Re: Vote on Valuess, Bet on Beliefs

From: mjg223 (mjg223@is7.nyu.edu)
Date: Mon Jul 31 2000 - 15:24:51 MDT


On Mon, 31 Jul 2000, Eliezer S. Yudkowsky wrote:

> It strikes me that the amount of money one has, and the willingness to
> bet, does not necessarily correlate with either competence or
> informational advantage. Unless idea futures are a social fixture, as
> in Stiegler's _Webmind_, one wild-eyed millionaire could blow the whole
> system.

This is one particular example of a general reservation I have about
using markets for optimization.

Take distribution of some finite resource, which each of a collection
of bidders values differently: Since the individual who values the
resource most will be willing to bid most, one hopes that the total
wealth of the system will be maximized by allocating goods to those
who can make best use of them. When all agents are reasonably equal,
such a system works quite well.

The system breaks down, however, when you have highly unequal players.
Rather than a good going to the agent to whom it is of greatest value,
it can end up going to the player to whom the payment means least.

For a practical example, take a market for transplant organs: I'm an
average guy with failing liver - a new one would be of great value to
me - and I can beg, borrow, and, steal perhaps $1,000,000 to
literally save my life. You, on the other hand, are a fantastically
wealthy industrialist, worth more billions than you can conceivably
spend the interest on, and you happen to like eating human liver,
cooked, whipped, and spread on crackers. Since $10,000,000 is worth
nothing to you, you easily out bid me.

-matt



This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 15:30:15 MST