Why Highly Intelligent People Are Better at Recognizing Intelligence in Others

Why Are Highly Intelligent People Better at Recognizing Intelligence in Others?

People make many judgments about one another in a very short period of time. In a first meeting, you may hear someone speak and immediately feel, “This person thinks quickly.” Conversely, when someone talks a lot but scatters the point and does not answer your question directly, you may wonder, “Do they actually understand this?”

Of course, this intuition is often wrong. People are influenced by appearance, volume of voice, title, educational background, clothing, age, gender, and the atmosphere of being used to speaking. We may see someone with glasses as intellectual, mistake a fluent speaker for a deep thinker, or underestimate someone who is not verbally smooth. Social judgment contains a great deal of noise.

Still, psychological research has also shown that human beings are not judging others’ abilities in a completely random way. Sometimes we can extract a certain amount of information from small fragments of behavior. Since the 1990s, psychology has developed a research area known as “thin slice” judgment. This field examines how far people can predict another person’s personality, ability, interpersonal style, teacher evaluations, clinical characteristics, and related traits from very brief behavioral fragments. In a 1992 meta-analysis by Nalini Ambady and Robert Rosenthal, even observations shorter than five minutes showed a degree of predictive validity. In other words, first impressions are risky, but they are not pure fantasy.

Within that line of research, a 2026 study published in the journal Intelligence by Christoph Heine, Johannes Zimmermann, Daniel Leising, Michael Dufner, and colleagues takes up a particularly interesting question. How accurately can people detect another person’s intelligence? And what kinds of people are especially good at doing so?

The short answer is that people with higher intelligence tended to judge others’ intelligence more accurately. In addition, people with higher emotion recognition ability and higher life satisfaction also tended to evaluate others’ intelligence more accurately. On the other hand, factors such as gender, empathy, openness, and social curiosity were not clearly related to judgment accuracy, contrary to what the research team had expected.

This result is not simply a matter of “smart people can recognize smart people.” The more important point is that highly intelligent people may use more valid cues when judging another person’s ability. Instead of relying on whether someone speaks loudly, appears confident, or seems friendly, they may place more weight on behavioral information that is actually related to intelligence: clarity of speech, the substance of an explanation, vocabulary use, and how the person handles concepts.

This has important implications for modern society. In hiring interviews, school education, workplace evaluation, debates on social media, YouTube and podcast communication, and the AI-era problem of “text that looks intelligent,” people are constantly estimating the intelligence of others. But the quality of that estimation differs greatly from person to person. Not everyone sees people in the same way.

Intelligence Is Not Just the Amount of Knowledge

The first thing to clarify is that intelligence in psychology does not simply mean being knowledgeable or having an impressive academic background. Intelligence is generally a broad concept that includes the ability to learn, reason, adapt to new situations, solve problems, hold and manipulate information, and handle language, numbers, space, and patterns.

In the early twentieth century, Charles Spearman noticed that performance on many different intellectual tasks tends to correlate positively and proposed the idea of a general intelligence factor. This explains the common foundation behind patterns such as a person who is good at mathematics also tending to perform reasonably well on verbal tasks, or a person with strong memory also being relatively advantaged in reasoning tasks.

Later work by Raymond Cattell, John Horn, John Carroll, and others led to the Cattell-Horn-Carroll theory, which is now one of the major frameworks for thinking about intelligence. In this theory, intelligence is not treated as a single block. It is organized into several broad ability groups, including fluid reasoning, crystallized knowledge, short-term memory, processing speed, visual processing, auditory processing, and long-term memory retrieval. Fluid reasoning is the ability to perceive patterns and relationships in new problems without relying only on existing knowledge. Crystallized knowledge is closer to vocabulary, knowledge, and conceptual understanding accumulated through education and experience.

In many IQ tests, scores are adjusted so that the average is 100 and the standard deviation is 15. In that case, roughly 68 percent of people fall between 85 and 115, and roughly 95 percent fall between 70 and 130. An IQ of 130 or above corresponds statistically to approximately the top two percent. However, IQ does not measure human worth. It does not directly indicate personality, ethics, creativity, perseverance, physical ability, social trustworthiness, or happiness.

