If someone wanted to design a technology capable of dismantling social institutions faster than corruption, structural fatigue, and quiet resignation combined, they would have to invent something very much like today’s artificial intelligence.
Not the romantic fantasy of a robot that brings you tea and solves your Sudoku. I mean the real AI: a statistical confidence machine that delivers answers with the swagger of an expert, even when it has no idea what’s outside the window.
Institutions — those old, boring, but vital organisms — run on slowness, conflict, accountability, and human stubbornness. AI runs on the opposite. That contrast is its elegance and its destructive power. It’s like placing a MIDI‑playing automaton into a philharmonic orchestra and then acting surprised when the violinists quit.
I. Expertise as the First Casualty
Institutions rest on expertise. Not the paper kind, but the kind forged through years of practice, mistakes, arguments, night shifts, and awkward meetings.
AI replaces expertise with the feeling of expertise. That’s a crucial difference.
AI doesn’t know. AI performs knowing.
And people — tired, overloaded, underfunded — are happy to believe it.
“Why should I think, when the model can think for me?”
And so the doctor becomes a click‑through operator, the teacher becomes a prompt shaman, and the journalist becomes someone who checks whether the model just invented a quote.
Expertise doesn’t disappear overnight. It atrophies.
And institutions atrophy with it.
II. Decision‑Making Without Decision‑Makers
Institutions are slow because they must be slow.
Decision‑making is a ritual: arguments, conflict, responsibility, signatures, consequences.
AI compresses that ritual into a single click.
And with it, responsibility evaporates.
When a human decides, we can ask: Why?
When an algorithm decides, we ask: How?
And the answer is: “It’s complicated.”
End of discussion.
AI creates a haven for cowardice.
No one knows anything, no one is accountable, but everyone pretends the system works.
Institutions lose their core function: being a place where someone must stand face‑to‑face with consequences.
III. Isolation as the New Default
Institutions aren’t just structures. They’re people in rooms.
Arguments in hallways.
Coffee in the break room.
The sense that we’re in this together.
AI replaces this social metabolism with a personalised monologue.
Everyone gets their own digital companion — one that never contradicts, never gets offended, never challenges.
A relationship without friction — and therefore without growth.
Institutions crumble when people stop being together.
AI offers a comfortable alternative: being alone, while feeling like you’re not alone.
That’s the most dangerous form of isolation.
IV. The Execution List of Institutions
Law
AI thrives on opacity.
Law thrives on transparency.
This marriage ends in a quick divorce.
Universities
AI is a machine for the average.
Universities are machines for originality.
Guess who wins.
Journalism
AI generates content.
A lot of content.
So much so that truth becomes a footnote lost in the noise.
Democracy
Democracy is built on trust.
AI is built on probability.
That combination won’t survive a single election cycle.
AI as Metastasis
AI isn’t just a tool.
It’s a process that rewrites the conditions under which institutions survive.
Institutions need:
- slowness
- conflict
- accountability
- human stubbornness
- shared space
AI offers:
- speed
- smoothness
- delegation
- illusion
- isolation
These aren’t compatible.
They’re antagonistic.
And if we let this run on autopilot, we’ll end up in a world where institutions exist only as stage props — and behind them hums a quiet, efficient, unaccountable machine.
Podle How AI Destroys Institutions by Woodrow Hart