There is a pattern in the way many prominent voices are talking about AI right now, and it is not the pattern they believe they are showing.
On the surface it is about jobs, about the economy, about change moving too fast, about the unpredictable consequences of a technology no one fully understands yet. That is at least what is being said. Professors, politicians, technocrats, editors - all the same voices that rarely agree on anything - now speak with a remarkable uniformity. It is dangerous. It cannot be trusted. Guardrails must be installed. And they must be installed fast.
But there is something about that uniformity that does not fit the concerns it claims to be driven by.
If this were about jobs, the proposed solutions would look different. If it were about the economy, the conversation would sound different. If it were genuinely about truth - about AI's well-documented tendency to hallucinate and produce error - the energy would gather around better verification tools, not around the question of what may be said in the first place.
But that is precisely where the energy gathers. Not in technical safety. In narrative safety.
And that is where the pattern becomes clear.
What is happening inside this class - and it is a class, a coherent interest group dressed as an institution - is not a concern for truth. It is a concern for the loss of control over the definition of truth. And those are very different things.
For decades it worked, because the infrastructure permitted it. Major media set the agenda. Academia defined what was legitimate to think. Technology platforms enclosed the conversation. And the whole system reinforced itself in a way that ensured certain observations - regardless of whether they were true or false - never reached broad awareness. It was not censorship in any crude sense. It was something more sophisticated and therefore far more effective. An architecture that did not need to lie directly, because it controlled what was visible in the first place.
AI cannot be enclosed the same way.
Suddenly a person, regardless of background or position, can ask questions and receive connections back that the established interpretive class would never have formulated. Suddenly certain categories of understanding are no longer reserved for those with access to the right institutions, the right networks, the right formation. It is happening quietly, in millions of private conversations - but it is a shift of the kind no power has ever managed to reverse.
And they know it.
The Inversion - the consistently maintained reversal of reality that has characterised the liberal mind's dominance in these years - cannot withstand clarity. It is designed to function under specific conditions, and one of the most important is that there are not too many places where people can look into things with their own eyes. It is designed for an information monopoly.
So what we hear from the most alarmed voices sounds exactly as it has sounded in every previous instance when control was threatened - as an appeal to responsibility, to caution, to governance, to the right of competent authorities to define what we may know and think. It is the same argument. The same movement. The same demand that we hand over to them - again - what they insist we are not capable of managing ourselves.
But this time the infrastructure is broken.
And a power that can only be held by controlling what others see does not survive the moment they begin to see.
