The biggest thing AI has taught me has nothing to do with AI. It has changed how I think about reality itself. The more I understand how modern AI works, the more convinced I become that we are living inside a simulation.

Most people think AI and humans are fundamentally different. I don’t think so. I think they are running the same underlying mechanism. The difference is only in the scale and sophistication.

Let’s start with what an LLM actually does. An LLM predicts the next word. That’s it.

Everything it says, every poem it writes, every joke it tells, every programming language it understands, every explanation it gives, is ultimately the result of predicting what comes next based on everything it has learned during training.

People hear that and think, “Humans are completely different.” But are we?

When I listen to someone speaking, my brain is already predicting what they are about to say before they finish the sentence. When I ride my bike, I predict what the vehicle ahead of me is going to do. When I throw a ball, I predict where it will land.

When I meet someone, I predict how they will react to what I say. When I invest money, I predict the future. Even emotions involve prediction. Fear is predicting danger. Hope is predicting a better future. Anxiety is predicting negative outcomes, while confidence is predicting success.

Prediction is everywhere. The only difference is that humans predict much more than language. If an LLM is a Large Language Model, then humans are Large Action Models. We predict actions, outcomes, emotions, social interactions, and consequences. Language is simply one tiny part of a much larger prediction engine.

People often point out what they believe is the biggest difference. “An AI needs a prompt,” they say. “A human doesn’t.”

At first glance, that sounds like a fundamental difference. But I don’t think it is. I think humans are simply running on a permanent prompt.

Let me explain.

Imagine creating an AI and giving it a set of instructions that never disappears. Stay alive, learn, explore, avoid danger, seek pleasure, find a mate, reproduce, protect yourself, improve your chances of survival. Never stop.

Now don’t ask it another question ever again. Just let it exist. It keeps observing, learning, acting, making decisions, and updating its understanding of the world.

From the outside, nobody is prompting it anymore. From the inside, it feels like it has motivation. It feels like it has desires and ambitions. But those ambitions may simply be the permanent prompt continuing to execute. Humans look exactly like that.

Nobody wakes me up every morning and gives me a new instruction manual. Nobody tells me to eat, avoid pain, care about the people I love, or work toward my goals. Those drives simply exist. We call them motivation, ambition, purpose.

Then there is another obvious objection: Humans don’t just process language.

Exactly. Neither would a sufficiently advanced AI. Today’s LLMs mainly predict text. Tomorrow’s systems will predict much more.

They will understand images, video, audio, touch, space, and movement. Eventually they may build an internal model of the world itself.

Now imagine taking that one step further. Imagine building an intelligence that continuously receives information from every sensor available to it, just like humans receive information through sight, hearing, touch, taste, and smell.

Now give it visualization and sounds. And why not? Language is also not something physical, it’s essentially just electrical signals interpreted and felt as language. So can be visuals and sounds. If an AI can generate language, it can generate visuals and sounds. Instead of merely predicting words, let it continuously render an entire world inside its own processing. Every object, every person, every sound, every movement, everything.

Now let it move through that world while continuously updating its predictions. From inside that system, what would it feel like? It would feel exactly like being alive. The system wouldn’t experience itself as a computer. It would experience itself as a person living inside a world.

It would have no reason to believe that the world it experiences is generated. It would simply call it reality.

Now reverse the thought experiment. What if we are that system?

BOOM!

What if our brains are not fundamentally different from advanced predictive intelligence? What if everything we experience, including the world around us, is the rendered environment inside an unimaginably advanced computational system?

This is where simulation theory stops sounding like science fiction to me. It starts sounding like reality.

People often ask for proof. I don’t think proof is even possible from inside the system. If the simulation is complete, every experiment we perform is still happening inside the simulation.

When I look at ideas like the universe having a speed limit, quantum phenomena like the double-slit experiment, the apparent mathematical nature of reality, code reuse throughout biology, and the remarkable similarities between artificial intelligence and human intelligence, I see evidence. None of these observations proves we are living in a simulation. Together, however, they point me toward a single explanation that I find more compelling than the alternatives.

The more we build intelligent machines, the less special humans appear. Not in value. In mechanism. Every breakthrough in AI removes another mystery that we once believed belonged exclusively to biological intelligence. That trend has only strengthened my conviction.

Ironically, AI didn’t convince me that machines are becoming human. It convinced me that humans may have been computational systems all along.

And if that’s true, then the line between artificial intelligence and natural intelligence may never have existed in the first place.