People ask me if the AI stuff I build at home changes how I work at the office. Honestly? Yes, but not in the way you'd think.
What Didn't Transfer
I haven't brought a single tool from my home lab into work. No open-source agents running on corporate servers. No experimental workflows in production environments. The IT governance people would have a heart attack, and rightfully so.
What Did Transfer
What I brought was a different way of thinking about problems. When you've watched an AI agent try to solve something and fail, then try again differently, then fail better, you start to see your own processes differently.
You start asking "why do we do this manually?" more often. You start noticing which decisions actually need a human and which ones just need a good set of rules. You get more honest about where people's time is being wasted.
The transfer isn't technology. It's a lens. And that lens changes everything about how you see operational inefficiency.
The Questions That Changed
Before I started building with AI, I'd look at a broken process and think "we need better training" or "we need more headcount." Now I ask different questions. Is this process well-defined enough that a machine could do the repetitive parts? Where is the human judgment actually adding value versus just adding time? What would this look like if we designed it from scratch today?
These aren't technical questions. They't operational questions that happen to be informed by a deeper understanding of what's possible. And they're making me a better leader, not just a more technical one.