My AI lab works while I sleep.
It runs 24/7 on hardware I own. Overnight, it researches, drafts, and checks. It proposes; I approve. Judges never grade their own model, and nothing ships without my final gate. I wake up to decisions, not chores. Every mistake lands on my electricity bill first.
I didn't learn AI from a course. I'm self-taught, and most of what I know came from building my own projects, breaking them, and fixing them myself. That is still how I learn. I want to know what a model is good at, where it fails, and which parts of the work should never depend on one.
I use AI for three jobs. Reading pulls signal out of noise. Writing turns rough material into something useful. Judging scores work against a clear rubric. Anything that must come out exactly right, every time, belongs in a deterministic workflow. A person stays at the final gate. That split shapes both my lab and the systems I help other people build.
The work came before the lab. I've spent 12+ years helping people turn enterprise software into something they use every day. That experience is why I care more about what survives after the demo than what looks good during one.
Most days start or end with a check on what the lab has done. Away from the screen, I strength train because the progress is measurable and compounds over time. Golf keeps me humble for the opposite reason. I live in Atlanta, and I went to Indiana University. Go Hoosiers!