Not topics — mapped tensions, still open. That is why they deserve close attention: they are likely to move considerably in the time ahead.
Moat Radar does not cover the AI industry by subject. It covers by tension. Every piece of news, every announcement, every earnings report is read through one of these six structural questions — still without a definitive answer. They are not categories that resolve over time. They are the axes around which the entire sector is organising itself right now.
The six split across two sides of the same equation: what drives Demand (who will pay, and why) and what sustains Supply (who is building, and at what cost).
Model vs. Channel
Who captures the value of AI: the one who builds the most capable model, or the one who controls distribution — the channel through which the end user reaches it? Models are becoming commodities faster than many expected. Distribution may be worth more than intelligence.
The Corporate Dilemma
Who manages to turn AI into results — productivity, revenue, margin — and who will only spend on licences without ever converting that into real gains. The technology matured before companies developed the capacity to absorb it. That gap is where the money is decided.
The Humanoid Race
Not only about robots with human faces. It is about any physical device that proposes to replace a function currently performed by a human — from industrial robots to self-driving cars. The question is never purely technical: it is when the cost drops far enough, and who is first to trust a machine with the function.
Capex vs. Bubble
The industry is investing at a trillion-dollar pace. The question is whether that investment reflects already-validated real demand, or whether it is an arms race where no one wants to be the first to slow down — even without certainty of returns.
Bottleneck Economics
Energy, chips, water, land. The capacity to train and run ever-larger models runs into finite physical resources. Whoever controls the bottleneck controls the pace of the entire sector — regardless of who has the best model.
How Long Will We Stay in Control
The frontier of model capability advances faster than the human capacity to audit, understand and contain what they do. This is not a question about science fiction — it is about practical governance, today.