Product judgment is the new moat
As models commoditize, the defensible edge moves from the technology to the decisions around it. Notes on why product judgment — not the model — is becoming the moat.
For a while, having an AI capability was itself the differentiator. That window is closing fast. The same models are available to everyone, and they get cheaper and more capable by the month. When the ingredient is a commodity, the advantage moves to what you do with it — and that's a judgment problem, not a technology one.
Same model, different products
Give two teams the identical model and they will build wildly different products. One points it at a vague, broad task and ships a chatbot nobody uses; the other points it at one sharp job, designs for its failure modes, and places it exactly where the work happens. The gap between those two outcomes is entirely judgment: what to build, what to cut, where the AI belongs, and how to handle it being wrong.
Judgment shows up as decisions, not features
We think of judgment concretely, as a series of decisions most teams make on autopilot:
- Which task to point the AI at — the narrow, painful one, or the broad, impressive-sounding one.
- Where it lives — inline in the workflow, or bolted on as a separate tab.
- What happens when it's wrong — designed for, or ignored until users find out.
- What to leave out — the discipline to not ship ten mediocre capabilities in place of one great one.
None of these are model choices. They're product choices, and they're where the durable difference is made.
Why this is defensible
A feature is easy to copy; a stack of good decisions compounding on each other is not. Judgment produces products that fit their users so specifically that a competitor bolting the same model onto a broader product can't match the feel without breaking their own. That's a moat — not because the technology is secret, but because the decisions are hard-won and hard to reverse-engineer.
What it means for founders
The practical implication: stop competing on "we use AI too," which is now table stakes, and start competing on sharpness — a narrower audience, a better-handled failure mode, a workflow the incumbent won't touch. This is the thread behind much of our thinking, from finding your wedge to validating the idea before a line of code.
Want the landscape and the wedge mapped for your idea? The free Market Need Analyzer and Competitor Landscape Teardown are a fast place to start.
More research
The trust gap: why users abandon accurate AI
A pattern we keep seeing: AI features get abandoned not because they're wrong, but because users can't tell when they're right. Notes on the trust gap and how design closes it.
Read →Research note · 5 minAI-as-a-tab is a dead end
The bolt-on chatbot ships fast and demos well, then goes unused. A note on why AI belongs in the flow of the work — and what changes when you move it there.
Read →