Chore Tracker
Up until this point, I had felt mostly ambivalent about AI tools. The odd recipe request to Gemini or car research questions for ChatGPT was as deep as I had gone.
The ethical and environmental concerns are still something I wrestle with, and the economics of it are certainly questionable.
But then I read a fairly motivating article about the state of AI (Something Big is Happening by Matt Shumer). The urgency I felt after reading it was tempered slightly by another article (Something big “might” be happening by Elvis Hsiao). Either way, I thought I should make a more concerted effort to understand the capabilities of these tools, given the industry I work in (software!).
Diving in
I got a subscription to the cheapest tier of Claude Pro and the first thing I set out to do was build a simple tool I had been thinking about for a while: a chore tracking app for my kids to log their daily and weekly duties. Using the chat feature, I described what I wanted very briefly and provided instruction for a few updates. It guided me through how to set up a database and deploy it as a standalone HTML file.
All of this took about an hour total and zero designing of screens on my part for the first iteration. Four or five detailed prompts, a couple of refinements, and a bug fix. A few more enhancements and bug fixes took up another two or three hours.
Reflecting on how the process would’ve gone just a few years ago, I likely would have taken a day or two to sketch some flows and wireframe some screens (and likely would’ve missed several that Claude captured at the start). UI design would’ve taken a few more days of exploration. And development! I’ll conservatively estimate weeks for that. Realistically, this would have been many thousands of dollars in billable time for a small agency. Instead, it was a few hours of my time and about $30.
I was amazed. And truthfully, a bit off-balance. It really made me wonder how this would impact my job, the industry, and well, everything (and it continues to).
What’s that saying about best laid plans?
I was so excited to share this with my family! I drafted an introductory text with the link, the PIN code to access it, a spreadsheet to capture bugs they found…and then?
Nothing.
I didn’t expect my kids to be super jazzed by it, but I thought they might find it somewhat interesting or useful. No. I practically had to force them to try it. I showed them how I was tracking my chores. After a few weeks of essentially no uptake, I prompted them again to use it. “Your allowance is contingent on seeing these chores tracked,” I warned.
Still nothing. I considered adding features to tie this to their allowance (money is a motivator, right?). But then I slowly came to a sad realization.
An efficient lesson
I don’t subscribe to the idea of needing rigid processes to build software. But I missed one really crucial step. I did not engage my kids in building this. I didn’t consider including them in the process. I didn’t ask them any questions at all. I was so enamored with the idea of speaking this software into existence, that I forgot a key principle of building software: It’s for people. To help them solve problems. So you should talk to them about it.
August, 2026 update: I shut down Chore Tracker. I had some nagging concerns about the data being exposed with the free option of the database service I was using, and since it was basically a zombie anyway, it was an easy decision.
