Very interesting. I still have this niggling belief that things like price and utility should be able to be derived from a fundamental matrix of possible unique activities and learnings and there collapse into actions or learnings enacted. I don't know enough mathematics or economics to even start to model it, though I did watch the khan academy on eigenvalues, lol. but building it as a computational model in the way described in the article seems way more accessible to me. I might look for some code.

The most convincing description of utility is as a representation of ranked choices or preferences.

Perhaps it’ll be an interesting exercise to think about two experiments, one asking for valuations of things, and one asking whether one thing is preferred above the other.

Then, consider how these outcomes may be related!

That's interesting. I did read quadratic voting more accurately reflects preference, along those lines. I wonder if there isn't some way of modelling a world with independent actors and a fixed number of unique activities, where availability of the activity changes with respect to preference, maybe conceptually your quadratic vote budget is effectively time units. I suppose you'd have to consider how innovation increases available activities. Is there a real world example of that type of model with pricing data?