Even so, many studies have shown that intelligence matters in social life. In industrial and organizational psychology, general cognitive ability has long been considered one of the strong predictors of job performance and training outcomes. The more complex a job is, the more it requires the ability to learn new information, abstract from it, detect errors, and revise judgments according to the situation. Research in education and health has also repeatedly reported links between cognitive ability or years of education and long-term outcomes in work, health, and lifespan.

In that sense, the ability to recognize another person’s intelligence is not just a conversational skill. It is connected to decisions such as who to work with, who should be trusted with difficult work, whose explanations should be relied on, and who might grow if given support. In companies it may affect hiring and promotion. In schools it may affect how teachers understand students. In research it may shape the choice of collaborators. In creative work it may even influence compatibility with editors and critics.

What the 2026 Study Examined

In the study by Heine and colleagues, 198 people participated as perceivers. Of these, 72 percent were university students, 140 were women, and the average age was 29. Participants watched videos of 50 “targets” whose abilities had already been measured by intelligence tests. Each video was one minute long. In the videos, the targets performed tasks such as reading a weather forecast aloud, talking about a recent enjoyable experience, explaining the meaning of the word “symmetry,” and taking part in a short role-play.

After watching each video, perceivers rated the target’s intelligence on a five-point scale. Later, the perceivers themselves took the same kinds of intelligence tests as the targets. The researchers also measured emotion recognition ability, empathy, personality traits, subjective well-being, life satisfaction, and related characteristics.

What the research team wanted to know was not simply whether people, on average, can guess intelligence. The more important question was: who guesses well? If people differ in judgment accuracy, where does that difference come from? Are highly intelligent people also better when they evaluate someone else’s intelligence? Or are people with high empathy, extraversion, curiosity, or life satisfaction better at it?

The results showed individual differences in judgment accuracy. In other words, some people could detect others’ intelligence relatively well, while others were not very good at it. Higher accuracy was associated with the perceiver’s own intelligence, emotion recognition ability, and life satisfaction.

What makes this study interesting is that it did more than report correlations. It also examined what cues perceivers used. The research team used Brunswik’s lens model to analyze behavioral cues that were actually related to the targets’ intelligence and the cues perceivers actually relied on when making judgments.

The lens model describes how people infer qualities that cannot be directly observed through cues that can be observed. “Intelligence,” for example, cannot be seen directly. But speech content, vocabulary, articulation, explanatory structure, response speed, facial expression, and posture can be seen or heard. The problem is whether those cues are truly related to intelligence, and how strongly observers use each cue.

In this study, good judges made stronger use of valid cues such as articulation, speech content, and vocabulary. This fits everyday experience. A person is not intelligent merely because they line up difficult words. What matters more is whether their word choice fits the situation, whether their explanation is logically structured, whether they can handle abstract concepts in their own words, and whether they can answer questions without missing the point.

Why Highly Intelligent People May Detect Intelligence More Easily

There are several possible reasons.

First, intelligent people may be better at reading the structure contained in another person’s speech. Conversation contains more information than the surface words. Does the person distinguish cause and effect? Do they consider exception conditions? Can they move back and forth between the abstract and the concrete? Can they anticipate objections? Can they admit that they do not know what they do not know? These elements are different from mere conversational smoothness.

For example, suppose someone says, “AI is dangerous.” That sentence alone does not tell us how well the person understands the issue. But if they can then explain that “danger” includes different kinds of risks in military use, job displacement, misinformation, personal dependency, and cyberattack support, we can see their ability to decompose a concept. If they can also add that productivity gains and medical support are real benefits, so the answer should not be a simple ban but regulation designed by use case, we can see their ability to handle opposing elements at the same time.

Highly intelligent perceivers may be more likely to pick up these structural cues. They do not merely receive another person’s words as an impression. They look at the skeleton of the explanation, the construction of causality, and the way concepts are handled.

