Information Brokers are not Filters

Bex over at bexhuff.com and I are still debating the (de)merits of information silos and brokers. The latest entry is an update that Bex makes on his original entry. The crux (as I understand it) is

Rather, I'm stating that the role of the "broker" has changed into being more of a "filter," and is arguably more important than ever.
I appreciate the clarification and I even agree on the importance and role of filters. However a broker is not a filter. A broker feeds/offers information. A filter discriminates between information pieces/sources. CNN.com is a broker. The Drudge Report is a broker. My RSS reader is a broker. Each of these brokers use filters to decide what gets posted. The ways I discriminate between selecting which stories to read and which to believe are the filters. These are in my head, but they are also implemented technologically. My RSS reader queries for certain feeds regardless of the silo they are in. The query is a filter, the sources are the brokers. Yes, they have their own filters because they are not primary data sources (usually), they are aggregators. Aggregators must implement a discrimination practice (aka filter) to ensure that what they're aggregating is basically topical. We should all be generally aware of the filters that our brokers use but that does not make brokers co-equal with filters. That is, unless we're relying exclusively on a single broker.

And this is where Bex and I diverge. If I rely on a single broker for silo'd aggregated data (that is not primary data which, by definition is neither filtered nor aggregated from disparate sources) then I implicitly rely on his/her/its filtering as well. I contend this is bad because, except in trivial, probably tautological cases, there is other nuancing or flavoring data available that ought to be considered (i.e. run through my own filters). This then makes the broker not a bottleneck but a locus of bias (based on the brokers own filters). I would rather use my own bias than adopt (trust) someone else's.

The one exception would be if my filters mesh exactly (or reasonably) with the filters of the broker. In this case the broker's filters become a proxy for my own. However, I never relinquish the ability (technological, moral, ideological) to re instantiate my own filters at my own whim. In either case, the primacy of my filters, my avatar, me as my own broker remains unthreatened.

What more people do need to understand about this way of thinking though, is that if I am my own broker, I alone am responsible for the information I consume and act upon. This is scary to some and it should be. But it is precisely the reason laws like Sarbanes-Oxley were put into place. CEOs go to jail if they sign off on bad numbers even if their CFO said the numbers were good. Relying on the excuse that the CFO as broker happened to be wrong does not matter. The CEO should have had a better (personal) filter and been his or her own broker.

Comments:

In practice, everybody uses both filters and brokers nearly constantly... For two very good reasons:
  1. one lone individual cannot consume all information, and
  2. even if you could, you couldn't verify the truth, relevance, or importance of every piece of information.
Every piece of information was initially brokered by somebody... At some point it was gathered, verified, labeled as important, then shared. Sometimes, people are starving for more information... so they want to broker to go away entirely so they can judge for themselves what data is important. It seems you're arguing from this perspective. Other times, people have way more than they can handle... because the initial demand for data -- ANY data -- has caused a flood of outdated, useless, and just plain false information. Too many possibilities, leading to over-analysis, wasted time, and bad decisions. I say, the solution to both is to free information, then use technology to empower the broker.

Posted by bex on December 09, 2008 at 10:15 AM CST #

Post a Comment:
Comments are closed for this entry.
About

Enterprise 2.0 and Content Management

Search

Archives
« April 2014
SunMonTueWedThuFriSat
  
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
   
       
Today