Information Cascades and the Wisdom of Crowds
In an article on the wisdom of crowds, Larry Hardesty uses the idea of information cascades to discuss ongoing research at MIT on “herding” in social networks. He reports that economists, electrical engineers, and computer scientists have shown that “as networks of people grow larger, they’ll usually tend to converge on an accurate understanding of information distributed among them, even if individual members of the network can observe only their nearby neighbors”. These decisions may be as simple as which phone to buy or as complex as which candidate in this upcoming presidential election will better serve our country’s needs.
This is interesting considering recent lecture topics because the article discusses how decisions made by large groups of people can stem from a very small portion of the crowd. As explained in class, through the use of the simple herding experiment of colored marbles in an urn and Bayes’ Rule, the decisions made by a select few can quickly propagate and assimilate in crowds of arbitrarily large size, given that people only know the decisions that others make and not the private information they used to make that decision. For instance, in the simple herding experiment, if just the first two people guess that the urn is a majority red, then everyone after them will also guess red regardless of what color they see.
http://web.mit.edu/newsoffice/2010/crowd-wisdom-1115.html