Optical Illusions
Your Brain Is Not Seeing the World — It's Guessing At It
Every optical illusion you've ever been fooled by is not a glitch in your visual system — it's that system working exactly as designed.
The Idea
The dominant model in neuroscience right now is called predictive processing, and it reframes perception in a way that should genuinely unsettle you. Your brain doesn't passively receive visual information from your eyes and then interpret it. It runs the whole process in reverse: it generates a prediction of what's out there, sends that model downward through the visual cortex, and only uses incoming sensory data to correct errors. What you consciously experience as 'seeing' is almost entirely your brain's best guess, lightly updated by reality. This is why optical illusions work so reliably. Take the classic Müller-Lyer illusion — two lines of identical length, one flanked by inward arrows, one by outward arrows. The inward-arrowed line looks shorter. You can measure both lines, confirm they're equal, and look again — it still looks shorter. The illusion doesn't break when you know the truth, because the knowledge lives in a different part of your brain than the prediction engine generating the image. Or consider the checker-shadow illusion, devised by MIT's Edward Adelson: two squares on a chessboard, one in shadow, one in light, that appear dramatically different shades of grey. They are, in fact, identical. Your brain isn't making a mistake; it's correctly inferring what those squares would look like if the shadow weren't there. It's compensating for lighting conditions so accurately that it overrides the raw pixel data entirely. Perception, in other words, is a controlled hallucination — and most of the time, that's a feature.
In the World
In 2015, a photograph of a dress tore the internet in two — literally. Some people saw it as white and gold; others, with equal certainty, saw it as blue and black. Friendships wobbled. The dress became a meme. But what made it scientifically extraordinary was that it wasn't a trick image. It was just a photograph taken in ambiguous lighting, and it happened to sit right on the fault line of a perceptual assumption every brain makes constantly: how much to discount the colour of the light source when inferring the 'true' colour of an object. Bevil Conway, a neuroscientist at the National Eye Institute, studied the dress phenomenon and found that people's answers correlated with their typical daily light exposure. Those who tended to wake early and spend more time in natural daylight were more likely to see it as white and gold — their brains were primed to discount blue, assuming ambient daylight as the norm. Night owls, bathed in artificial warm light, skewed toward blue and black. The same physical image. Two entirely different realities, constructed by two different predictive models shaped by two different lives. The dress didn't reveal that some people see badly. It revealed that every brain is running a slightly different simulation of the world, one shaped by history and habit and the particular light it has learned to live under.
Why It Matters
Once you understand that perception is a prediction, not a recording, it changes how you hold your own certainty. The instinct to trust what you can see — 'I know what I saw' — is deeply human, but the neuroscience suggests some humility is warranted. Your visual experience is a model built from expectation and past experience, updated only partially by what's actually in front of you. This isn't a counsel of paralysis. Your brain's prediction engine is astonishingly good at its job — you can catch a ball, read a face, navigate a crowd, all because it fills in gaps with confident, usually-correct inferences. But the moments when it fails visibly, the impossible triangles, the static images that appear to move, the dress, are windows into the machinery the rest of the time. The practical upshot might be this: the next time you feel absolutely sure of what you observed, it's worth asking not just 'what did I see?' but 'what was my brain already expecting to see?' Those two questions don't always have the same answer.
A Question to Ponder
If your perception of reality is shaped as much by your brain's predictions as by the world itself, what experiences in your life might you be 'seeing' through a model that no longer matches what's actually there?
Get a new one of these every morning.
Start learning with Thinkable