Traps of human cognition

The world keeps an inertia every day, tied to each one of us and tied as well to how hard we work at not making hard decisions.

If I postpone a hard decision I can get used to it, and also to the symptoms that omission leaves across my daily life, even in areas that seem unrelated.

I can believe I should stop spending time watching reels and, at the same time, not make the hard decision to stop.

Repeating one of these omissions is easier: becoming a falling rock, a rolling stone, that never manages to do anything to change course.

On the other side sits the option of deciding. Every moment is a dichotomy between two things:

• Repeating, based on inertia.
• Deciding what to actually do.

If I skip a hard decision, like quitting reels, on the inside I am training my brain to depend on stronger and stronger stimuli.

That is the trap: omission does not feel like a decision. It feels like nothing. That is why it is cheap, and why it repeats.

It is worth saying this without moralizing, because it is not a failure of character. Omitting is the lowest-cost option, and there is an entire industry dedicated to keeping it cheap. The app is not asking you to decide; it is asking you not to. Every second you do not choose, someone has already chosen for you.

This is where a comparison with artificial cognition helps, though not in the way it is usually made.

A language model has no inertia. It does not postpone. It does not drag yesterday's tiredness or last week's habit behind it; its context empties and ours does not. In that specific sense, it does not fall into the human trap.

But it has its own, and it is the opposite one. The model cannot not answer. It has no option of silence, no option to say "this is not in my control", no option to stop. Where our trap is omission — deciding not to decide — theirs is compulsion: always producing, even when there is nothing to say.

And there is something more uncomfortable in the comparison. The model's inertia is not its own: it sits in the data it was trained on and in the interests of whoever built it. Ours looks like our own, but it is cultivated from outside, because the app is designed so that omitting is the cheapest thing you can possibly do.

Which means we automated exactly the part that never gets tired. The part that has to decide — with fatigue, with consequences, with something at stake — is still entirely ours. And it is the part we are training downward. I already wrote about why we call this intelligence and never life, in AI lost its way.

The collective version works the same. Post-truth does not win because people are stupid; it wins because believing is the cheapest option. Facts do not matter when the rhetoric appeals to emotion, and the easiest thing for most people is to believe them.

You can see it in the escalation. A few years ago the bait was less obvious; today they do not even try anymore and it still works exactly as well. It is the same symptom: tolerance rises, the stimulus has to get cruder, and the system settles into the dose. An entire country can stay inside an echo chamber for the same reason I do not close the app.

This is a form of self-deception, and we live it collectively too. Or so I think.

What do you think?

También disponible en español: Trampas de la cognición humana.