There is a principle which is one of the the basic in statistics and opinion polls. If you take a large number of people and ask them their opinion on something, most of them will probably give you an answer close to reality. It is also called common believe or wisdom of the crowds. Modern societies (mainly democratic ones) are based on this principle. Their goverments are elected by the majority of voters. However, the question of the source article is how unbiased is the crowds opinion on something when they know what the majority of the society is thinking on it.
Take for example a pop star that you don't really like. You can't say you enjoy his/her songs. You have a neutral and maybe slightly negative opinion on him/her. What if you knew that 70% of your friends think s/he is talented and like him/her. Whould you give him/her a second thought? Whould you alter your opinion to his/her favour? Well... this is another phenomenon called mass psychology. We often tend to follow large masses' opinions even if we are not 100% sure about them.
Now combine the latter with social media like facebook and twitter where you can easily see what large groups of people follow. Even better you can identify frineds of yours that you might respect their opinion in those groups. There you have it. A much easier and more effective way to bias the wisdom of crowds. So the question is... how wise are the crowds today with all this social media information sharing? The source article discusses the question and presents an experiment which shows that maybe crowds are not as wise as they were without all this social information sharing.