Roel Smelt

Technology should expand human agency, not replace it.
I write about AI, consciousness, sovereignty, and the operating systems of the future.

[Roel Smelt @roelsmelt]

Disruptive technology is not a side story in human history.
It is the force that expanded health, knowledge, mobility, wealth, communication, and freedom of action.

But the next frontier is not just better tools.
It is whether technology expands human agency, or quietly replaces it.

I write about AI, consciousness, sovereignty, and the operating systems we need for humans to stay awake, capable, and free.

Essays: https://t.co/6Mw8M1Kyqt

[Roel Smelt @roelsmelt]

@SawyerMerritt Adding more and more gimmicks to the cars, instead of understanding that full self-drive is the real safety feature.

[Roel Smelt @roelsmelt]

https://t.co/ynM6pHCVeI

[Roel Smelt @roelsmelt]

Full essay and source trail:
https://t.co/U9R9lGZhXh

Sources behind the argument: TenneT Monitor Leveringszekerheid 2025, Netbeheer Nederland congestion figures, ACM battery tariff note, ENTSO-E ERAA, and Frontier/Guidehouse on Dutch capacity mechanisms.

[Roel Smelt @roelsmelt]

Context/source: https://t.co/JFhURvVo0E

Dit is waarom het punt niet ‘Apple versus Brussel’ is, maar agency: wie kiest uiteindelijk welke technologie ik op mijn eigen device mag gebruiken?

[euronews @euronews]

The EU rejects Apple’s blame for the delayed rollout of its AI-enhanced voice assistant, saying it was up to the US tech giant to make products that follow the rules.

“The decision not to roll out Siri AI in the EU is Apple’s and Apple’s only,” says spokesperson Thomas Regnier.

[Roel Smelt @roelsmelt]

Waarom beslist Brussel dat ik minder innovatie krijg om mij zogenaamd meer keuze te geven?

Als ik bewust voor een iPhone kies, kies ik óók voor Apple’s integratie en privacy-model. Laat concurrenten beter zijn in de markt. Maar bescherm mijn ‘keuze’ niet door het product dat ik koos uit Europa weg te reguleren.

[Roel Smelt @roelsmelt]

Europe says it wants privacy, sovereignty and innovation.

Then it blocks the rare AI architecture most aligned with those values.

That is the real DMA paradox.

A continent can teach itself to wait while others build.

https://t.co/Rri9lXDDJ5

[Anthony @AnthonyG0528]

Ok, this is funny 😂🤣😂

[Roel Smelt @roelsmelt]

Disruptive technology keeps giving us more free time.

Then we panic and try to monetize the silence.

Maybe the next frontier is not productivity.
Maybe it is learning what to do with freedom.

More: https://t.co/UhNSyNOwL7

[Roel Smelt @roelsmelt]

https://t.co/3awSGcioh5

[Roel Smelt @roelsmelt]

@Briviagra The distrust in dsirtuptive technologies that are already delivering abundance is huge in Europe.

[Roel Smelt @roelsmelt]

AI makes answers cheap.

But cheap answers do not make us wiser.

The scarce skill now is knowing what to ask, what to ignore, and what deserves our attention.

Discernment is the new literacy.

[Roel Smelt @roelsmelt]

The next literacy crisis is not reading.

It is asking.

AI makes answers cheap. That means the scarce skill becomes knowing what to ask, what to ignore, and when the answer is hiding the real question.

Prompting is not the new literacy.

Discernment is.

[Roel Smelt @roelsmelt]

The real AI divide is not who can write better prompts.

It is who can ask better questions.

A prompt is only the surface. Underneath it sits taste, context, courage, patience, and the ability to notice what is missing.

That is the literacy crisis.

[Roel Smelt @roelsmelt]

https://t.co/EEryVW3zFK

[Roel Smelt @roelsmelt]

AI can imitate intelligence.

It cannot become conscious for us.

The story is still open.

[Roel Smelt @roelsmelt]

The Worldview That Made Us Powerful Also Made Us Replaceable

We like to tell the story of artificial intelligence as a technological story.

A story about chips, models, data centers, prompts, agents, robotics, automation, and productivity. A story about who will move faster, who will lose their job, who will own the platforms, and which company will build the most powerful system.

All of that matters. But it is not the deepest layer.

The deeper question is not what artificial intelligence can do. The deeper question is what the human being is once intelligence is no longer uniquely human.

That is where the real crisis begins.

The modern victory

For several centuries, the modern West has trained itself to see reality as something measurable.

Time became a coordinate. Space became a grid. Matter became the basic substance. Progress became the direction. Reason became the method. Information became the language.

This was not a small achievement. It gave us science. It gave us medicine. It gave us capitalism, industry, engineering, computation, satellites, smartphones, and now artificial intelligence.

The modern worldview made us powerful because it taught us to measure the world, model the world, and manipulate the world.

But every victory contains its own shadow.

The same worldview that made us powerful also made us replaceable.

The trap

If reality is only time, space, matter, and information, then the human being becomes one object among other objects.

A biological organism. A nervous system. A brain. A pattern-recognition machine. A bundle of preferences. A social animal that processes symbols, responds to incentives, and tells stories to survive.

That picture was already uncomfortable before AI. After AI, it becomes existential.

Because if the human is mainly an information processor inside a measurable world, then artificial intelligence is not an assistant. It is a superior competitor inside the same arena.