Second, intelligent people may be less easily fooled by invalid cues. People are often pulled by confident demeanor, fast speech, titles, appearance, fashionable terms, and technical jargon. When someone in a meeting speaks in a decisive tone, they can look as if they understand. On social media, people who make short, strong assertions can appear smart. In video, persuasive force increases simply because the person speaks at a good pace, the editing is clean, the voice is pleasant, and the captions look polished.

But a truly intellectual judge pauses there. They separate the strength of assertion from the strength of evidence. They separate difficult vocabulary from depth of conceptual understanding. They separate explanatory fluency from logical correctness. This is not only intelligence itself; it is also close to metacognition. Metacognition is the ability to inspect one’s own thinking and others’ thinking from one level above.

Third, intelligent people may be sensitive not only to the upper limit of another person’s ability, but also to how that person’s limits appear. On easy topics, almost anyone can speak reasonably well. A person’s real quality becomes visible when the question becomes a little more complex. Do they get confused when conditions multiply? Can they handle exceptions? Can they understand an opposing view and revise their own position? Do they become cautious in a domain they do not know? In these moments, the quality of intellectual processing tends to appear.

This connects directly to hiring interviews. A resume can be filled with polished words, but that alone is not enough. When you ask about concrete incident response, design decisions, failure experiences, and trade-offs, you can see how far the person has structured their own experience. A highly skilled engineer does not merely say, “We introduced microservices.” They can explain why the system was split, which boundaries were wrong, how operational costs increased, and where a monolith would have been better.

Why Emotion Recognition and Life Satisfaction Were Related

In the study, people with higher emotion recognition ability also tended to judge others’ intelligence more accurately. At first glance this may seem surprising. If we are judging intelligence, logic and vocabulary might seem sufficient. But in real human evaluation, cognition and emotion cannot be separated.

In conversation, distinguishing whether another person is anxious, speaking carefully while thinking, pretending to understand, or merely nervous requires reading facial expressions, voice, pauses, gaze, and subtle reactions. If we judge someone who stumbles because they are nervous as “not smart,” we are wrong. Conversely, if we judge someone who is confident but shallow as “excellent,” we are also wrong.

People with high emotion recognition ability may be able to correct for this noise. They can more easily distinguish whether the other person is failing to process intellectually or is simply psychologically inhibited. This is especially important in education. When a student cannot answer, is it because they do not understand, because they are embarrassed, because the relationship with the teacher makes them tense, or because they have misread the question? If this is misjudged, instruction can quickly become unfair.

The relationship with life satisfaction is also interesting. People who are satisfied with their lives may have less excessive defensiveness or hostility when evaluating others. They may have less need to maintain themselves by looking down on others, and less need to overvalue others and depend on them. The more psychologically stable a person is, the more they may be able to observe another person’s behavior as it is.

However, this point must be handled carefully. This study alone cannot determine whether higher life satisfaction makes judgment more accurate, whether accurate judgment stabilizes relationships and thereby raises life satisfaction, or whether a third factor is related to both. The scientifically important point is not to confuse “there is a relationship” with “this is the cause.”

What It Means That Empathy Was Not Related

One surprising result is that empathy was not clearly related to judgment accuracy. It may seem natural to think that people who can emotionally attune to others would also detect others’ intelligence well. But empathy and judgment accuracy are not the same.

Empathy includes an aspect of feeling another person’s emotions oneself and an aspect of understanding the other person’s perspective. If the former is too strong, evaluation may become overly generous, or the observer may be pulled along by the other person’s anxiety. When someone is speaking earnestly, the observer may take that alone as evidence that the person is thinking deeply. Conversely, a calm person with little facial expression may be misunderstood as cold or as not understanding.

Kindness alone is not enough to detect intelligence. One must respect the other person’s effort and emotions while still inspecting the validity of the explanation. This often happens at work. A person with a good personality, a pleasant manner, and consideration for those around them is not necessarily strong at complex design judgment. Conversely, someone quiet and not very sociable may possess a very high level of abstraction ability.

