You reach into the fridge for the milk carton, expecting it to be full. It's nearly empty, and your arm shoots upward like it's been launched. For a split second, you feel genuinely betrayed by dairy. That little jolt reveals something big: your brain had already worked out how heavy the carton would be before your fingers even touched it.

According to the predictive processing framework, this isn't a quirk. It may be how the whole brain works. Rather than passively recording the world like a camera, your brain constantly guesses what's coming next and checks those guesses against reality. Let's look at how this idea explains perception, why your senses sometimes fool you, and why surprise is secretly your brain's favourite teacher.

Guessing Before Seeing

The traditional picture of perception works like a factory line. Light hits your eyes, signals travel up through the brain, and somewhere at the top a finished image pops out. Predictive processing flips this around. Your brain starts with its best guess about what's out there, based on everything it has learned, and sends that guess downward to meet the incoming signals.

When the guess matches the evidence, there's not much to report. When it doesn't, the brain registers a prediction error, basically a memo saying, “Heads up, something's off here.” In this view, these errors are the main news worth processing. It's wonderfully efficient. Why spend energy re-describing your familiar kitchen every morning when you can simply flag the one mug that's out of place?

The idea has old roots. In the 1800s, the scientist Hermann von Helmholtz suggested that perception involves “unconscious inference.” Modern researchers such as Karl Friston and Andy Clark have built that hunch into a full framework. The neuroscientist Anil Seth puts it memorably: perception is a kind of controlled hallucination. Your brain imagines the world, and your senses keep that imagination on a leash.

Takeaway

What you see is not a live feed of reality but your brain's best guess, constantly edited by the evidence. Perception is less like recording and more like imagining with a fact-checker.

Why Illusions Happen

If perception is guesswork, then illusions are what happens when a confident guess wins an argument it should have lost. The brain weighs its predictions against sensory evidence, and how much it trusts each depends on how reliable they seem. When the signal is weak, blurry or ambiguous, the prediction tends to take over.

Take the hollow mask illusion. Look at the inside of a mask and you'll usually see a face bulging outward. Your brain has seen thousands of convex faces and almost no concave ones, so it overrules your eyes. Or consider phantom phone vibrations: that buzz in your pocket when your phone is sitting on the table. If you're expecting a message, your brain happily reads any twitch of fabric as a notification. Surveys suggest most regular phone users have felt one.

Then there's hearing your name called in a noisy café when nobody said anything of the sort. Your name is one of the strongest predictions you carry around, so your brain is eager to find it in random chatter. These aren't malfunctions. They're the price of a system that's usually right and impressively fast.

Takeaway

Illusions aren't signs of a broken brain but of a brain doing exactly what it's built to do: trusting experience when the evidence is thin. The stronger your expectation, the less evidence it needs.

Learning from Surprise

If prediction errors are the brain's news alerts, they're also its homework. Every mismatch is a chance to update the model so the next guess is better. In the 1990s, neuroscientist Wolfram Schultz found that dopamine neurons in monkeys fired when they received an unexpected squirt of juice. Once the monkeys learned that a signal predicted the juice, the neurons fired at the signal instead, and the juice itself barely registered. Withhold the expected juice, and activity dipped below normal.

In other words, the brain doesn't just care about good things. It cares about good things it didn't see coming. This helps explain why novelty is so appealing. New places, new music, a plot twist you never predicted: they all deliver prediction errors that feel rewarding because they offer something to learn.

But too much surprise is exhausting. Imagine a day where nothing behaves as expected: doors push instead of pull, your coffee tastes like soup, your colleagues speak only in rhyme. Routine is comforting precisely because it keeps errors low and lets your brain rest. Most of us live somewhere in between, seeking just enough surprise to keep learning without feeling lost.

Takeaway

We learn most from what we didn't expect, which is why a good life might need both a comfortable routine and a regular dose of the unfamiliar. The trick is finding your own sweet spot between boredom and overwhelm.

Seeing your brain as a prediction machine changes how you read everyday moments. The jolt from the empty milk carton, the phantom buzz, the thrill of a plot twist: each is your brain comparing what it expected with what it got.

It's also a gentle reminder to hold your perceptions loosely. Albert Bandura showed that our beliefs about what we can do shape what we attempt. Predictive processing goes further: our expectations shape what we see in the first place.