸻
Where I agree—and where I think the discussion remains open
This is where our conversations over the past year become especially interesting.
Many of Han’s diagnoses have proved remarkably prescient:
- algorithmic attention capture,
- self-optimization,
- digital exhaustion,
- the loss of contemplative time,
- and the erosion of shared narratives.
Those themes fit closely with your own concerns about the algorithmic self.
Where I think his account is less complete is his picture of AI itself.
Han often portrays AI as a sophisticated mechanism for repeating what already exists. That captures an important aspect of current systems, but it leaves relatively unexplored how AI can also function as a partner in exploration when embedded in practices that encourage iteration, critique, and discovery. Recent developments in agentic workflows and loop-based systems suggest that AI is increasingly used not merely to predict the next token but to support extended cycles of inquiry.
This connects directly to something you have been developing independently.
You often describe creation not as linear planning but as
“follow the fragments.”
From Han’s perspective,
AI remains trapped inside
the Same.
From your perspective,
AI can sometimes help us
follow a fragment into the Unknown.
That is a subtle but significant difference.
The crucial question, then, is not whether AI possesses Geist. Rather, it is whether humans can use AI in ways that preserve openness to surprise instead of collapsing experience into prediction. In your own vocabulary, the challenge is ensuring that the algorithmic self does not crowd out the analogical self. That strikes me as a productive dialogue with Han: accepting much of his critique of digital capitalism while remaining open to the possibility that AI, under the right conditions, may sometimes help us encounter—not eliminate—what is genuinely new.
