Schools look for genius. The rest of us need to learn for ourselves

I increasingly wonder whether we overestimate the extent to which schools are designed to teach every individual well, and underestimate their role as a system for identifying and selecting talent.

Schools serve thousands, millions, even hundreds of millions of students, rather than any one particular person. They need a common curriculum, a relatively fixed pace, standardized exams, and a way to compare results. We often find this arrangement too rigid, too indifferent to individual needs. But seen from another angle, that standardization is what allows schooling to operate at scale.

Only when enough people encounter similar bodies of knowledge can the system begin to see the outliers. We often ask why someone who will never become a mathematician needs to study so much mathematics, or why someone who will never do physics research needs to learn physics. From an individual’s perspective, these are reasonable questions. From society’s perspective, the answer is fairly straightforward: until a child has actually encountered mathematics, physics, music, or literature, no one knows whether they might have an exceptional ability in one of those fields.

So universal education and the discovery of exceptional talent are quite compatible. Precisely because extraordinary ability can emerge anywhere, in any family, society needs to give as many people as possible access to a broad range of knowledge. A child who has never encountered something has little chance of discovering a gift for it. In that sense, one of the important missions of schooling may always have been to avoid overlooking people whose abilities might otherwise go unnoticed.

But there is an easily missed distinction here: the goals of a social system do not automatically coincide with those of an individual.

Society has good reason to want as many people as possible to take part in the same education and selection system. The wider its reach, the more likely it is to discover talent. For an individual, though, the central question is whether they are gaining the abilities they need to do what they actually want to do, rather than whether they can keep climbing the rankings within that system.

In the past, these two concerns were difficult to separate. The reason is simple: outside school, most people struggled to find better conditions for learning. Entering an unfamiliar field meant first finding textbooks, courses, and teachers. When something made no sense, there was not necessarily anyone available to explain it. Standardized education did not suit everyone equally well, but for a long time it was one of the best affordable ways to acquire a systematic body of knowledge.

That is the premise AI is beginning to change.

I do not think AI will replace schools anytime soon. Schools also provide qualifications, a community of peers, socialization, laboratory resources, and the formal training many professions require. A chat window cannot easily replace those things. But AI is beginning to make something once very expensive widely available: for the first time, ordinary people can get something approaching one-to-one learning support at very little cost.

You can tell it what you already know, what you are working on, and why you need to understand a particular subject. When an explanation does not make sense, you can ask for another approach, or go back to the background knowledge you are missing. Learning no longer has to follow the sequence laid out in a course. It can develop around a real problem.

This may be AI’s most important effect on how ordinary people learn.

Learning used to be largely driven by the curriculum: first learn what school says you should know; whether you will ever need it can be settled later. Increasingly, it can be driven by a problem instead: start with something you actually want to solve, then follow it wherever it leads, filling in the knowledge you need along the way.

If you want to build a product, learn the design, development, marketing, and business skills it calls for. If you get halfway through and realize you do not understand databases, study databases. If you need to read financial statements, learn accounting. If you want to enter an overseas market, work on the language and local culture. Learning becomes a way to develop the ability to solve real problems, rather than simply to complete a course someone else has designed.

That is why I think the question worth reconsidering in the age of AI is how heavily we should depend on schools for our learning, rather than whether we should still attend them.

Schools still matter, and they still deserve to be taken seriously. But if you gradually realize that you are not the kind of person traditional education is particularly good at recognizing, you may not need to spend ever more time proving that you can perform better within its selection process. Pushing an exam score from 90 to 95 may well have value. It will not always be a better use of that time than developing an ability you actually need.

Schools need standardization because society relies on them to educate large numbers of people and to keep identifying those with exceptional abilities. An individual’s learning does not have to follow exactly the same logic.

You can still attend school, take exams, earn qualifications, and complete the necessary training with care. At the same time, it is worth taking more ownership of your learning: stop assuming that anything school has not taught you must wait until later, or that completing the learning school requires automatically means you have acquired the abilities you actually need.

Schools need to find society’s next genius.

For the rest of us, who may never turn out to be that outlier, the age of AI may offer a new opportunity. We can begin answering a different question sooner, instead of waiting for an institution to tell us what to learn next:

What do I need to learn to do the things I actually want to do?