Why Filtering Doesn't Matter Redirected from Why Filtering Doesn't Matter

Why filtering doesn’t matter (2H) 

Convener: Philippe Blanc, Blacktonic 

Notes-taker(s): Dan Chiao, Fliptop 

A. Tags for the session - technology discussed/ideas considered:   

• filtering,  
• market fragmentation, 
• social games,  
• context switching 

B.  Discussion notes, key understandings, outstanding questions, observations, and, if appropriate to this discussion: action items, next steps: 

This discussion centered on a thesis presented by Philippe.  His main points (as I understood them) were: 

A)   New consumer applications for filtering the realtime web are unnecessary because humans already have the right habits and requisite instincts (complemented by existing tools such as traditional search engines) to distill the information we need.  The advent of television is a supporting analogy; we have organically integrated television consumption into our lives even though it was feared to be information overload in the ‘50s and ‘60s. 

B)   These filtering applications are also inaccurate because they generally don’t take into account market fragmentation (i.e. individual preferences) or the characteristic features of each content source.  Feed sources like Twitter are unique “social games” with distinct, complex rules that aren’t easily codified by an “outside observer.” 

C)   Realtime sources are inherently untruthful because users are incented to market themselves rather than dispassionately share information.  As an example, while Twitter may be useful for monitoring an influenza outbreak, it would never be useful for tracking an outbreak of herpes, because no users would self-report it. 

D)   Analytics/data mining and data transport are the only areas where there are real market opportunities for filtering applications.  It is possible to filter realtime streams in order to extrapolate trends and draw big-picture conclusions that are useful to businesses and marketers.  It also makes sense to use realtime filtering technologies to reduce machine-to-machine chatter in B2B applications. 

Philippe’s arguments evoked a mixed reaction from the group.  A few specific rebuttals: 

A)   Filtering tools can classify and cluster, allowing us to monitor more topics in parallel, and to more easily “switch gears” between topics. 

B)   As a counter-argument to the TV analogy, consider that modern cable and satellite TV, with it’s hundreds of channels, would be unwatchable if it wasn’t for “filtering tools” like interactive guides and Video On-Demand. 

We ran out of time before we could develop a consensus or come up with any action items.