I mostly share thoughts here: things I find interesting, what I'm exploring, or whatever feels worth saying.
testing 1233524
AI agents behave differently from human developers, exposing gaps in enterprise APIs around speed, context, reliability, and state.
