AI is discussed almost entirely as a technology story — which model is best, which prompts work, which jobs disappear. But the questions that will actually shape the next decade are structural: who will produce the energy AI needs? Who will own the infrastructure it runs on? Who will control the data that trains it? And who will end up dependent on someone else's technology to compete?
For countries and companies alike, AI is becoming a matter of economic competitiveness and strategic independence — not just adoption. That stack runs deeper than most conversations acknowledge: from the energy that powers compute, to the infrastructure it runs on, to the agents that will increasingly act on our behalf. We call this full stack the agentic cloud — infrastructure built from energy up to the agent layer. Owning it, rather than renting it from someone else, is the difference between AI capacity and AI dependence.
This session steps back from the tools to look at the forces: energy, infrastructure ownership, and sovereignty — and what it actually means to build, or depend on, AI capacity in this new landscape.