My day job is mapping actors in a system, mostly for development aid programming. This means looking at things like who has to act, who can, and who wants to. Listening to a lot of podcasts on AI (including recent interviews with Kevin Roose on his book 'The AGI chronicles'), I wanted to see if we could map the human system around AI in that way too.