This is why I added an ai index page to my site. It’s heavier on citations, niche q&a, and technical specs that would bore the sh*t out of the average person. I also ran it through my app’s own AI grader, which actually suggested adding even more details, so there you go.
This reframed something I’ve been thinking about for a while. The companies that will survive agent-mediated evaluation are the ones whose value was never in the friction to begin with. It was in enabling a better output. I spent years at LiveRamp where most clients weren’t using us because of the UI and many may not even know they ever used LiveRamp at all since it was baked into many other platforms. They were just using us as infrastructure to make something possible that wasn't before. Agents will find those companies immediately.
I of course agree that we should be thinking and preparing for this. At the same time, with a lot of the (tech and consumer product) companies I’m advising, they’re not even doing the “101” stuff right. Soooo there’s a lot of catching up to do. And with consumers any time I exit Substack or LinkedIn, I feel like only a small subset are even thinking about AI at all. It’s very bizarre / contradictory and I can’t quite describe it. I can’t see the consumer gen pop or even business leaders outsourcing tool evaluation parts to agents. Do I think it’ll happen? Yes. But unclear on the timeline if people and executives are still using AI as a chat/search feature. Regardless, we should be thinking and developing what you’ve suggested. And it’s a good practice for codifying decisions and rationale too. That’s actually the part of AI I enjoy the most - REALLLYYYYY thinking through how to create a skill and breaking it down to a strong .md file an agent can learn from. Thanks for taking the time to write this!
Just had a convo last week that a client of mine needs to prepare to expand their TAM - time to serve the robots
This is why I added an ai index page to my site. It’s heavier on citations, niche q&a, and technical specs that would bore the sh*t out of the average person. I also ran it through my app’s own AI grader, which actually suggested adding even more details, so there you go.
great stuff!
This reframed something I’ve been thinking about for a while. The companies that will survive agent-mediated evaluation are the ones whose value was never in the friction to begin with. It was in enabling a better output. I spent years at LiveRamp where most clients weren’t using us because of the UI and many may not even know they ever used LiveRamp at all since it was baked into many other platforms. They were just using us as infrastructure to make something possible that wasn't before. Agents will find those companies immediately.
Talked about last week:
https://designisinthecode.com/posts/2026-04-14-the-web-has-a-new-user/
Or
https://jjoaquim.substack.com/p/the-web-has-a-new-user-it-isnt-human
I of course agree that we should be thinking and preparing for this. At the same time, with a lot of the (tech and consumer product) companies I’m advising, they’re not even doing the “101” stuff right. Soooo there’s a lot of catching up to do. And with consumers any time I exit Substack or LinkedIn, I feel like only a small subset are even thinking about AI at all. It’s very bizarre / contradictory and I can’t quite describe it. I can’t see the consumer gen pop or even business leaders outsourcing tool evaluation parts to agents. Do I think it’ll happen? Yes. But unclear on the timeline if people and executives are still using AI as a chat/search feature. Regardless, we should be thinking and developing what you’ve suggested. And it’s a good practice for codifying decisions and rationale too. That’s actually the part of AI I enjoy the most - REALLLYYYYY thinking through how to create a skill and breaking it down to a strong .md file an agent can learn from. Thanks for taking the time to write this!