MechanismArticle

A crowded AI frontier won't make AI a pure commodity, but it will drive down the value of the best models sold on a meter, so AI profits will accrue mostly to those who build the best products on top of the models and those who own the compute.

The author explains why crowding erodes the value of metered model sales without fully commoditizing the technology, shifting profits to product-builders and compute owners. ✦ AI generated

Alex Kantrowitz · Big Technology · 2026-07-31 · original ↗

Share card ↓Export .md Export .txt

A crowded AI frontier won't make the technology a pure commodity, because specialization and compute resources will always give certain AI labs edges in certain areas, but it will drive down the value of the best AI models (at least if your plan is to sell them on a meter). The profits in AI will thus accrue mostly to those who build the best products on top of the models, and those who own the compute that enables them to serve these products.

Read full article ↗excerpt · fair-use quotation

✳ AI models ✳ compute

Around this claim

Evidence · 2

The market for frontier intelligence has consolidated into a duopoly of Anthropic and OpenAI that can charge premium prices, while everything else is commoditized to the point where you can only charge for compute, inference, and consulting — not the model itself.David Sacks · All-In Podcast · conf 90%The business realities are already pushing AI labs upmarket into competing with their own clients — as seen in Anthropic's Claude Code and Claude Design and OpenAI's ChatGPT superapp with Codex — and if the labs hoard their best models, they will have a strong chance to build the best AI-native products and outcompete struggling legacy-software incumbents.Alex Kantrowitz · Big Technology · conf 60%

Extends · 1

The AI race is no longer a one- or two-lab competition: open-weights competition from China and profit-insensitive initiatives from Google, Meta, and SpaceX mean there may now be five, six, or seven labs at the top.Alex Kantrowitz · Big Technology · conf 80%

This moment responds to

supports → There may eventually be six or more frontier AI developers rather than just two or three, which would dramatically cut model-layer margins and shift the real value in AI toward the best products built on top of models.Alex Kantrowitz · Big Technology extends → A frontier lab's strategy should follow from the competitive landscape: with one or two dominant players, sell models at a high markup; with several tied labs, sell products; and with a narrow lead like OpenAI's and Anthropic's today, pull up the ladder and build products competitors cannot match with older tech.Alex Kantrowitz · Big Technology explains mechanism → It is time to delete the assumption that frontier AI labs will always license their best models; increasingly it may be in their interest to hoard the intelligence and sell the products built with those models, not the models themselves.Alex Kantrowitz · Big Technology supports → The single biggest unknowable factor determining who captures value in AI is the market structure of the model companies — how many frontier labs compete determines whether token prices stay high or fall.David · a16z Podcast