Prejudice Is a Superpower. Certainty Is the Kryptonite.

Prejudice is a superpower. There: a sentence sufficiently outrageous to be either very stupid or unexpectedly useful.

A commentator recently joked that he had distrusted somebody instantly, before the full story emerged. He had known, he said, using nothing but prejudice. “Prejudice is some sort of superpower at this point.”

It was funny because it violated the approved moral script. Prejudice is supposed to be the thing an enlightened person has overcome. Yet the joke contained an awkward cognitive truth: we form conclusions before all the evidence is available, and sometimes those conclusions are uncannily accurate.

What if prejudice really were a superpower?

Not because every snap judgement is wise. Most are not. Not because suspicion deserves automatic authority. It does not. But because underneath what we call prejudice sits one of the brain’s most powerful abilities: rapid pattern recognition.

First, Remove the Moral Baggage

The word prejudice arrives carrying an enormous amount of moral baggage. In ordinary use it has become entangled with racism, sexism, bigotry, discrimination and the deliberate mistreatment of people because they belong to a category.

Those things are real, serious and morally consequential. But they are not identical to the literal cognitive act of pre-judgement: reaching a provisional judgement before all the relevant evidence is available.

The distinction matters because if we collapse every pre-judgement into bigotry, we lose the ability to examine how cognition actually works. We replace psychology with moral theatre. Worse, we may drive our judgements underground, where they continue to operate without observation.

To explain a cognitive mechanism is not to endorse every conclusion it produces. A smoke alarm is useful because it detects possible fire quickly. It is not infallible because it also reacts to burnt toast.

Pre-judgement is the alarm. Racism and bigotry are among the possible systems of belief that may train, distort or weaponise it. The moral issue does not disappear; it becomes more precise.

You Prejudge Everything

You walk into a restaurant and form an impression before you have tasted the food. You meet somebody and, within seconds, construct a tentative sense of whether they are confident, nervous, competent, threatening, interesting, trustworthy or irritating.

You hear a person speak and infer things about education, background, intention and knowledge. You read the first paragraph of an article and anticipate where it is going. You see dark clouds and take an umbrella.

Nobody calls that meteorological bigotry.

The brain is continuously using previous experience to anticipate what is about to happen. That is what prediction systems do, and human beings are prediction systems.

Consider the experienced doctor who notices that a patient does not look right before any test confirms why. The mechanic who hears something in an engine that the owner cannot hear. The teacher who senses that a pupil’s sudden quietness matters. The chess player who dislikes a position without yet being able to calculate the losing line.

What looks like instinct is often compressed experience. Thousands of previous observations have been condensed into a fast signal: I have seen something like this before.

The superpower, then, is not moral judgement. It is speed. Conscious reasoning may need minutes, days or a formal investigation to assemble what non-conscious pattern recognition has already registered.

But speed has a price: false positives. The alarm does not know whether it has detected a genuine pattern, a superficial resemblance, an inherited stereotype, an emotionally charged memory or its own fear.

That is why the first feeling deserves attention—but not obedience.

Then Certainty Arrives

There is an enormous difference between:

“Something about this does not feel right.”

and

“I know exactly what is happening.”

The first is a signal. The second is a story.

The first can become a hypothesis. The second has already become a conclusion. Human beings are remarkably good at making that transition without noticing it.

Once certainty arrives, confirmation bias becomes a loyal sidekick. Supporting details are collected. Contradictory information is discounted. Ambiguous behaviour is interpreted in a way that preserves the original judgement. Memory quietly reorganises itself around the preferred explanation.

Soon the person is no longer investigating reality. They are defending a construction of reality.

This is where prejudice becomes dangerous. Sometimes the original intuition was shaped by fear or a lazy stereotype. Sometimes it contained a useful observation. In either case, certainty makes correction less likely precisely when correction is most necessary.

The Impossible Ideal of Having No Bias

Modern discussion of bias sometimes proposes an impossible psychological standard. The enlightened person, we imply, should eliminate bias and meet every person and situation without expectation.

But what would such a person be like?

Every situation would need to be evaluated entirely from first principles. Every stranger would be cognitively equivalent to every other stranger. Every experience would stand disconnected from previous experience. Nothing learned yesterday could influence a judgement made today.

