Re: [decentralization] decentralization and Mobile phones
- Christensen's second book, "The Innovator's Solution", presents, in
passing, an interesting framework to apply to questions like this one.
He talks about "modular" and "interdependent" (I think he applies it to
the wrong thing, or I don't understand him, my own interpretation is: )
Companies are incentivized to build interdependent markets, i.e.
markets where individual components are tightly glued to each other.
Where markets become modular (typically a later stage, once technology
has developed sufficiently and competition shows up from the bottom of
the market), innovation can explode, at the cost of losing market
dominance by the vendors, which is of course something they'll fight.
Good examples: Apple offers interdependent computers (OS and HW etc.
are tightly integrated, and not products on their own), while PCs are
modular (although PC subsystems, notably the CPU and the Windows OS,
are not -- Linux is)
If one applies this model to the mobile phone industry, it appears to
me that in the US, there is interdependence between the mobile phone
network of a certain operator, and the handsets that they offering for
that network. (I think in Europe things are more modular). I'd argue
the difficulties Nokia has been running into lately under the label of
"Nokia has not done private-label phones" has been that the operators
control the interdependent components while Nokia does not, and the
profits split up accordingly -- but I digress.
The mobile operators are incentivized to create as many
interdependencies of new technology (Julian's example of location-based
services that must go through a server) into the existing, defensible
product (the mobile phone network) as possible, in order to avoid
modularization which would likely create an explosion in innovation
(e.g. what Julian wants to do) but kill the operator's control over the
market, and thus their control on where to put the toll booth and thus
To follow Christensen, the dynamics to get out of this are to find a
crappy modular technology competing in the very low and of the market,
which finds volume appeal, which will unroll the market from the bottom
and modularize it. If you have any ideas on how this vector is going to
happen, I'd be very interested.
- Interesting tangent:
"The Internet Corporation for Assigned Names and Numbers (ICANN)
announced Tuesday that it had begun adding IPv6 addresses to the DNS
root server system, which maintains the master list of domain names."