────────────
average · forgettable
vs
specific · yours
Introducing Digitopia Labs
We build the voice, audience and positioning layer that makes Claude produce on-brand work from the first prompt. Ready-made skills, an AI Audit, and the full Blueprint build.
Read the article →Why AI writes like a brochure for advice firms, and what to give it instead
Most advisers try AI once, get something that reads like a brochure, and stop. The problem isn't the prompt. It's that nobody gave the model your approved wordings, your audience, or the things you'd never say.
I built an app that tells dad jokes. It's the most serious thing I've made.
Landing a dad joke reliably is a brutal voice test. Pulling it off is the exact voice infrastructure I build in the first stage of an AI Blueprint.
I lost 40 chillies to a bad build. So I started again, properly.
A bad build wiped 40 of my chillies. Rebuilding the Chilli Index with Claude taught me why the plan, not the prompt, is the real work - and what that means for an AI Product Sprint.
Why your AI content sounds like everyone else's
Most AI content is bad, and it's almost never the AI's fault. The model didn't fail - it did exactly what it was asked, which was nothing. Here's the layer underneath the prompt that changes everything.
Get The AI Layer
A monthly note on building AI that works - the posts, plus the bits that don't make the blog. No spam, unsubscribe anytime.