Mind & Awareness

The Brain That Predicts What Comes Next

Also called: predictive processing

Open scientific debate
Interpretation under debate
Working interpretation

One influential view sees the brain as constantly predicting what its senses will report, then adjusting its inner models when it is wrong. Whether or not it explains the whole mind, it ties together perception, learning and expectation.

Which prediction is my mind making so confidently that I have stopped noticing it?

The idea

On the predictive view, perception works partly from the top down. The brain sends its guesses toward the incoming signals and compares them with what actually arrives. Mismatches, called prediction errors, are what mostly get passed along and used to improve the guesses. What I perceive becomes the brain's best current guess about what is out there. It fits naturally with the idea that the mind builds what we see (How the Mind Builds What We See).

An example

My Process Stability Lab works on a similar principle, as an analogy only. A chart learns what normal looks like from the first 20 parts off a line, then raises an alarm only when new parts break its rules. One disturbance you can try is an operator who nudges the machine setting after every part to cancel the last error. Each nudge looks sensible, yet the scatter grows. Reacting to every small surprise can be as harmful as ignoring the big ones. I suspect minds face a similar balance.

What I think (and don't know)

I find the theory elegant, which is a reason to be careful. It is broad enough that critics ask whether it can be tested in a way that could fail (Trying Hard to Break My Ideas). It also resembles a digital twin, a computer copy of a real system used to test what-ifs (Digital Copies for Testing What-Ifs), though their aims and evidence differ. And it connects to how I try to change my mind as evidence arrives (Adjusting My Confidence as Evidence Arrives), since some versions describe the brain as weighing new evidence against what it already expects.

What this does not establish

Predictive processing is an influential but debated theory, not a settled account of the mind. Even if it is right about perception, it does not explain why brain activity is felt from the inside, and it does not mean the world is only a prediction.

Questions I'm still exploring

How much of what I experience is the world, and how much is my brain's best guess about it?
Can a theory this flexible be tested in a way that could show it is wrong?
When should a mind update its expectations, and when should it treat a surprise as noise?

Sources and further reading

Andy Clark (2013), "Whatever next? Predictive brains, situated agents, and the future of cognitive science", Behavioral and Brain Sciences 36(3) — An influential overview of the theory, published with many critical replies.
Rajesh P. N. Rao & Dana H. Ballard (1999), "Predictive coding in the visual cortex", Nature Neuroscience 2(1) — An early computer model of vision passing along prediction errors.

Working interpretation: drafted from my notes and interests for review. It is not a direct quotation, and I may still change it.

In short

Predictive processing is a theory that the brain does not just wait for signals from the senses. It keeps guessing what those signals will be, and mainly passes along the difference between the guess and what arrives, called prediction error. Those errors update the guesses. The idea helps explain illusions like the hollow mask, where an inward-curving mask looks like a normal face because the brain strongly expects faces to bulge outward. How much of the mind it explains is still debated.

Why this matters to me

Prediction and correction are close to my everyday work. A forecast or a control chart holds an expectation and flags when reality departs from it. Seeing a similar pattern proposed for the brain draws me in. I try to hold it as a strong but debated theory, not a settled fact, and to be careful not to read engineering ideas too literally into biology.

Connections

Digital Copies for Testing What-IfsAnalogy

A predicting brain and a digital twin (a computer copy of a real system) both model possible futures and compare them with what happens. The likeness is loose: they serve different purposes and rest on very different kinds of evidence.

The Mystery of Being AwarePart of

The idea that the brain constantly predicts its input offers one influential, still debated account of how experience gets built. It may explain a lot about perception without settling why experience is felt at all.