Notes, explanations and build logs
Writing
Long-form essays about AI systems, engineering leadership and building Tell Me Ted.
-
Why Our First AI Architecture Was Too Complicated
Nothing in the first architecture was obviously wrong. That was what made it difficult to replace.
-
How We Turn Product Ideas Into Codex Tasks
“Add page-turn detection” sounds like a feature request. To a coding agent, it is an invitation to make dozens of unstated decisions.
-
The Most Important Product Decisions Were Not Technical
The biggest Tell Me Ted decisions were not which model or API to use. They were decisions such as whether a child should press a button.
-
Hardware Is Where Software Assumptions Go to Die
The build succeeded. It also warned that the firmware had almost no useful headroom.
-
Building AI for Children Changes the Engineering Brief
A child is not an adult user with a shorter prompt and a friendlier interface.
-
From Chained AI Calls to a Realtime Voice Experience
A child does not care how many APIs are involved. They notice only that the teddy did not answer.
-
What’s Inside an AI Teddy Bear?
An AI teddy bear is not a model in a soft enclosure. It is a small networked computer, camera, audio system and battery product competing for the same limited space.
-
Codex Can Write the Code. You Still Own the System.
The refactor worked. The code was dramatically simpler. One problem: a lifecycle-critical maintenance capability no longer survived the change intact.
-
Building an AI Hardware Product With ChatGPT as a Daily Collaborator
ChatGPT did not become another member of the team. It changed what our small team believed it could attempt.
-
Building a Teddy Bear That Can Read With a Child
Most AI products begin with a text box. Ours begins with a child opening a book beside a teddy bear.