|
Friday, February 21, 2003
|
|
|
The wireless commodity economy.
Forgive me if this is obvious, but some intuitions about the future of the wireless and internet economies seem to have crystallized, for the moment at least. I come to these thoughts while considering a number of business ventures related to wireless networking and reflecting on recent history. Comments?
I think the wireless digital commons will (and should) become a globally ubiquitous amenity like drinking water, air conditioning, public libraries, dial tone, etc. These are world-changing commodities that we take for granted (others, less fortunate do not). As important as they are (we die without water) they are cheap. Not a great for-profit business to be in, unless perhaps youíre a utility or a behemoth.
It is, rather, during periods of expansion, when one of these soon-to-be-ubiquitous amenities is taking over the world that there is significant money to be madeÖtypically by supplying picks and shovels, know how, etc. These are good businesses to be in, if you can ride the wave, but remember: every good wave eventually crashes on the beach. When the wave is played out, you either need (a) to take your toys and go home, or (b) find a new wave. Or (c) perhaps your business was not about riding a particular wave, but about riding waves in general. That is a long-term business model for the new information economy, because in this domain, ìlong termî may be a matter of months. But thereís nothing wrong with (a) or (b) as long as you can avoid being pummeled at the end of the ride, and or disappointed.
And thereís nothing wrong with (d) either (au contraire): help make it happen because itís a good thing, not because it may be a profitable thing.
11:07:46 AM
|
|
Powerlaws are a means to, not an end of, knowledge.
In broad strokes I'd suggest that
Power Law data are suggestive of interesting underlying processes and phenomena, they are not hugely interesting in and of themselves.
Mathematical analysis is valuable and "cool" but it should not preempt or trump more "direct" forms of perception.and analysis.
In particular, I believe that our visual systems are the best pattern recognition and pattern classification systems around. We should apply them here.
So I suggest that Power Law data are best used to inspire "organic information visualizations" of the underlying phenomena.
Granting that linking structures are usually not perfect trees (they are typically webs: branches merge as well as bifurcate) a tree shapes and plants shapes are useful and suggestive approximations of what "fan out" looks like.
So here's my conjecture: just as plant shapes often fit Power Law functions while covering fascinating diversity of very specific patterns, so must Power Law-consistent linking (and other) data from span a huge and fascinating diversity of patterns...interesting in their own right, and interesting as clues to the dynamics that form and constrain them. Having identified Power Law phenomena in cyberspatial phenomena, it would be a huge error to dwell overmuch on the mathematical phenomena while failing to characterize and visualize the underlying phenomena themselves (in all their idiosyncratic vividness and specificity). A huge an error, perhaps as detecting signals of intelligent life under on another planet and then focussing attention on the structure of the signals without attempting to determine the structure of the aliens of themselves.
So let's get on with it!
Some relevant references on this topic (which I found in a very cursory google search and have not read) include:
- Niklas, Karl J. Plant Allometry: The Scaling of Form and Process. xvi, 396 p., 3 halftones, 130 line drawings, 18 tables. 1994 http://www.press.uchicago.edu/cgi-bin/hfs.cgi/00/12648.ctl
- T akashi Matsuo, Masato Nakakubo and Keijiro Yamamoto. Scale Invariance of Spatial Distributions of Tree Branches, Leaves, and Petals Forma, Vol. 12 (No. 1), pp. 91-98, 199.
Abstract. The box-counting method was applied to measure the fractal dimensions for ramification patterns of trees and the spatial distributions of leaves and blossoms. The relationships between box size and the number of boxes necessary to cover the pattern for tree branches were found to obey the power law in a scale length of one decade. On the other hand, the distribution of leaves and petals showed scale invariance over a wider scale length of 2 orders when observation conditions were properly selected.
-
- and this, which may illustrate my point better than my words.
10:46:22 AM
|
|
|
|
© Copyright
2003
Jon Schull.
Last update:
10/23/03; 7:41:09 PM.
|
|
February 2003 |
Sun |
Mon |
Tue |
Wed |
Thu |
Fri |
Sat |
|
|
|
|
|
|
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 |
|
Jan Mar |
|