(Even if it’s not directly about learning this time – indirectly, it’s about nothing else.)
Sometimes the most exciting developments are not directly learning-related topics, but they change everything we know about learning, working, and future viability.
Just like now: AI is spreading across the globe, revealing a fascinating pattern. On the one hand, its use is concentrated in a few countries and corporations, while on the other, it is simultaneously becoming massively accessible. Concentration meets democratization, and this very tension shapes where the world of AI – and thus learning – is headed.
Concentration: Small Countries, Big Usage
The new Anthropic Economic Index (September 2025) shows: The strongest AI usage occurs in small, highly digitized countries like Israel, Singapore, Australia, New Zealand, and South Korea. While the USA remains the largest single market, smaller states extract more from their tools relative to their population – an indication of the quality of digital competence.
Companies also vary in their progress: Around 77% of API usage is automated – meaning AI is deeply embedded in processes, not just used as an assistant. This depth of integration is the real secret to success and a lesson for learning too: It’s not about “using AI,” but about integrating AI.
Democratization: When Access Becomes the Norm
In parallel, global usage is exploding. According to the NBER Working Paper (September 2025), around 10% of the adult global population actively uses ChatGPT, and in low-income countries, usage is growing four times faster than in wealthy nations.
Two reasons drive this:
- The cost per token has fallen 280-fold since 2022 – AI is now everyday infrastructure.
- Easy to use in the browser or via app – no prior knowledge required.
Result: AI has arrived everywhere. But: Value creation remains concentrated. Or, looking at it from a learning perspective: Knowledge becomes democratic, but competence does not automatically follow.
Two Worlds, One System
What seems contradictory at first glance is a classic pattern of technological maturity:
- Breadth vs. Depth: AI is everywhere – but real impact occurs where it is understood.
- Augmentation vs. Automation: Mature markets use AI as a partner, less mature ones as a replacement.
- Compute as a Power Factor: Computing power is the new raw material – whoever controls it controls progress.
For education and HR development, this means: We must not only enable people to work with AI, but teach them to think in AI contexts.
Europe: Between Ambition and Connection
Europe plays an intermediate role in this geography. Scandinavia and the Netherlands are at the forefront, but the entire EU has only about 5% of global AI computing power. Projects like “AI Gigafactories” aim to change this, but until then, the fact remains: We cannot scale everything, but we can empower people to make the most of access.
The Next Wave of Democratization
The question is no longer who uses AI – but who truly understands it.
To achieve this, we need:
- Local data & languages – AI must know our contexts.
- Digital learning competence – understanding not just tools, but principles.
- Open models & access – so that learning does not depend on capital.
Conclusion: The Divided World as a Learning Task
“AI spreads like water – but it flows downhill.”
Access alone is not enough. True democratization means that people understand how to use AI to learn, design, and create new things. And that is – regardless of whether we are talking about economics, politics, or education – the greatest learning task of our time.