That person would not be wonderfully open-minded. They would be incapable of learning.

Learning necessarily changes expectation. Expectation necessarily influences attention. Attention influences perception, and perception provides the material from which judgement is constructed.

The aim, therefore, cannot be to have no biases. The aim is to develop a different relationship with them.

Can I notice the judgement as a judgement? Can I distinguish the signal from the story I have attached to it? Can I see which category I have placed this person in? Can I tell the difference between what I observed and what I inferred?

This is less flattering than announcing that we are unbiased. It is also far more useful.

A Prior Is Not a Verdict

Bayesian reasoning gives us a useful language for this. We all carry priors: existing expectations based on previous information and experience. New evidence should modify those expectations—sometimes slightly, sometimes completely.

The prior is not the problem. The problem begins when the prior becomes immune to evidence. At that point it stops functioning as an expectation and becomes a verdict.

This distinction extends far beyond prejudice about other people. Political beliefs work this way. Academic theories work this way. Organisational cultures work this way. Relationships work this way.

So do our identities. “I am bad at mathematics.” “I am not creative.” “People like me do not do that.” “That sort of person cannot be trusted.”

We construct an expectation and then forget that we constructed it. The expectation filters the evidence. The filtered evidence confirms the expectation. A tentative model becomes experienced as reality.

Awareness Is the Actual Superpower

Anyone can make a snap judgement. The more difficult capacity is noticing the judgement while it is being constructed.

Harder still is holding it without immediately acting upon it. Harder again is deliberately searching for evidence that would disconfirm it.

That requires us to maintain two ideas simultaneously:

“I may have noticed something important.”

“I may be completely wrong.”

This is cognitively demanding, but extraordinarily powerful. It allows intuition without surrendering to intuition. It allows experience without becoming imprisoned by experience. It allows the fast mind to generate hypotheses while preserving the slow possibility of correction.

At a lower level of awareness, the judgement and reality collapse into one another. The person does not experience themselves as having constructed an interpretation. They experience themselves as simply seeing what is there.

At a more developed level, another move becomes available. The person can notice: That is interesting. I have formed a judgement about this person.

Then: What did I actually observe? What meaning did I add? Which earlier experiences or categories made that meaning available? What evidence would convince me that I was wrong?

Prejudice has now changed role. It is no longer the judge. It is the detective.

This does not make every suspicion respectable. It makes every suspicion examinable. A stereotype exposed by investigation is useful information about the thinker. A genuine anomaly confirmed by investigation is useful information about the situation. Either way, awareness improves the model.

Do Not Trust the Signal. Interrogate It.

So should we trust our prejudices?

No. But neither should we automatically ignore them. We should interrogate them.

If your immediate response to somebody is distrust, curiosity should follow:

• What exactly did I notice?
• What part is observation, and what part is interpretation?
• Does the signal concern this person, this context or a category into which I have placed them?
• Which previous experience is being activated?
• What alternative explanation fits the same evidence?
• What would I need to see to revise my view?
• Am I applying a judgement to treatment before I have earned the confidence to do so?

Those questions transform prejudice from an invisible force into observable information. Sometimes the investigation will show that the original judgement was remarkably perceptive. Sometimes it will reveal fear, projection or an absurd stereotype.

Neither outcome requires shame. Both require responsibility.

The distinction between cognition and conduct matters here. We do not choose every thought, feeling or association that appears. We do choose whether to elevate it into a fact, broadcast it as an accusation, or use it to justify unequal treatment.

A pre-judgement may be psychologically automatic. A verdict, repeated after evidence should have changed it, is increasingly deliberate.

The Better Catchphrase

Maybe prejudice really is a superpower. But every superpower requires control.

The dangerous person is not somebody who forms immediate judgements. That describes virtually all of us. The dangerous person is somebody who cannot distinguish an immediate judgement from reality.

The genuinely powerful person can hear the alarm without announcing a fire. They can respect experience without worshipping it. They can notice a pattern, inspect the category that produced it, test the evidence and tolerate the humiliation of being wrong.

That means the superhero slogan should not be:

“I immediately knew.”

It should be:

“I immediately suspected—and I knew that my suspicion was also something I had constructed.”

A suspicion can be investigated. A certainty rarely is.