I have been thinking about the role of filters in signal processing and time series analysis, but the more I think about it, the more I see filters in many other places, principally as a metaphor for perception, thought, memory and communication. To filter is to remove some aspect of reality, and this happens all the time.
If we were to absorb all the information from the world around us: sights, sounds, voices, faces, facts – we would probably be incapable of functioning and go insane. This reminds me of the schizophrenic woman described in the The Man Who Mistook His Wife for a Hat who imitated all the people she met on the street: she absorbed it all, and lost her own identity, just assuming the quirks and mannerisms of everyone around her. So there must be some kind of mental filtering mechanism going on: but how does this filtering take place? (Unfortunately I do not know of any work in this area. I also wonder if any compressive sensing ideas been applied to this. It seems like they should be).
While people are able to filter out a lot of the information around them, I think they often filter out the wrong information. There is often a confusion between the essential and the unimportant (the noise and signal), and this problem will only get worse as information becomes more easily accessible. More information is not better information when we don’t know how to use it. It can be easy to drown in a sea of irrelevant details.
The presence of noise is a huge problem in the news. Unfortunately most news tries to entertain, shock or surprise, but not to inform, and a result journalists do not report on things that truly matter. As a result, people watching or reading the news often focus on details which the media make seem important or which might seem to matter today, but which are of no real consequence.
With the Internet we are now able to distribute as much information as we’d like, to as many people as we’d like, for a very low cost. This is a really amazing development that people often take for granted. But the sudden growth in the ability to communicate means that people are not as selective about what they put online, which means that there is more noise to filter out. It seems that there is a law of diminishing returns here: the more information there is, the less is the real information content and value of each additional bit added. (Incidentally, Clay Shirky gives an interesting talk about filters in technology and society here.)