Becoming the Robot Default
The next great distribution advantage is becoming an AI's default choice.
Ever notice how AI-built websites all look the same?
The same rounded cards and little status badges appear everywhere. The aesthetic has become something of a joke. But the sameness reveals something interesting: robots have preferences.
In February 2026, Amplifying studied 2,430 Claude Code responses across 20 technology categories. The prompts didn't name products. They simply asked the agent to solve problems.
The results were striking:
- shadcn/ui: 90% of UI component selections.
- Tailwind CSS: 68% of styling selections.
- Stripe: 91% of payment selections.
- Resend: 63% of email selections.
- Zustand: 65% of state management selections.
Meanwhile, Redux received zero selections in 88 state-management decisions. Jest was selected in just 4% of testing decisions.
The robots knew the incumbents. They preferred something else.
It's just a single study of Claude Code's choices, and doesn't prove real-world adoption. But it reveals the potential of a powerful new distribution channel.
From ten blue links to one decision
Google spent years earning our trust by consistently delivering good search results. That trust transferred to the websites it ranked. Appearing among the ten blue links conferred authority; being the first result became an endorsement.
AI agents take this a step further: they can act on their own recommendations.
Ask Codex to build an app and it chooses the technologies, installs the dependencies, writes the code, and delivers something that works.
Sometimes those decisions never reach the operator. Other times, a less technical person sees the choices and says "lgtm 👍".
Either way, the robot has made the decision.
And that changes distribution. Google handed you a visitor. An agent can hand you a customer. Discovery, selection, and adoption can all happen within a single workflow.
What makes a robot confident?
I suspect the underlying mechanism is machine confidence: how certain an agent can be that choosing a product will lead to a successful result.
Products that are easy to use, well documented, and widely represented in public code repositories give agents plenty to work with. Every developer publishing a working implementation adds another example that future models might learn from.
Familiarity reinforces preference, and preference drives further adoption.
The strategy isn't radically different from SEO. What's good for humans and good for Google is generally good for robots.
But the tactics change when the ambition becomes: Get this robot to choose my product for this class of problem.
As agents take on more of the work, their preferences increasingly determine which products get used.
Become the thing the robot reaches for first.