This shows the difficulty of human evaluation. We should not look at people coldly by ability alone. But we should also not confuse goodness of character with intellectual ability. Kindness, sincerity, cooperativeness, intelligence, creativity, and execution ability overlap, but they are distinct qualities.

Distortions in Self-Evaluation and Other-Evaluation

The problem of detecting another person’s intelligence is connected to the problem of how we see our own intelligence. In 1999, Justin Kruger and David Dunning published research showing that people with low ability tend to have difficulty recognizing their own lack of ability and are prone to inflated self-evaluations. This is the so-called Dunning-Kruger effect.

This effect is often used carelessly. The phrase “incompetent people are overconfident” tends to spread on its own, but the core is more precise. People who lack ability in a domain often lack the criteria for recognizing what good judgment in that domain looks like. As a result, they have difficulty recognizing their own errors as errors. This also happens when evaluating others.

For example, a beginner in programming is likely to be impressed by code that is short and works. An experienced developer, however, looks at error handling, maintainability, security, testability, dependencies, and future change cost. Problems invisible to the beginner are visible to the experienced person. The same applies to fiction. Beginners are often impressed by beautiful phrasing, but experienced writers look at consistency of point of view, scene transitions, character desire, placement of foreshadowing, and continuity of style.

In other words, evaluating a particular ability requires some degree of that ability itself. If you do not understand mathematics, you cannot appreciate the beauty of an advanced mathematical proof. If you do not know music theory, you are less likely to notice the sophistication of complex modulations or rhythmic structures. If you do not understand system design, you are less likely to recognize the value of a plain but robust architecture.

The result of this study, that intelligent people are better at recognizing others’ intelligence, is consistent with this general principle. To detect intelligence, one must identify the quality of behavior that intelligence produces. That identification involves the observer’s own cognitive ability.

The Trap of Appearance, Title, and Fluency

The most dangerous thing when judging another person’s intelligence is jumping to easy cues. Because humans often need to judge quickly, they tend to use salient information. Good appearance, a pleasant voice, high academic credentials, employment at a famous company, many followers on social media, technical jargon, and polished presentation skill all influence evaluation.

But these cues do not perfectly match intelligence. Attractive people are more likely to be seen as competent. Glasses can create an intellectual impression. Fluent speech can make a person look as if they have deep understanding. Titles can weaken the listener’s critical thinking.

Of course, titles and credentials are not meaningless. People who graduate from highly selective universities or work in advanced professions have often passed through some kind of selection process. But that is probabilistic information. It does not directly show the individual’s current thinking ability. If we rely too heavily on titles, we stop looking at the quality of the explanation in front of us.

In the present day, this trap has become even stronger. On social media, short and forceful assertions are easy to spread. In video, editing, music, lighting, and subtitles reinforce persuasiveness. With AI, anyone can quickly produce expert-looking prose. The distance between polished writing and the writer’s own understanding is wider than it used to be.

Generative AI research from 2025 to 2026 has also pointed out the possibility that while AI can improve task performance, it can distort people’s assessment of their own ability. For example, in logical reasoning tasks, AI use may improve participant performance while also making participants more likely to overestimate how well they themselves are doing. This happens because AI takes over intellectual output, separating the person’s ability from the quality of the final product.

In the coming era, detecting another person’s intelligence will become increasingly difficult. Surface-level writing, surface-level explanations, and surface-level presentations can all be easily polished by AI. What matters will be not the finished product, but the depth of understanding shown in the moment. Why did you think that? How would it change under different conditions? What are the counterexamples? Where are the weak points in your explanation? The ability to answer these questions will become a more important cue.

Can the Ability to Detect Intelligence Be Trained?

This study showed that people with higher intelligence tend to be more accurate judges. But that does not mean that only people born smart can have a good eye for people. The quality of judgment may be improved by changing how we observe.

