Comments on: Future microchips based on collective intelligence

By: Paul M. Banas

Paul M. Banas — Tue, 18 Mar 2008 01:06:21 +0000

Thanks for the explanation. I’ve always been entranced by the concept of Condorcet’s Jury Theorem ever since reading Cass Sunstein’s book “Infotopia”. However, my interest has always lagged my detailed understanding of its underpinnings and limitations.

Thanks again for the clarification and added insight.

By: robojiannis

robojiannis — Mon, 17 Mar 2008 12:25:07 +0000

The criticism on the theorem is actually based on basic logic: people usually don’t give yes/no answers, there are more variables involved. So if the theorem’s validity is based on this binary principle, it is probably false.

But Godin’s & List’s research indicates that even when the answers are beyond a yes/no level, the Condercet Theorem is applicable.

As far as I have read - and if I understand your question correctly - the criticism wasn’t on the fact, that binary questions cannot be used to produce collective intelligence, but mostly on the binary restriction itself.

By: Paul M. Banas

Paul M. Banas — Mon, 17 Mar 2008 12:09:53 +0000

Very interesting. Whenever I’ve thought about the wisdom of crowds in the past, the crowds all had human faces. That may have to change.

I was curious about the criticisms you mention concerning the Condorcet’s Jury Theorem. Your previous post mentions binary limitations with people, but even with human decision making, can’t you daisy chain a string of dependent binary questions and still lever the wisdom of crowds?