Michael Battalio

Friday, February 12, 2016

Discussions on Wealth (part 12): Chapter 6: summary part 1

This discussion on wealth is an offshoot of  Serious Conversations parts 53 and 54.  We are considering the book  The Origin of Wealth by Eric D. Beinhocker.  (I do not profit from clicks).  (Ed.:  we will be taking the general format of outlining the major points of the chapter and then discussing what we believe to be important or intriguing points.)

This Chapter is about how economics views humans.  We are finally addressing the ridiculous proposition of the completely rational human.  We begin with the absurdity that Traditional economics thinks each human goes through when it considers a purchase.  We should, according to Traditional theory, consider the purchase of any item in the context of all possible future items we might need with a logical plan about how purchases effect our budget and with logical consistent preferences.  Also considered how that money spent on the purchase will effect all future decisions.  Complexity economics recognizes that humans are impulsive and might buy something on a whim.  We are introduced to a new theory, bounded rationality, which says while humans are smart, we aren’t that smart.  We have imperfect computing capability and imperfect information.  We “take the information we have, and we do the best we can.”  Finally, behavioral economics is noted that shows people do not follow the Traditional economics paradigm of human behavior, but there is no way to model it.

As an example of irrationality we have been given a version of the ultimatum game.  Traditional economics says that any offer should be accepted by the second participant because whatever the amount they are still richer for it.  This does not happen because humans are programed with a degree of fairness as a result of our evolution.  We each feel it our duty to hold other accountable.  

Another flaw of Traditional economics is that humans never make mistakes.  A series of common biases which lead to mistakes is described:  framing bias, representativeness, availability biases, difficulties judging risk, superstitious reasoning, mental accounting.  A perfectly rational human could not take advantage of other’s biases because it is impossible to build a perfectly rational machine.

Inductive (reasoning by pattern recognition) and deductive (reasoning using logic) reasoning are contrasted.  Traditional theory considers humans perfectly deductive when we are certainly not.  We reasoning by induction, using past experience to influence decision making.  An inductive model is described as an agent that has a set of goals and a way to determine if its decisions bring it closer or further from those goals.  A list of condition-action rules (if-then statements) dictate the course of action.  The agent can add to the list of rules based on experience.  If an action has deleterious consequences it is applied less frequently.  This model is explained by way of a frog.  It has simple actions including [flee], [pursue], [extend tongue], etc.  It has various inputs [moving], [striped], [near], [blue], etc.  And various rules IF [small], [flying] THEN [extend tongue].  These rules can be changed; however, if the rules are ever in conflict the frog has credit assignment where rules are scored based on past helpfulness.  The higher scored rule is more frequently used.  This simple model can also be used to plan out long term goals using a market system.  Rules compete with each other by bidding using their credit.  Finally we assume that the rules self-organize into hierarchies.  Rules become associated with each other (e.g. [large], [flying], [flapping] are associated with [bird]).  This allows the system to respond to new situations and allows for reasoning by analogy.  

No comments:

2003-2016 Michael Battalio (michael[at]battalio.com)