Mises Wire

A
A
Home | Blog | Comments on MSNBC "Mind Reading" Article

Comments on MSNBC "Mind Reading" Article

July 13, 2004

MSNBC has an interesting article on Mind Reading. Usng a functional MRI (fMRI) machine, which is able to show brain-activity in real-time, Prof. Quartz (neuroscientist) and Prof. Camerer (economist) have been using the machine to understand problems of decision-making.

Prof. Glimcher, a a neuroscientist at NYU, has shown that:

the rate at which these neurons fire is proportionate to the expected utility of the juice payoff. The implication is electrifying, especially to economists: an abstract, mathematically derived formula appears to be literally hard-wired into the primate brain.

We can be thankful that most mainstream economists seem to have very little understanding of the scientific method, and couldn't design a real (as in, acceptable by Nature/Science) experiment if their lives depended on it. Otherwise, we'd have every mainstreamer trying to "quantify" in biological terms "utility functions" for people. (actually, this is moot for economics, since even if "utility functions" can be biologically characterized and quantified, they are still constantly changing...you can't take out the element of personal freedom of choice). Unfortunately, mainstream economists can collaborate with neuroscientists, who do know how to design experiments. The problem is that mainstream economists aren't going to provide very useful input for designing neurological experiments relevant to economics, as can be seen from a few considerations.

Regarding the experimental games of investing/etc, they are interesting, but fundamentally flawed. There is a fundamental error in these experiments. In these experiments, two partner's play a game with 10 rounds, and they know that there are 10 rounds. In real life, we don't know when the "game" is going to end. So, the idea of trying to cooperate until the last, and then screw your opponent over on the last round, is flawed, since we don't know when the last round is. Furthermore, real life is not a game. Even if we manage to pull off a real "win" and screw over a creditor, borrowing money, and then dying, there are still consequences: for our heirs, who will inherit less or nothing.

A way to do these experiments which would be more true to real life would be to randomly determine how long each "set" is going to be, and not telling the players this, nor even telling them what the probability is that the set is going to be certain lengths (that is, you would randomly select from different randomized selection-processes which are differently distorted). Also, we have players who are playing thousands of miles apart. How about we put these players in the same room and introduce them to eachother, to introduce a tiny element of consequence for one's actions. When you know your opponent, are you just as willing to be a Scrooge on the last round?

These are still fascinating ground-breaking experiment, however. They could be improved by Austrian insights.

Some relevant links:

Follow Mises Institute

Add Comment