Seems to me number one problem with the Rational Expecations Hypothesis REH promoted most heavily by Freshwater Economists is one person model.
A diversity of internally diverse (with mix changing over time) agents is the obvious reality.
internal operators could include include REH, Herd Following, Random, Pay-Forward, Lazy, Altruistic
Pay-Forward is do what's universally right for it's own sake, but uses conventional definitions of what's right that are not entirely altruistic
Altruistic is do first what's best for others, regardless of social conventions, bounded
Lazy is do as little as possible
for most of those internal operators, there is also an anti-operator, as some 10% or so are simply inclined to be contrarian.
Thinking about Myers-Briggs 4 major types SJ, SP, NT, NF
SJ by far most common type, 50/50 herd following and pay forward
SP random and anti-lazy (enthusiastic)
NF Altruist or Pay-forward at 100%