[i’m guessing entering that space directly would be aided by modeling people’s learning processes separate from other things such as their senses, so maybe modeling what somebody sees, what somebody hears, maybe what knowledge somebody is presently holding, and then drilling down on their choices, maybe modeling good vs bad choices —- and then experientially, detecting good and bad choices is not hard (because of emotional and result-rooted behaviors), so we could skip the mind simulation and model directly when choices are good and bad
here the focus on good and bad choices can build a micromodel of their learning process via simple observation [device power low, sending