One widely replicated finding has remarkable implications for education across the globe. And yet, no one really talks about it. It's the Flynn Effect.

There has been a massive increase in "IQ" as measured by intelligence tests over the last century. This is across countries, north and south, east and west. It does not have to do with access to schooling. In fact, if anything, it might have the opposite effect.

Now, the intelligence quotient is a dubious and reductive measure. It measures "something" related to what we recognize and label as intelligence or smarts. Intelligence itself is fuzzy, and the link to IQ is fuzzier still. Yet enough companies around the world have found it useful to retain intelligence tests as a filtering mechanism. That does not make them unimpeachably good, but there's a signal in there, and it's worth figuring out what it tells us.

The surprising truth about what drives intelligence

Why would intelligence increase so dramatically across the board? First, it is clearly not fixed. There's no way you could see such enormous increases in the space of a few decades. This means it is something that can be acquired. That's what's called the growth-mindset these days.

One obvious candidate is schooling. You would think that what the Flynn effect tells us is that it is the schooling system that has been responsible for these gains. But here's where things get perverse: it's actually leading to worse outcomes. Instead of letting kids come in contact with the world or a close simulation of it, we are shielding them and exposing them to narrow, artificial drills.

It turns out increases were seen even among kids who did not have access to regular conventional schooling. What else could explain it? It's the culture, but not in the narrow definition of that word, but the wider definition encompassing our physical world and the world of ideas. It's the web of information, knowledge, know-how and understanding we all live inside.

Our lived world is now enormously complicated relative to the pre-industrial era. There are symbols and abstractions everywhere. From traffic signals to restroom signs. And understanding them matters. And being steeped in them allows us to soak them in.

In other words, the world you inhabit teaches you. By experiencing and exploring it, we learn and become smarter.

What's also been observed is what's called the reverse-Flynn effect. A dip in the last two decades. That is actually the clearest signal that it is not universal education that is making the difference. It is the offline web we live inside.

We have replaced the world with screens, first with TVs and then smartphones. While these bring information, they do not allow us to actively explore and engage and construct meaning. It makes passive consumption easy. And that does not lead to learning.

Education as apprenticeships

What are the lessons? We need our education and schooling systems to understand that learning happens with active participation in our world. And learning systems that capture the world in its totality make a lot more sense than narrowly defined skills.

Only superficial learning is possible when kids are drilled on these. It is very easy to pay lip service to holistic, meaningful learning. I am sure every school has those words plastered on their web page. But how do you bring those to life?

We need to reimagine education as apprenticeships—where children engage with the complexity and richness of the real world, not sanitized versions of it.

Two metaphors for meaningful learning

How do we build on this understanding? We find two metaphors that are hugely useful in understanding how we learn and how we can enable meaningful learning. The Whole Game. And The Long Journey.

The Whole Game

We learn by playing the whole game, not by polishing up individual skills in isolation. This is what we call the Whole Game approach. A game is played in its entirety. We never play a game by first learning narrow skills. We learn by playing the game, first in easy mode, and then advanced.

A game is about ACTION. Learning is fundamentally active. It is not the passive consumption of information. It is the active construction of knowledge, meaning, and joy by engaging with the task at hand.

Approach subjects as we are learning to play the game at a junior level. This requires a deep understanding of the game, and then creating a version that simplifies both the game AND the tools to developmentally appropriate stages.

A game naturally has the idea of error built into it, but in a way where there is no labeling or narrow judgment of failure. Miss scoring a goal, that's feedback. Miss a tackle, that's feedback. But it is never too narrow.

The Long Journey

And this brings us to our other metaphor. The Long Journey approach. Not everything is goal oriented. Sometimes it is about exploring the world with your unique perspective. It is an act that unfolds over time and space, often many years and cities and countries.

Learning here is not guided by the notion of error, but by experience. By seeing new aspects of the world, internalizing them. The journey can take time. Insights and understanding do not have to appear right away.

This fits well with balance. The game perspective focuses on the here and now and the journey perspective allows us to keep an eye on the future. These are different time and space scales.

AI as a mini-world maker

We are now in an era where we can bring the world into the classroom in a way that can appeal to each child. AI makes this possible. The astonishing Flynn effect style advances can be greatly accelerated with AI by allowing the world to be explored.

The Flynn effect is intelligence rising 30 points over a century. Imagine such an increase over a decade. That's possible.

We are not saying the AI will teach us. What makes us human is we like to learn from humans. And we teach each other collectively. We share concepts and care in language. AI cannot do that yet.

But they can build these learning worlds for us, which we can explore along with our grownups. The promise of AI is not automating assessments and quiz generation but creating these worlds. This, if done right, is going to be spectacular.