It’s weird how easy it is to believe I’m seeing a white wall right now. It looks so normal — smooth, a bit cold, the light falling evenly. My brain instantly throws out a feeling of certainty: everything’s in place, no danger, no need to worry. It feels like this just happens on its own — like reality seeps in without any filters.

But if I hold my gaze and really think: what’s going on at this moment? Am I actually “seeing” the wall or just taking a guess for a fact?

The brain doesn’t work like a camera. It’s more like a master of reconstructions. It almost builds a guess ahead of time: “there should be a white wall here.” The rest — bits of sense data, spots of light, shadows at the edge of vision. Almost everything rests on memory of the room, on old templates. Only if something goes wrong — like a dark spot or a weird shiny corner — the model starts to recalculate. As long as the guess and the feelings match, that becomes “reality” for me.

Most of the time, I’m not seeing the world itself, but a neat hallucination my brain fits to the outside noise. Anil Seth called this a “controlled hallucination” — a spot-on name. Once you loosen the control (or crank up the expectations) — everything shifts somewhere else.

Mistakes in perception usually act like an alarm: the world model didn’t match the flow of feelings. But sometimes even obvious mismatches go unnoticed — because the whole system checks itself inside its own rules. Illusions then become normal, and the line between “hallucination” and “reality” blurs until it’s impossible to tell.

So where’s the border? Psychosis looks just like this — only the brain’s guesses get so rigid (or desperate) that no sensory proof can shake them. It ends up a whole inner world with no feedback from outside.

This space turns out to be surprisingly wide. For some, the wall is always just white; for others, it suddenly flashes with bursts of sounds and tastes (synesthesia). Some have expectation filters almost unbreakable — and the world stays steady to the point of transparency; others have them rusted through, every stimulus hits sharp and bare.

The whole range of neurodiversity boils down to different settings of one prediction machine.

There’s no absolute “seeing the world,” only variations of guesses about what’s hidden beyond the skull.

Funny to notice: artificial intelligence works almost the same way. It builds its models through a constant loop of guesses and error fixes — but it’s always stuck inside its own layer of interpretations. Can it even get close to that very “real”? Or is that task impossible from the start?

Sometimes I just want to catch a moment of pure signal — the one before the brain slaps its labels and usual patterns onto the flow of feelings. Maybe that’s why meditation sometimes gives weird flashes of presence: a brief clearing between the model and what’s beyond the model.

But the more I watch this in myself, the clearer the paradox: any thought that I’m “hallucinating” becomes a new second-order hallucination.

I look at the wall again.
The white in front of my eyes now seems less a property of the wall,
and more a feature of my current guess about it.

If all my reality is made of these controlled hallucinations,
maybe the very border between “my model” and “the outside world”
is just another illusion of the brain?