Nick Szabo  🔑

Which leads to oversimplification. Will need to shift to complex adaptive mechanisms. Cybernetics more useful than game theory.

CAS especially important when entities themselves have models of the systems they are in.

All Economics can do is Rational Expectations theory. This absurdly assumes each entity has same model of system, the one the economist has.

GT & Economics assume so-called rational behaviour. This assumption inappropriate for computational systems because code can be buggy.

...preference is an adequate basis for reproducibility.

Also remember that the axioms of Game Theory can derive probability theory and Bayes Theorem. Use prob. to express knowledge of outcomes

Prob theory is just one among many formalisms for representing uncertainty. Not necessarily the best for any particular application domain.

The maths of game theory also does not scale well. Many players, action-options, payoffs, leads to intractable math.