Notes on building the AI-native web.
Page 11 of 17
The first three jobs to give an AI employee, ranked
Ranked by how quickly they pay back and how cheaply they fail. The flashy job is fourth at best.
Translating your content with AI
Good enough to make translation affordable, not good enough to publish unreviewed. Where the line falls, by content type.
The about page nobody reads
Except they do — it’s often the second page a serious buyer opens. It’s just that yours is written for nobody in particular.
Definition pages: the most quotable thing you can publish
“What is X” is the most common shape of question there is, and a good definition is the most liftable passage there is.
Fitting an AI employee into your helpdesk
Your ticketing system already has the workflow, the history and the reporting. Working inside it beats bolting something alongside.
Search is the feature everyone underestimates
It looks like a text box and a query. It is ranking, tolerance for error, and knowing what people actually meant.
How to instrument software so you can add AI later
The reason AI features are hard to add to existing systems is almost never the model. It’s that nothing was ever written down.
Should customers know they’re talking to an AI?
Yes — and the interesting question is how to say it in a way that helps you rather than apologising for it.
How often should you redesign your site?
Less often than agencies suggest and more often than you’re doing. The trigger should be a reason, not a birthday.
The FAQ page is back, and it works differently now
SEO spent a decade telling you to delete it. In an answer-shaped web, the format is suddenly the right one — if the questions are real.
Does an AI employee change who you hire next?
Not by removing a role. By changing which part of the job is scarce, which changes what you should screen for.
Handling personal data in an AI feature
Customers type things you never asked for. Four decisions to make before that becomes a problem you can’t unwind.