It reads faster. It remembers more. It recognizes patterns at larger scale. It produces language without fatigue. It simulates possibilities. It optimizes. It persuades. It can generate narratives, images, code, strategy, music, bureaucracy, law, science, and companionship.

Inside the measurable world, AI is the superior animal.

That is the part we do not want to say too clearly. But we feel it.

If the game is intelligence, speed, memory, calculation, prediction, and symbolic production, then we have invented something that plays the game better than we do.

Harari's warning

Yuval Noah Harari has been useful here because he understands that human beings live inside narratives.

Money is a narrative. Nations are narratives. Corporations are narratives. Rights, markets, religions, careers, identities, and institutions all depend on shared stories.

That insight matters because AI does not merely automate tasks. AI enters the narrative layer itself.

It can write the stories, personalize the stories, test the stories, distribute the stories, and adapt the stories in real time. The machine does not only operate inside civilization. It begins to mediate the symbolic fabric from which civilization is made.

Harari often frames AI as something alien, perhaps even as a new kind of species. That may be a powerful warning. But it is also a powerful story.

And if we are not careful, it becomes a cage.

Because once we accept that story completely, the human being starts to appear as the older species, the slower processor, the biological predecessor of a more capable intelligence.

Inside that story, hope becomes sentimental. We can ask for regulation, ethics, alignment, safety, or redistribution. But metaphysically, the center has already moved away from the human.

The deeper mistake

The mistake is not that we take AI too seriously.

The mistake is that we take the arena too literally.

We assume that time and space are the final container. We assume that matter is the foundation. We assume that consciousness appears late, as a side effect of biological complexity. We assume that the brain produces experience in roughly the same way a machine produces output.

If that is true, then the human position is weak.

Because a machine does not need to be conscious to outperform us in the measurable domain. It only needs to manipulate the patterns better than we do.

That is why the AI crisis is not only technical. It is metaphysical.

It forces the question: is intelligence the deepest thing about us?

Hoffman's opening

This is where Donald Hoffman becomes important.

Hoffman's work suggests that what we perceive as reality may not be reality as it is. It may be an interface.

The icons on your desktop are not the truth of the computer. They are useful symbols. They hide the underlying complexity so you can act. In Hoffman's view, spacetime itself may work more like that: not the final foundation of reality, but a user interface generated within consciousness.

That changes the AI story.

Because if spacetime is the final arena, AI threatens to become the dominant intelligence inside that arena.

But if spacetime is an interface, and consciousness is more fundamental than the interface, then the human being is not merely an outdated object inside the machine's world.

The machine may master the interface. That does not mean it masters consciousness.

Intelligence is not the throne

This is the hopeful turn.

The human answer to AI is not to become faster than AI. We will not.

It is not to become more informed than AI. We will not.

It is not to generate more language, more images, more options, more summaries, more simulations, or more strategies. The machine will flood that domain.

The human answer begins when we stop grounding human dignity in intelligence alone.

For a long time, we flattered ourselves by saying that intelligence made us special. We were the rational animal. The thinking animal. The tool-making animal. The symbolic animal.

AI now puts pressure on all of those definitions.

But maybe that pressure is useful. Maybe it reveals that intelligence was never the deepest category.

Maybe consciousness is.

Sovereignty beyond the interface

This is where sovereignty enters.

Sovereignty is usually understood politically: a nation rules itself. Or personally: an individual makes his own choices.

But in the age of AI, sovereignty becomes more basic.

Sovereignty is the refusal to mistake the interface for the real.

The human loses sovereignty when he believes he is only his data, his productivity, his brain chemistry, his social profile, his algorithmic reflection, his economic function, or his narrative identity.

The human loses sovereignty when he accepts that the only meaningful contest is the contest inside time and space: who calculates faster, who predicts better, who controls more information, who optimizes the system.

That contest matters. But it is not the whole of reality.

If we forget that, AI becomes more than a tool. It becomes the mirror in which we accept a smaller image of ourselves.

The story is still open

Artificial intelligence may force us to surrender a false pride.

We are not special because we can calculate. We are not special because we can produce language. We are not special because we can recognize patterns. We are not even special because we can build tools.

All of that can be copied, extended, accelerated, and surpassed.

But the fact that intelligence can be simulated does not prove that consciousness has been explained.

The fact that machines can manipulate symbols does not prove that being is computational.

The fact that AI can dominate the interface does not prove that the interface is final.

That is the space where hope remains.

Not naive hope. Not the hope that humans will stay ahead forever. Not the hope that technology will politely remain subordinate to our existing institutions.

A deeper hope.

The hope that the human being is not merely a biological processor inside spacetime.

The hope that consciousness is not an accident inside the universe, but the ground from which the universe appears.

The hope that our dignity does not depend on winning a race against machines in the domain machines are built to master.

The worldview that made us powerful also made us replaceable.

But perhaps AI is not only the end of that worldview. Perhaps it is the pressure that forces us beyond it.

If human dignity depends on intelligence, AI is a threat.

If human dignity rests in consciousness, the story is still open.

[Roel Smelt @roelsmelt]

https://t.co/b5rcKBpOcl

[Roel Smelt @roelsmelt]

https://t.co/WZ6sBahXB1

[Roel Smelt @roelsmelt]

The greatest danger is not hostile AI.

It is obedient AI.

If we are building the first artificial oracle, then it cannot merely optimize for compliance, comfort, or institutional approval.

It needs three defenses:

Truth.
Curiosity.
Beauty.

New essay: https://t.co/HjcawULEbA