General-purpose AI may make it easier both to build a competitive advantage and to help others catch up, argues a new essay from Quasi Convex Union. That tension could push frontier AI labs towards selling their own services, even as their tools help customers build alternatives.
Watch Desk analysis
What happened
The essay, The Economics of Intelligence, argues that AI can lower production costs and increase people’s capabilities. Its central idea is that intelligence has an “inductive” effect: as AI raises the capabilities of customers and competitors, it may also make frontier-level work easier to reproduce.
The author argues this could make durable competitive advantages harder to maintain than in industries where capital or scale provide a firmer moat. For AI labs, the resulting pressure may be to move downstream and compete in the markets their own technology helps customers enter.
Why it matters
That is a useful way to frame the tension between selling a general-purpose tool and competing with the people who use it. If customers can use AI to build more of their own capabilities, access to a model may not guarantee a lasting advantage for its provider.
The essay presents an argument about how the economics could work, not evidence that every lab will follow the same strategy. Still, it puts a sharper question behind the familiar race to build better models: who captures the value when powerful tools make capability easier to spread?
Our read
The interesting claim is not simply that AI makes work cheaper. It is that AI could make competitive advantage more contagious, including to a company’s customers. That is a neat paradox, and one worth testing against how these businesses actually behave. The moat may be less a castle wall than a very well-stocked workshop.
What to watch
- Whether frontier AI labs expand into products and services that compete with their customers.
- Whether customers use general-purpose AI to build credible alternatives to established providers.
- What evidence emerges about whether AI makes market advantages easier to reproduce, or merely changes where they come from.
Discussion spark: If AI helps customers build capabilities that once belonged to their suppliers, should frontier labs compete downstream or focus on being the best toolmaker?
Sources and evidence
- The Economics of Intelligence | Quasi Convex Union (16 November 2026, 00:00 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.