First, separate behavioral cues. When evaluating someone, do not stop at global impressions such as “seems smart” or “seems unreliable.” Break down what made you think that. Was it vocabulary? The order of explanation? The response to questions? The way they gave examples? Their form of confidence? Appearance? Title? Simply becoming aware of which information is valid and which is noise can make judgment calmer.

Second, observe multiple situations. Even if one-minute videos can support a certain amount of judgment, real people vary greatly by context. A person may be nervous, talking about a favorite field, being asked about something they do not know, receiving an opposing opinion, working under time pressure, or collaborating with others. Intelligence often appears not in a single utterance but in the response to change.

Third, improve the quality of questions. If you want to see another person’s intelligence, knowledge quizzes are not enough. More effective questions include: “What is the premise of that idea?” “How far does it hold?” “What is weak about it from the opposite position?” “What would be a similar example and a contrasting example?” Intellectual ability appears not only when someone gives the correct answer, but also when they handle ambiguity.

Fourth, record your own biases. In hiring and evaluation, it helps to write down first impressions and compare them with later performance. Do I overvalue people with loud voices? Am I pulled by academic background? Do I underrate calm people? Do I assume young people are immature? Unless you test your own evaluation habits, your eye for people will not improve easily.

Practical Meaning in Work, Education, and Creative Practice

This study has fairly direct implications for workplace evaluation. Especially in intellectual labor, the ability to detect who really understands is important. In software development, law, medicine, research, management, finance, education, editing, planning, and similar domains, outcomes are shaped more by the ability to decompose problems, find blind spots, handle counterarguments, and integrate others’ knowledge than by surface-level confidence.

It is also important for managers. A manager who cannot recognize excellent subordinates cannot make use of the team’s intellectual capital. If they evaluate only loud speakers, skilled presenters, and people who are good at showing themselves to superiors, the people quietly solving essential problems will be buried. Conversely, people who truly think are often cautious, slow to assert, and not flashy at first impression.

The same applies in education. If teachers judge children or students’ intelligence only by how much they speak or how cheerful they are, they will make mistakes. A child who answers quickly is not always thinking deeply. A quiet child is not necessarily failing to understand. Good teachers look not only at whether the answer is correct, but also at why the student thought that, where they got stuck, and which explanation changed their understanding.

This perspective is also useful in creative work. In novels and scripts, writing “he was a genius” in exposition does not make a character convincing. Intelligence must be shown through action. How does the character organize a complex situation? How do they detect a lie? How do they admit their own mistake? What do they prioritize in a crisis? Intelligence appears less in difficult dialogue than in the quality of judgment.

For an activity like Pocholab, which handles writing, video, AI, and personal stories, this problem also matters. When turning a person’s life into writing, that person’s intelligence does not appear only in academic background or titles. How did they interpret hardship? How did they overcome problems with family or work? What did they learn from failure? How did they handle their own limits? There, the person’s unique intellectual outline appears.

But Judging Intelligence Is Also a Dangerous Tool

Up to this point, the ability to judge another person’s intelligence accurately may seem extremely useful. But it is also dangerous. The moment people think “I can see through others,” they easily become arrogant.

Intelligence evaluation always carries ethical risk. If we classify someone early as “smart” or “not smart,” our way of interacting with them changes. People who are expected to do well receive more opportunities and grow further. People who are not expected to do well lose chances to speak. This is related to the expectation effects well known in educational psychology. Evaluation is not merely observation; it is also an action that affects another person’s future.

Intelligence also differs by domain. A person who is extremely strong verbally may be weaker in spatial cognition or interpersonal judgment. A person strong in mathematical reasoning may be rough in story understanding or emotional understanding. In practice, intelligence is not the only factor. Sincerity, persistence, cooperation, physical condition, experience, values, and fit with the environment all matter greatly.

Furthermore, high intelligence does not guarantee good character. A logically sharp person is not necessarily honest. A person who thinks quickly does not necessarily use that ability to help others. High intelligence can also be used for excuses, manipulation, and self-justification. When evaluating intelligence, we must also ask what that intelligence is being directed toward.

