Last updated June 15, 2026
Making AI Useful
Aaron Makelky tests AI tools on actual work and writes down what saves time, breaks, or holds up.
What this site is for
I try AI tools on actual jobs. If they save time, I keep notes. If they break, I write that down too.
Right now I am testing what I would use again next week: a prompt from a workshop, a tool that saves a few minutes, or a workflow that makes one annoying step disappear.
My filter comes from Wyoming classrooms, coaching, and the move into tech. If a workflow only works in a demo, I am not interested for long. I want the version that saves time, helps a student get unstuck, or keeps a team from checking the same thing all day.
The useful path usually starts small. Read one note, try one tool, or bring me into a workshop where people leave with something they can repeat on Monday.
Workshop proof
The latest workshop feedback was plain: people wanted practical examples, repeatable prompts, and time to check the work.
In the LUM Studio feedback, 10 of 10 respondents said they left more confident using AI at work. Eight marked the session extremely useful and said they had clear next steps.
The most useful critique was also practical: people wanted more time for examples, Q&A, and discussion.
Start here
The blog is where I keep the longer thinking. The tools page is where I keep small utilities, Codex skills, and prompts I made after the normal way got annoying. The shop is for OpenClaw kits and agent reliability resources. The about page explains the teacher-to-tech route behind the work.
Keep going
Sources and references
Short answers
Who is Aaron Makelky?
Aaron Makelky is a former Wyoming teacher and football coach who now works in tech at Descript and writes about practical AI use.
What does Aaron help people do with AI?
He tests AI tools on real jobs and turns the useful parts into notes, workshops, and small tools.