The player moves into a bigger box, on purpose.

This week the software moved into a bigger box: a red GuSi radio enclosure that plugs into the wall. It isn’t the product. It’s a tidy bench, so the software can get finished without the hardware getting in the way.

The first handheld is retired for good. It had started to run hot, and the smell of warm plastic coming off a speaker isn’t one to have in a kid’s room. I never found the cause, so that unit isn’t going back into use. The next handheld will need a different design.

The red box uses the GuSi radio’s own parts, nearly one for one. I went with it because someone had already made it work. That lets me focus on the software, and later a companion app, without debugging hardware at the same time. It also has what the handheld was missing: a real speaker, a knob, and a power button, so it shuts down cleanly instead of being unplugged mid-story. This week I also started having AI write some of the code.

The red box open on the bench, October 15, 2025.

Last week I worked out what the smallest version worth finishing looks like: the hardware together, playing a preloaded list of episodes. And what done enough to be real looks like: a kid who uses it regularly, and one unit posted to a friend. Everything else can wait.

I’ve built things before that didn’t become products. This one is different. It’s for my son, he’s already helping, and we test it together. The player got a name this week too, after him.

Next: make each control do more than one thing, and clean up the code so it’s ready for an app.