Great discussion all.
From reading all of the comments and previous posts from various authors above, as well as my own thoughts, the key is really making the distinction between popularity and authority.
Current influence measurement is primarily numeric based on social activity and little else. If Author Rank is implemented properly (and as per the patent etc.) then social activity will be tempered by/a tempering influence on actual authority. The two will have to co-exist and interact but must be weighted.
I have written before about relative influence and context. Currently we are all treated the same regardless of who we are, how we use the web or what we do. We are just a number!
There must be different calculations of influence based on who we are and what we do: the creator and curator can both be influential but will be so in different ways.
While I don't necessarily agree with TweetLevel from Edelman it does at least look to identify
different types of user based on what they do (it is still flawed, however).
Whatever systems we use must recognise and allow for these differences to work. Google is in a particularly advantaged position in that it is a data host (rather than an external service trying to extrapolate from available data) so has more data points on which to make calculations.
Current systems reward the TweetSheep, as mentioned above, but Google is able to count the number of poeple who actually clicked on the links shared within its ecosystem (as is Twitter, Facebook et al.) - don't just count the number of shares but also the number of people who actually took positive action based on them. Then look at the dwelltime and bounce rates - did people actually find what they clicked on interesting enough to actually
read it? (See +AJ Kohn
's article linked above)
Peer review will be hugely important and there must be weighting based on who
is reading/liking/linking but there must also be sentiment analysis as mentioned above - if the recognised experts in the field all say you're an idiot then perhaps they're right. But this, in itself, can pose a problem with alternative views being dismissed because they do not conform with prevailing opinion. But then that has always been the same in society/science etc. so there are no easy answers.
I have previously written that creators must be recognised regardless of social interaction - take the example of the leading scientist in their field. They may write a research paper which is considered the defining work on that topic but, just because they don't write new papers every week, does it make them and that work any less influential? No.
At least Google is trying
to take a different approach and treat people differently. Author Rank may not be perfect at present but it can iterate, will iterate based on experience and feedback.
As +Mark Traphagen
says we willingly play the game because, at present, there is no other game worth playing. If the game doesn't offer rewards then people will stop and that is their perogative but without enough people playing it can't improve.