Was thinking about this today. I had to do a simple wedding planning task - setting up my wedding website with FAQ, cobbling the guest list (together from texts, photos of my father’s address book, and excel spreadsheets), directions and advice for lodging, conjuring up a scheme to get people to use the on-site cabins, and a few other mundane tasks. No phone calls, no “deep research” just wrote browser-jockeying. Not even any code, the off-the-rack system just makes that for you (however I know for a fact an LLM would love to try to code this for me).

I know without a single doubt that I could not simply as an “AI” “agent” to do this today and expect any sort of a functional result, especially when some of these were (very simple) judgement calls or workarounds for absolutely filthy data and a janky wedding planning website UI.