Kieran Calder of UBP said China’s AI models are cheaper yet comparable to more expensive U.S. rivals, a blunt reminder that the race for machine power is being driven by institutions with deep pockets and even deeper control over the infrastructure. He also outlined potential risks facing OpenAI while staying constructive on the company’s long-term AI infrastructure and capex buildout.
Who Holds the Levers
Calder’s comments put the hierarchy on display. UBP is analyzing OpenAI scenarios while the company keeps pouring money into infrastructure and capex, the kind of spending that concentrates power in the hands of firms that can afford to build at scale. The discussion wasn’t about ordinary people shaping the future of AI from below. It was about which corporate actors can keep up, which ones can dominate the field, and which ones might get squeezed as the costs rise.
China’s AI capabilities came up as part of that calculation. Calder said its models are cheaper yet comparable to more expensive U.S. rivals. That matters because price and scale aren’t neutral technical details. They decide who gets access, who gets locked out, and which companies get to set the terms for everyone else. The whole setup rewards the biggest players and leaves the rest to scramble.
The Risks They’re Gaming Out
UBP discussed scenarios for OpenAI and the broader AI story, with Calder identifying both risks and opportunities as AI develops across sectors. That’s the language of boardrooms and capital allocators, where the future gets reduced to scenarios, exposure, and upside. The people who will live with the consequences don’t appear in the model.
OpenAI’s long-term AI infrastructure and capex buildout also drew a constructive reading from Calder. That’s the polite financial version of saying the company’s power depends on massive investment in systems most people will never control. Infrastructure sounds clean. Capex sounds technical. But behind the jargon sits a familiar arrangement: a few firms accumulate the means, and everyone else is expected to adapt.
The article offers no grassroots response, no mutual aid, no horizontal organizing. Just the usual top-down conversation about what the powerful might do next, and how analysts can price the fallout. The public gets the bill in the form of higher barriers, tighter control, and a future shaped by corporate decisions made far away from the people affected by them.
What the AI Story Really Means
Calder framed the discussion around both risks and opportunities as AI develops across sectors. That broad reach is exactly why the stakes are so high. When a technology spreads across sectors, the institutions that own it don’t just sell a product. They gain leverage over work, communication, and daily life.
The comparison between cheaper Chinese models and more expensive U.S. rivals also shows how the AI race is being organized as a contest between power blocs, not a public good. The language of competition hides the same old structure: capital concentrates, infrastructure expands, and ordinary people are told to marvel at the progress while the terms are set elsewhere.
UBP’s read on OpenAI stays constructive, but the facts in the discussion point to a system built on expensive buildouts, corporate rivalry, and control over the machinery that increasingly shapes the world. The models may be getting cheaper. The power behind them isn’t.