The Devastating Effect of Information Brokers in Multi-Agent Systems

12^th November 2013, 16:00 Ashton Lecture Theater
Dr. David Sarne
Department of Computer Science
Bar-Ilan University
Israel

Abstract

In many multi-agent systems we find information brokers - agents that can
disambiguate noisy signals upon payment of a fee (e.g., Carfax that provides
reports on car histories, credit report companies, experts that provide
accurate estimates on auctioned items). In this talk I will show that better
information can hurt: the presence of the expert, even if the use of her
services is optional, can degrade both individual agents' utilities and
overall social welfare.
The talk will focus on information brokers in two specific domains: auctions
(where the provided information relates to the common value of the auctioned
item) and distributed agent matching. For the first, I'll show that with the
existence of the information provider, in conflict with classic auction
theory, the auctioneer will not necessarily find it beneficial to have more
bidders participate in the auction and similarly bidders will not
necessarily prefer less competition. Also, bidders' unawareness of the
auctioneer's option to purchase the information does not necessarily play
into the hands of the auctioneer and, similarly, bidders may sometimes
benefit from not knowing that the auctioneer has the option to purchase such
information. For the distributed matching application I'll show that the
externality imposed by the fact that others are consulting the expert can
lead to a situation where the equilibrium outcome is that everyone consults
the expert, even though all agents would be better off if the expert were
not present. In both cases the agents may find it beneficial to pay the
information-provider in order to leave the market (or alternatively publicly
increase the price of the information she offers).