Therefore, this study should not be read as “only smart people have the right to judge others.” A more accurate reading is that judging another person’s intellectual ability requires a certain level of cognitive and socio-emotional ability on the observer’s side, and that training in the use of valid cues is important.

In the AI Era, the Value of “Looking Smart” Declines

This topic is especially important in 2026 because generative AI has made it possible to mass-produce output that looks intelligent. Emails, proposals, articles, code, presentation materials, social media posts, sales copy, apology statements, and research summaries can all be made polished by AI. As a result, humans become more vulnerable to surface fluency.

In the past, the ability to write clearly itself indicated a certain degree of understanding. Now, even if the person does not understand, they can ask AI to polish the material and receive writing of a reasonable standard. This is convenient, but it also makes evaluation harder. Simply looking at a finished text makes it difficult to detect the person’s intelligence, understanding, and judgment.

For that reason, process will become increasingly important. What did the person ask AI to do? How did they verify the output? What did they adopt, and what did they discard? Can they notice when AI is wrong? Can they take responsibility for the final judgment as their own? In the AI era, intelligence appears not only as the ability to produce an answer, but also as the ability to doubt an answer.

The same applies when looking at another person’s intelligence. The strong person is not merely someone who can produce polished writing, but someone who can explain where polished writing is dangerous. The strong person is not someone who runs AI-generated code as is, but someone who can add boundary conditions, security, operations, maintainability, and testing perspectives. The strong person is not someone who believes an AI summary, but someone who can compare it with the original and see what is missing.

In other words, the value of “looking smart” declines, while the value of being able to inspect things that look smart rises. This is a significant change.

In Conclusion, What Is an Eye for People?

Highly intelligent people can judge others’ intelligence more accurately. At first glance, this finding may sound a little cold. But read more deeply, it teaches something quite gentle about what it means to have an eye for people.

An eye for people is not the ability to rank others instantly. Rather, it is the ability to separate surface impressions from essential cues. It separates confidence from understanding. It separates fluent speech from deep thought. It separates kindness from intelligence. It separates titles from present judgment. It separates nervousness from lack of ability.

Highly intelligent people may detect others’ intelligence more easily not because they simply recognize their own kind, but because they use information differently. They look not at salient cues, but at valid cues. They look at structure rather than impression. They look not only at the answer, but at the path to the answer. They look not at assertion, but at conditions. They look not at the amount of knowledge, but at the ability to handle concepts.

And this ability matters in every part of society: hiring, education, management, creative work, communication, AI use, and human relationships. Whom do we trust? Whom do we entrust with work? From whom do we learn? Whose potential do we fail to notice? These judgments affect the quality of both life and organizations.

Still, we must not see human beings only through intelligence. Intelligence is an important ability, but it is not the whole person. Smartness gains social meaning only when combined with kindness, sincerity, courage, perseverance, humor, and responsibility. The ability to detect intelligence becomes truly mature when it is used not to judge others from above, but to understand their strengths accurately and help place those strengths where they can be used well.

This 2026 study shows that first impressions are not merely prejudice, but a form of information processing. At the same time, it shows that the quality of that information processing differs greatly from person to person. That is why what we need in the coming era is not simply the ability to “spot smart people.” We need the ability to inspect what cues we use when evaluating people, and to distinguish what merely looks intelligent from what has truly been thought through.

That ability will remain at the center of human society no matter how advanced AI becomes.

---

References

• Christoph Heine, Johannes Zimmermann, Daniel Leising, Michael Dufner, The good judge of intelligence, Intelligence, 2026.
• Nalini Ambady, Robert Rosenthal, Thin slices of expressive behavior as predictors of interpersonal consequences: A meta-analysis, Psychological Bulletin, 1992.
• Justin Kruger, David Dunning, Unskilled and unaware of it: How difficulties in recognizing one's own incompetence lead to inflated self-assessments, Journal of Personality and Social Psychology, 1999.