Why does Pariser suggest that experts are most vulnerable to confirmation bias?

Eli Pariser’s book, The Filter Bubble discusses the growing trend of personalization online. Google knows enough about you to personalize your search results so that those links you are most likely to be interested in show up at the top. Facebook only shows updates from those friends you have interacted with most recently. Advertisers show you ads they think will be most attractive to you. You are increasingly living in a “filter bubble,” only exposed to those things you like the most — and agree with. Two different people with identical search terms will get completely different results, just because their invisible filter criteria are different. The problem is most people are unaware of these filters — they did not deliberately set them up themselves.
There are some real attractions to this model, but dangers as well. Read Chapter 3 of his book (long read) and answer the following questions.
answer these five question

Why does Pariser suggest that experts are most vulnerable to confirmation bias?
In what way does he suggest that reducing exposure to news and information (filtering) that doesn’t really interest you will limit your creativity?
Can this filtering be useful? When might it be useful and when might it be harmful?
Humans are social creatures. We are most comfortable when our ideas and beliefs are in consensus with those around us, even if that doesn’t always happen. [Consensus is a broad societal agreement on any given topic.] How would this filter bubble affect that consensus?
If the filter bubble has blocked some information from your view, how will most people find it again? We’re not only asking what people should do, but also what they will do.