a16z: OpenAI wins by creating new kinds of customers, not by having the best models

a16z Newsletter · rss · 2026-09-28

a16z's newsletter argues OpenAI's moat isn't models, chips, or even product — it's creating new kinds of customers and having the most durable distribution strategy.

The four levers for durable frontier-scale AI business:

Levers 3-4 are dead ends: model switching is trivial (a Red Queen's Race) and price-performance follows scale. So it comes down to 1 and 2, where OpenAI leads on both.

Why OpenAI keeps finding "the next thing":

General solutions depend on distribution: going generic is costly and against focused-product best practice, but only general exposure reveals the breakthrough primitives. Standalone products control the experience; partnerships give reach but lose economics and trust; platforms are slow but most durable.

Conclusion: intelligence is now everywhere — whoever awakens new behavior and wins distribution wins, and OpenAI is playing the long game correctly.

Original post →

More from Companies & People

Companies & People channel →