Robert Zaremba Blog

Word of the day: Praxeology

Why do we do the things we do? Why do some people like tea while others prefer coffee? Why do act as we act (communicating, learning, driving, working...)? This and many more questions are important part of any business and artificial intelligence.

Praxeology is the science of human action, which is purposeful behavior.

It’s one of the greatest AI challenges to learn the human behavior. It’s necessary for all kind of artificial assistants which we will see more often in our households and offices in the next decade.

Mises Wiki explains what the purposeful human behavior is:

A human acts whenever he or she uses means to achieve an end that he or she subjectively values. Human action is thus teleological or intentional; a person acts for a reason. Not all human behavior is action in the praxeological sense: purely reflexive or unconscious bodily movements (such as coughing when exposed to tear gas) are not examples of action. Praxeology starts from the undeniable axiom that human beings exist and act, and then logically deduces implications of this fact. These deduced propositions are true a priori; there is no need to test them in the way that a physicist might test a proposed “law” of Nature. So long as a praxeological statement has been derived correctly, it must necessarily contain as much truth as the original axioms.

For example, when we throw a ball, we do not reason that it is guided in a teleological way by some mystical spirit or “prime mover.” Instead we use the laws of mechanics and causality to examine the position, velocity, and forces acting on the ball, in order to predict the future position and velocity of the ball.

Technology side

I would like to link a great Mark T Littlewood talk, who is explaining technology without psychology can actually be a source of dangerously misapplied effort. We need to apply praxeology when designing all part of technology.