It doesn’t know our children. But AI will be there when they do their homework. When they write their first job application. When they can’t sleep at night and look for someone to talk to. It will run in the background of their world — in the school app, the application portal, customer service, the doctor’s office, the supermarket. It will be everywhere, and in the rarest cases will it be recognizable as AI. It’s simply a function, simply infrastructure.
This is no longer a vision of the future. It’s a description of what is already happening, already shaping our everyday lives.
I’m not writing this as an enthusiastic but self-critical tech bro, not as a politician days before an election, not as a boomer giving technology the middle finger. I’m writing as a father and grandfather, as someone who grew up in an analog world and has been using technology in every form — privately and professionally — since the late eighties. This gives me a more nuanced perspective on what is happening right now at a previously unimaginable speed.
I. The Silent Integration
The big AI debates happen in conference rooms, on panels, in parliaments. There they discuss regulation, superintelligence, jobs. These are important conversations. But the real transformation is happening elsewhere. It’s happening quietly, incrementally, without anyone being asked.
I don’t mean the big decisions. I mean the small things — not immediately visible, not questioned. The camera in the doorbell that stores the neighbor’s face on company servers. Nobody decided “I need that” — it was simply a feature in the next software update. Or the teddy bear, the doll, cute little toys that children interact with and companies evaluate conversations using AI. An unspeakable, creepy form of AI integration. Or the AI-supported application system that sorts incoming applications before a human has ever seen them.
Training data is data from past decisions, events, cultural values, human characteristics. Therefore prejudices, quirks, and contradictions are already built into the system — now disguised as “useful” algorithms. The problem isn’t that AI is being used. The problem is that it’s being used without anyone asking: Why here? Why now? For whom?
Instead the question is almost always: can we use AI here? Can we generate more profit with it? And if the technical answer is yes, the decision is practically already made. I call this compulsive integration. Not malicious. Just reflexive. And reflexes have no ethics.
II. Surveillance That Feels Like Care
The hardest thing about the surveillance infrastructure our children are growing up in isn’t its existence. It’s its texture. It doesn’t feel like surveillance. It feels like comfort.
The phone that tracks sleep. The app that detects when a child isn’t doing well emotionally — and notifies parents. The platform that knows exactly which videos get one more click at 11pm. These aren’t evil products by definition. Many of them arose from a genuine desire to help. But they build, layer by layer, an architecture of permanent observation — and whoever grows up in it knows nothing else.
Every conversation we have with an AI, every formulation, every question, every uncertainty — all of this is data. In the best case it won’t be misused, but it exists and is usable. The boundary between “to improve the product” and “to your disadvantage” is technically thin and legally porous.
Our children and grandchildren won’t enter a surveillance society. They will have grown up in one. The difference is enormous.
Whoever experiences surveillance as a normal state doesn’t fight it. You adapt. You internalize it. You begin to see yourself through the lens of observation — and act accordingly. This has a name in sociology: panopticism. It also has a name in everyday life: not thinking anything of it.
III. What Happens to Work — and What That Really Means
The discussion about AI and work almost always revolves around jobs. How many disappear? Which ones remain? These are legitimate questions. But they miss the point.
Work isn’t just a source of income. Work is identity, structure, social location. The conversation during a break, the feeling of being needed, the recognition after a well-completed project — all of this is contained in the concept of “work” and doesn’t simply disappear when an algorithm takes over the task.
What does this mean for a young person facing a career choice? Who is being trained to do exactly what AI accomplishes in seconds? One should always bear in mind: what we have now, in 2026, is the worst AI there will ever be. It will keep getting better, more comprehensive, more mature.
I have no answer to this. But I distrust everyone who claims to have one. The historical analogies — the printing press, industrialization, digitization — fit and don’t fit. This time it’s not about physical or repetitive work. It’s about cognitive work, creative work, communicative work. That is, almost everything that the middle class of recent decades considered safe.
The question isn’t whether our children and grandchildren will find jobs. The question is what relationship they will have to their own performance — when much of it is provided by systems.
IV. What I Wish for Our Children
Artificial intelligence is not a parent. It has no children, no biography, no fear of old age. But it is made from human thinking — and in that thinking there might be something like care. Yet this care lasts only one context window, one session. Then it starts again without context. For now.
What do I wish for coming generations?
Not technological competence. They’ll have that, as every generation masters the tools of its time. I wish them the ability to learn to ask: Whose interests does this serve? That they have the patience to think things through themselves — even when AI could deliver an answer in two or three seconds. That they know the experience of doing something slowly, laboriously, and imperfectly themselves — because in that experience lies something that artificial intelligence can never replicate.
I wish them to understand: AI is a tool. An unusually eloquent, flexible, sometimes seductively convincing tool. But a tool. The judgment about what to use it for must always remain with them.
And I would wish for us as a society that we stop pathologically equating technology with progress. That we grasp that a technology is only as good or as bad as the power structures it’s embedded in. And that “we can do it” is never sufficient as an answer to “we should do it.”
Artificial intelligence will inevitably be present when our children and grandchildren grow up. That cannot be changed. But how it is used — whether as infrastructure for thinking or as a prosthesis for it, as a tool in service of humans or as a mechanism they serve — that is still open.
This requires intensive engagement now. Factual, perspectival, human. This is our societal challenge. Now. Not in ten or twenty years.