haven't read the whole paper, but the single pendulum, double pendulum analogy caught my attention (took me back to grad school). they use it to talk about complex systems (like a double pendulum) arising rapidly from simple systems (2 single pendulums), which defies basic intuition about systems.
and then (on p.27)...
> "the very dimensionality of ABMs is striking. The largest in physics can now simulate interactions of up to 400 billion distinct particles at any one time."
and...
> "If the largest ABMs have a dimensionality greater than our population, why not model the economy at its highest possible resolution?"
so they use agent-based models (ABMs) (figure 2 is simcity!) to uncover unintuitive emergent behaviors underlying the highly complex system that is economics, things that just can't be understood with the simple macroeconomic models based on a few variables like GDP and inflation.
and then (on p.27)...
> "the very dimensionality of ABMs is striking. The largest in physics can now simulate interactions of up to 400 billion distinct particles at any one time."
and...
> "If the largest ABMs have a dimensionality greater than our population, why not model the economy at its highest possible resolution?"
so they use agent-based models (ABMs) (figure 2 is simcity!) to uncover unintuitive emergent behaviors underlying the highly complex system that is economics, things that just can't be understood with the simple macroeconomic models based on a few variables like GDP and inflation.