Context:
2026 is when the AI story stops being about a fairytale of infinity and starts being about proof. Proof of economics. Proof of defensibility. Proof that scale compounds instead of hollowing out the core. Who will survive when the narrative meets reality?
Meta
Can they turn it on beyond ads? Or are they permanently stuck inside their own garden? AI helps only if it expands the surface area of value—not if it just makes the feed a little better.
OpenAI
Can they stay ahead on models and make money? Being first doesn’t matter if the marginal economics never work. Frontier tech without margins is a research project, not a company.
Google
Can AI drive real economic lift across the stack faster than it eats search? This is a race between reinvention and self-cannibalization—and the clock is very real.
Nvidia
Can they stay dominant on chips, or do TPUs and custom silicon start to matter? And more importantly: can they turn an absurd cash position into actual defensibility, not just “we’re winning right now”?
Apple
Do they even care? Or are they effectively short the entire AI arms race—letting everyone else light money on fire while they capture value quietly at the platform and hardware layer?…
SpaceX
Can they keep accessing capital to do the thing? And can they convince people that ideas like GPUs in space aren’t science fiction, but inevitable infrastructure
ROW (everyone else)
Who can move out of “experimental” AI budgets and into core operating spend? And once AI is table stakes, who actually has margins and moats?
Trade, overall
Will people keep believing? Because belief is still doing a lot of work. When that cracks, it won’t unwind slowly.