Source: approximations, Feb 2026

Look at the image above for a second. If its approximations are even roughly right, the message is startling. The vast majority of humanity still has not meaningfully used AI. A smaller group is using free chatbots. A tiny sliver is paying for AI. And an almost microscopic number of people are using AI for coding scaffolds and more advanced applied workflows.

That is the paradox of this moment. We are still extremely early in adoption, yet the implications already feel existential. The technology has arrived before society, companies, and institutions have properly absorbed what it means. In other words, the world is still near the beginning of the usage curve, while already standing inside the strategic consequences.

This is what makes the current moment so unusual. Normally, existential shifts feel obvious only once adoption is everywhere. With AI, the opposite is happening. Even at this early stage, the pressure is already visible in software, services, education, media, research, and knowledge work more broadly. Entire sectors are starting to realise that this is not just another tool layer. It is a new operating layer.

In software, the change is already tangible. Writing code, debugging, refactoring, scaffolding, agentic development environments, and AI-assisted product design are changing how work gets done. But software is only the first sector to feel it clearly because it is closest to the tools. The bigger story is what happens as these capabilities move into professional services, operations, finance, logistics, healthcare administration, education, and any environment where people spend their days moving information between systems, decisions, documents, and workflows.

That is why the image matters. It shows just how much of the world is still untouched by advanced AI interaction. If only a small percentage of people are using AI today, then most of the economic, organisational, and social consequences are still ahead of us. We are not late. We are early enough that the infrastructure is still being built, the standards are still forming, and the winning operating models are still up for grabs.

And yet, despite being early, companies cannot afford to be casual. The reason is that the landscape is no longer defined only by better models. The surrounding infrastructure is maturing fast. Tool connectivity standards are improving. Orchestration frameworks are getting more usable. Memory and state layers are becoming more robust. Durable execution is making it easier to move from experiments to live workflows. The stack is becoming real enough to shift AI from pilot projects into production systems.

That changes the question for leadership teams. The question is no longer, "Should we play with AI?" The question is, "How will our company need to refactor as these tools become embedded in real operating workflows?" That is a much more serious question. It touches product design, service delivery, headcount shape, margins, customer expectations, defensibility, and speed of execution.

This is why the moment feels existential even while adoption is still low. The size of the current user base understates the size of the coming impact. Small numbers can still signal a very large wave when the capability curve is steep, the infrastructure is improving, and the cost of intelligence keeps falling. What looks niche today can become normal much faster than incumbents expect.

For investors, this means the real opportunity is not just in consumer adoption numbers today. It is in understanding the second-order effects. Which sectors are most exposed? Which workflows will be rebuilt first? Where will new value accrue? Which companies will use AI to compress cost, expand capability, or redesign the economics of their category? Capital needs a route to the frontier not because the market is mature, but because it is still forming.

For operators, the image is a reminder not to confuse low current penetration with low future importance. Most of the world may still be outside the advanced AI economy, but that does not reduce the urgency. It increases it. The organisations that learn fastest now, while things are still fluid, will be in a far better position than those who wait for the shift to become obvious to everyone.

So the right response to this image is not hype, and it is not complacency. It is clarity. We are early enough that there is still time to learn, position, and build. But we are far enough in that the implications are already profound. That is what makes this moment so important.

This is what the first chapter of a major reset looks like.