I’ve spent a lot of time focusing on what AI does for individual thinking. From a student who stopped struggling to a professional who stopped believing in their own judgment, it was eye-opening. And while these are real concerns, something bigger could be going on.
Here is the main question. What happens when millions of people make the same “cognitive trade” at the same time? It is not a failure of any consciousness, but manifests itself in concentration.
An important point is disagreement
In general, complex systems do not self-correct through consensus. They correct themselves through the effective friction of people who see things differently. And when that happens, they sometimes come to different conclusions for reasons they can’t fully articulate.
Markets are a great example of this. The financial market is not inherently a calculation. It is a mechanism for aggregating different human judgments. In any transaction, the buyer and seller disagree about the value. This discrepancy is not a flaw in the system, it is the system. Take away friction and you don’t have a smart market. You get something more fragile.
Science, medicine and markets all depend on the same hidden and very human assets. They are full of people who look at the same evidence and come to different conclusions. And that can be valuable beyond measure.
Passive convergence
The calculator has expanded our computing power. The search engine has expanded our access to information. Neither got into the architecture of how people judge.
AI does. And when that happens, the spiral begins. When the same models, trained on the same data and processed by the same data, inform investors’ risk assessments, doctors’ diagnostic judgments, and writers’ editorial judgments, something structurally changes. Individuals remain. A variety of leaves. And as he leaves, the work these people produce will be similar. And this will feed the next generation models, which will further narrow the range.
I wrote about cognitive surrender— a way for humans to delegate judgment to AI not because it’s hard, but because it’s easy. I wrote about collapse of the modelWhat happens when AI is recursively trained on AI-generated data and originality is averaged. Synchronized blindness is when both things happen at the same time, at the population level. One person does not give up, but everyone unites.
What is lost in tails
My point is that statistics are important. Financial bubbles, intelligence failures, medical assumptions that persisted long after the evidence contradicted them. This was not a data breach. They were failures of perception. There are enough people looking at the same facts and missing the same thing at the same time. An experienced investor in the markets will notice something wrong before the data and confirm that it is working pattern recognition built through memory How much does a bad position really cost? This knowledge resides where models cannot reach. It cannot be extracted and encoded. Gotta get it.
When this perception is lost in a population of participants, the system loses its ability to provide early warning. Gone are the anomalies from which a sufficiently diverse field of human thought can be captured. And it’s not because they don’t exist, but the hardware that detects them has converged around the average. There is no tail, it is not said.
Systematic risk we don’t measure
We have frames and clinical instruments for individual measurements cognitive decline. We have little to measure this narrowing of collective perception. This is a hidden problem because the risk of syncope blindness is not visible at the individual level. Everyone can work better than before with the help of their convenient and amazing artificial intelligence tools. Aggregate, at the same time, can see less.
AI will not only change the way people think, but also what we can perceive. And when that perception narrows in general, the danger is not that AI will defeat us. We all begin to lose something important without realizing it.
