He gives the example that if Google delivers search results that arent blocked by robots.

txt, people find out about it and draw a lot of attention to it, whether that be by blogging or forums, or whatever. We try to be relatively careful, says Cutts. With search results, Google tries to make sure theyre properties arent too spammable, and if they are worried that something might be potentially abused, they will often use a nofollow attribute. Cutts says with Google Knol, most of the external links started out as nofollow, but then over time as they learned which Knol authors were more trustworthy, they lifted the nofollow on those Knol authors. Sometimes we do flow pagerank, but we try to be relatively careful because were mindful of the people who do try to abuse it, he says. Google has a lot of properties, and this might not answer everybodys questions on the subject, but I would bet that the trust factor mentioned with regards to Google Knol, could probably be applied to other products.