Ideally, the web should emulate the best practices people have used for many years. Authority should be no different. For example, you probably define authority the same way I do – by watching who experts consider to be experts.

For example, before I became a member of Seth’s Triiibes.com – I had never heard of you – nor had I visited your blog. However, Seth has a few recommended blogs listed – one of which happens to be yours. Placing tremendous value on Seth’s recommendations – I automatically consider you to have tremendous authority.
Now I visit your blog and pay attention to what you write.

Unfortunately, the web has yet to develop a good system to emulate this. Google’s PageRank is probably the closest – but it’s unreliable and suffers from several flaws. The biggest flaw – IMHO – happens when only a few top blogs in a certain niche consistently regurgitate links to one another. We all do it – it’s much less time consuming to have a pool of X number of blogs we follow and 95% of our links go to those blogs. This creates a type of “good ole boy system” where authority is skewed tremendously.

The solution I propose would be based on the way we determine authority naturally (as described above).

However, I’m not sure it could be completely automated.