Almost Entirely Human

There are four levels of AI coding available today, from fancy autocomplete to full agentic engineering. The interesting stuff starts at Level 2. And Level 3 is where it gets weird.

I built a web app called Palm Springs Eats AI, where you can chat with a map of restaurants near the convention center.

TL;DR - A friend built me a Chrome extension that generates alt text for images using the surrounding page content for context. It took less than an hour.

I demo how to build an Experience Builder widget using Claude Code and it works!

This week, to celebrate one year of AEH, Holly and I review our favorite posts and I make some comments about what it is like writing a newsletter every week!

I used Claude Code to build a voice dictation app called Spiel. It uses OpenAI's Whisper for transcription, has an optional AI cleanup step with a prompt that I can control, and it only took me one evening to make!

Join me while I build GeoScribble, a draw the country from memory app!

This is a story about how I build apps. It's about a process I've been refining that enables quality and speed. I want to walk you through exactly what I did because it's repeatable, and I think more people can and should be building prototypes before meetings instead of just talking about ideas.

Large Language Models don’t actually "remember" your conversations. Every message you send includes the entire conversation so far. When that gets too long, applications have to make choices about what to keep, and sometimes that goes sideways.

This is what happens when you let an engineer use AI to try to play Jingle Bells. Spoiler: it's bad. But also kind of wonderful?