On a more serious note, I’m now in the process of reading David Weinberger’s Everything Is Miscellaneous. I brought this book to read on vacation, not knowing that it would be directly related to this particular line of thinking. It is and I will post my thoughts. Stay tuned…

Your post along with some of the other information-centric posts you have done got me thinking about visualization models and tools to find, discover, and build other ideas like the FM station finder that are out there. This might be old hat, but thought it was worth the time to send over anyway. It is like the chaining in a rule engine in some respects… So what if there were ways to easily get a complexity/difficulty score of automatically relating any two bodies of analog information–with the longer term goal being the ability to support perpetual hands-off automatic update (within the bounds of depending on others continuing to provide their part of the information trail)? This is the creation of real value: automatic transformation and assimilation from analog (i.e. creating the ability for users to teach their software how to unambiguously relate things based on ambiguous source materials). So it starts by assessing the transactional cost of relating that type of information in analog (i.e. searching for it by hand and interpreting text or images). The cost or distance of creating each relationship required could be automatically generated by aggregating interference patterns for combinatoric queries branching out from each end of the “information trail” using search engines–similar model to how directions between two street addresses are generated under the hood. So the user is left with a choice to get their answer, they could deal with each analog pivot on their own (assuming a way to store the analog pivot can be remember and re-applied). From the analog pivot, perhaps there is a computer-assisted road to automation through automatic transformation, depending on how digital the “information signal” is.

Rough example of what should exist:

• Start with list of all FM radio stations in US
• Lookup reference information for towers, frequencies, and power levels from various sources
• Assimilate into table of tower location, frequency, hours of operation, and station power
• Assist a user pivot into 2D map of radial signal coverage by frequency based on the lat/lon of the tower locations and power level (This would look like that scene in war games where the simulated bombs make discs on the map as they explode.)
• Assist a user pivot into spatial least-likely-to-be-interfered-with signal (e.g. similar to a topographical map, but with frequency voids instead of elevation). (Aside: There is probably a very cool 3D “biggest abscess hunter” visual version of this using lat/lon/freq as the 3-axis for the rendering. This would be like virtually swimming through seltzer water looking for biggest area without any bubbles.)

Next step: create “knowledge service” that accepts lat/lon location and returns the biggest abscess of frequency interference based on geometry of signal patterns and previously generated matrix. The service execution should be automatable against the generated information in prior steps.

Next step: have an automated way to “outsource” required analog conversions going forward to lower cost resources (i.e. overseas) if you choose to keep the information alive. Perhaps the whole thing could be “watched” so that a pay per refresh model could also be supported.

So the tool would visually let people build structured information as they would like it to exist, and the system would figure out how ways to achieve it, even if human intervention and resolution is required along the way. The experience would be no different than getting directions on Google Maps, but instead of getting directions from city A to city B, it would be for getting to information X–and the underlying service figures all the pivots required.

These “information trails” double as “knowledge services” for others ex-post. Organically this could be a model for masses to digitize their knowledge in the process of achieving their own goals! Oh wait… I think they already call that the semantic web? Either way, at some point my computer will finally start asking questions instead of always giving answers.

Oh yeah, last and most important step: mandatory merging of all Taco Bell Locations on the planet at all times (wink)

Any thoughts you have or resources you can share on the topic would be greatly appreciated

-Loren