The Era of Our Best Possible Average | We will never evolve beyond it

I have been watching people, myself included, use AI long enough to notice something. But before I start my rant, let me ask you a question. When was the last time you actually wrote something? Honestly. Independently wrote something useful and thoughtful without coauthoring it, in private or in public, with AI?

Take a second and think about it.

It is not that we are getting dumber. It is that we are getting more comfortable with the dumb version of ourselves.

Congratulations. We built the most powerful thinking machine in human history and the first thing we did was use it to outsource our thinking.

I use AI. Constantly. I am not writing this from some cabin in the woods with a typewriter and a superiority complex. I am inside it every day, using it, benefiting from it, and also watching what it is doing to us in slow motion.

My friend gave me the best analogy I have heard for any of this. She said what is happening now is exactly what happened to architecture. We went from the Renaissance, buildings that were essentially arguments in stone, every column and arch and carved detail saying something is worth the effort of beauty, into minimalism. Glass boxes. Clean lines. The aesthetic of an airport terminal applied to everything we build. It is functional!. Yes it makes sense. But walk into the Alhambra and then walk into a modern office lobby and tell me we did not lose something real. We just decided not to talk about it. We called it progress and renovated our grief into a design principle.

We are doing the same thing to thought. And we are calling it efficiency.

Here is what AI actually is, underneath all the marketing. It is the average of everything. Every article, book, every post. This is not a criticism. This is how to machine is designed. It’s designed to produce the most probable version of a sentence. The most probable version of an idea.

and we were over the moon, we called this intelligence, and maybe it is? but when we noticed that the output was average, instead of thinking more, we invented an entire discipline dedicated to asking the machine better questions. Prompt engineering. There are courses. Certifications. People list it on their CVs. We built a profession out of learning how to talk to a calculator. and no matter how hard you will ever try to engineer a prompt, truth is the machine will not do the actual human thinking on your behalf. The output is only ever as good as the input. You cannot prompt your way to an original idea. You can only get a faster, cleaner version of the one you already had.

What is human thinking? And how is it different than machine thinking?
Well according to my Claude agent, Human thinking is not primarily about producing correct answers. It is about the process of not knowing, sitting in that discomfort, and arriving somewhere unexpected.
the thought changes you on the way to the conclusion. That is the part that matters. You start thinking about one thing and discover something about yourself you did not go looking for. That is not a side effect of human thinking. And it matters. Quite a lot actually.

Not in the success of a project and not in meetings deadlines but it really does matter for our continuity as human beings..
Wait, I am probably being dramatic here.
Our survival as a species has nothing to do with our capacity to think, right?
The reason why we did not go extinct, the reason we are still here at all, has absolutely nothing to do with our intelligence. Nothing to do with our ability to think through problems nobody had seen before. To sit with the unknown and generate, from nothing, something that did not exist yet.

But it’s Fine. Keep outsourcing it. It will be fine.

Let me get back to my original point. When you hand an idea to AI before it has had time to breathe, before it had time to become something you did not expect, you stop it in place. You freeze it at its most basic form. You take the seed and immediately grow it in a greenhouse where the temperature is controlled and the outcome is predictable and nothing surprising can happen. The idea never gets to move. It never gets to keep you up at night and come back different in the morning. It never gets to be wrong first and then slowly right. You completed it. And now it will never be more than that.

You cannot prompt an idea into motion. Motion is the part that is entirely yours.

And as someone who uses AI every day, I can spot AI content almost immediately. Not because of the telltale dashes or the words like "delve" and "quite" that models apparently cannot resist. It is something underneath all of that. Content that arrives technically correct and emotionally absent. It says the right things in the right order in the right time and produces in you exactly nothing. No friction. No surprise. No sense that a specific person, with a specific wound or obsession or weird fixation, thought this specific thought at three in the morning and had to write it down.

You can feel when someone wrote something. You can feel when something was generated. The difference is the same as the difference between a letter and a terms and conditions document. Both are words. Only one of them has a person in it.

What worries me is not the output. The output is fine. What worries me is the input. The thinking that is supposed to happen before you open the tool. The sitting with something uncomfortable until it clarifies. The bad first draft that contains, buried somewhere in its mess, the actual idea. We are skipping all of that. Every day. In increments small enough to feel harmless. And we do not notice what we are losing because the result still looks like a result.

The more you outsource the thinking, the less you trust yourself to do it. The muscle weakens. It does not announce itself. You do not wake up one day suddenly unable to think. You just notice, slowly, that you reach for the tool faster. That the blank page feels more threatening than it used to. That you cannot remember the last time you sat with an idea long enough for it to surprise you.

And then one day you sit down to write something that actually requires you, and you are not sure you are still in there.

This is the part that frightens me, truly frightens me.

Not that AI is too powerful. Not the job displacement, not the regulation debates, not the existential risk discourse that academic men have at conferences. What frightens me is smaller and more personal. It is the possibility that we will forget what it felt like to think something through. To not know and stay with that. To earn the idea.

I don’t really have a beautiful impact closing line, so I’ll leave it as it is.