A kitchen sink, kettle, dishes and utensils beneath an existing overlay of glowing green binary digits.

Cover image supplied from the owner-authorized Psychology Notebook reserve.

Predictive Processing and the Strange Comfort of Being Right

Imagine that you believe you're fundamentally unsuccessful.

Then, one day, something extraordinary happens.

You succeed.

You achieve something difficult. Someone recognises your work. An opportunity opens up. Reality presents you with evidence that perhaps you're not quite as hopeless as you thought.

Excellent news, surely.

Except your mind immediately gets to work.

It was luck.

They don't know how incompetent I really am.

It won't happen again.

And within minutes, the familiar picture of yourself has been successfully restored.

Congratulations. You've managed to turn success into further evidence of failure.

But why would the mind do something like that?

One possible explanation begins with an intriguing idea from neuroscience.

Your Brain Is a Prediction Machine

We tend to imagine perception as a relatively straightforward process.

The world sends information through our senses. The brain processes it. We see what's there.

But a major framework in contemporary neuroscience, known as predictive processing, suggests a more complicated arrangement.

The brain doesn't simply wait for information to arrive.

It continuously generates predictions about what might be happening and compares them with incoming sensory signals.

When the prediction fits, relatively little adjustment may be necessary.

When something unexpected happens, the difference creates a prediction error.

This error provides information that can help revise the brain's model.

In hierarchical predictive coding models, higher levels send predictions downwards, while lower levels communicate information about mismatches upwards.

Perception becomes a continuous negotiation between expectation and evidence.

Neuroscientist Anil Seth has popularised the provocative metaphor of perception as a controlled hallucination.

The important word is controlled.

Our experience is actively constructed, but ordinarily constrained by information from the real world.

We aren't simply inventing reality.

We're making educated guesses about it.

And usually, those guesses work remarkably well.

Two Ways to Deal with Surprise

Neuroscientist Karl Friston developed a broader theoretical framework known as the free-energy principle and active inference.

One of its central ideas is that living systems can reduce discrepancies between predicted and observed sensations through processes involving perception, learning, and action.

In simplified terms, there are two interesting possibilities.

Change your prediction.

You expected something to happen. It didn't. So you update your understanding of the situation.

Or:

Act to change what happens next.

You expect to feel a cup in your hand. You reach towards it, grasp it, and bring about the anticipated sensory experience.

The brain doesn't literally convene a meeting to decide whether learning or action would be preferable.

These are interacting processes described by a computational theory.

But the distinction is psychologically fascinating.

Because sometimes changing our understanding is considerably more uncomfortable than continuing with a familiar explanation.

Even when that explanation makes us miserable.

The Strange Comfort of a Negative Self-Image

Now return to the person who believes they're unsuccessful.

Their expectation isn't simply about an external event.

It's about who they are.

I'm not good enough.

People always reject me.

Whenever something goes well, something terrible follows.

Such beliefs can influence attention, interpretation, memory, and behaviour.

A compliment may be dismissed as politeness.

A rejection may be remembered for years.

A success may be explained away as a lucky accident.

And ordinary uncertainty may be interpreted as confirmation that failure is approaching.

Social psychologist William Swann studied a related phenomenon known as self-verification.

His research suggests that people sometimes prefer feedback consistent with their existing self-conceptions, even when those self-conceptions are negative.

Familiarity can carry a peculiar psychological value.

A painful but predictable identity may sometimes feel more coherent than a positive identity we don't quite believe.

This doesn't mean the brain has a dedicated neurological programme for rejecting success.

But it illustrates how expectations can influence what information we accept and what we dismiss.

Sometimes the difficulty isn't that reality refuses to improve. It's that our interpretation of reality hasn't caught up.

When Expectations Influence Behaviour

The next step is particularly uncomfortable.

Expectations don't only affect what we notice.

They can influence what we do.

Imagine someone who expects rejection.

They approach relationships cautiously. They avoid expressing their needs. They interpret ambiguity as disapproval.

Perhaps they withdraw before another person has the opportunity to respond.

The resulting distance then appears to confirm their original belief.

See? Nobody really wants me.

Or consider someone who believes they cannot succeed.

They delay applying for opportunities. Avoid challenges that might expose their abilities. Dismiss encouragement and abandon projects when uncertainty appears.

Eventually, they have fewer opportunities to experience success.

And their negative self-image remains intact.

This kind of self-reinforcing pattern is psychologically plausible, and related processes have been studied in self-verification research.

But it would be a mistake to claim that predictive coding has scientifically established a universal neural mechanism of self-sabotage.

Real failures, discrimination, exhaustion, limited resources, and difficult environments also shape what people expect and what they can achieve.

Not every negative belief is irrational.

And not every disappointing outcome was created by the person who experienced it.

The Problem with Always Being Right

Perhaps the most interesting feature of predictive processing is that internal models must balance stability with flexibility.

If they changed with every unexpected signal, our perception would become unreliable.

But if they never changed, we'd become unable to learn.

The system has to determine which discrepancies matter.

Some unexpected events are noise.

Others contain information that should change our understanding.

This is partly why researchers study precision, the weight given to different predictions and sensory signals.

If a prior expectation carries too much influence relative to new evidence, updating may become difficult.

The precise role of these mechanisms in complex psychological beliefs remains under investigation.

Still, the broader psychological lesson is useful.

We need stable assumptions to navigate the world.

But we also need the capacity to recognise when those assumptions are wrong.

Especially when they concern ourselves.

Learning to Be Wrong About Yourself

Changing an entrenched belief isn't necessarily a matter of repeating positive affirmations.

If you believe you're worthless, telling yourself you're wonderful may simply feel absurd.

A more useful beginning might be noticing the moments when reality doesn't fit your expectations.

Someone offers genuine appreciation.

You complete something difficult.

A relationship survives a disagreement.

You make a mistake, and the world doesn't collapse.

Instead of immediately explaining away these experiences, you allow them to remain possible evidence.

Perhaps your old model isn't entirely accurate.

Perhaps you can act differently.

Perhaps the next outcome doesn't have to resemble the previous one.

This isn't guaranteed to transform deeply rooted beliefs, and persistent psychological difficulties may require professional support.

But it introduces an important possibility.

You don't always have to make reality agree with your expectations. Sometimes you can allow reality to change your expectations.

And perhaps psychological freedom includes the ability to tolerate being wrong.

Not only about the world.

But about the person you've spent your entire life believing yourself to be.

Framework and scope

Predictive coding, the free-energy principle and active inference are related frameworks, not interchangeable names for one established mechanism. A prediction error need not be a conscious feeling. Seth popularised the controlled-hallucination metaphor and acknowledges earlier uses; it does not mean ordinary perception is unconstrained fantasy. Self-verification is an independently studied social-psychological phenomenon, not direct proof of predictive coding or inevitable self-sabotage. Furutachi and Hofer’s 2026 review highlights unresolved empirical and conceptual questions about predictive processing, especially its extension to complex cognition and beliefs.

References

Keep reading

Explore the writing archive →

Stay in touch

More essays and updates on ABVX Substack.

Read and subscribe on Substack →