roundtable: $witched digital?

roundtable: $witched digital?

$witched digital?

natlwtrgrp@aol.com
Mon, 28 Feb 94 17:29:15 EST 

From: natlwtrgrp@aol.com
Message-Id: <9402281729.tn129316@aol.com>
To: roundtable@cni.org
Date: Mon, 28 Feb 94 17:29:15 EST
Subject: $witched digital?

For all their differences, many Roundtable participants seemed to agree 
that "switched digital" is an idea whose time has come.  It's not hard 
to understand why. Committed to providing a wide variety of services to 
virtually everyone, switching seems to make sense. It is, after all, the 
way our telephones work and they provide service that is interactive, 
versatile, usually affordable and almost universal. The phone's only 
problem seems to be that it's narrow band and analog. Shouldn't we stick 
with a winner when we go broad band and digital?

Not if switched digital provides service that is more expensive and less 
versatile than some alternative. Not if it creates a war between 
technological haves and have nots or creates legally mandated obligations 
and "in-kind services" that shift money between groups depending on the 
latest whims of politicians. There are other ways you, I and anyone who 
wants could establish a wide band connection. We need to concentrate on 
giving birth to a technology that will make our dream inevitable, not in 
trying to mandate bits and pieces of it by legislative fiat.

In the days when telephone wires really were wires, switches made a lot 
of sense. If I called someone at the other end of the country, literal 
stepping switches created a continuous copper connection between the two 
of us. When long-distance service began to be multiplexed on microwave 
signals, a less literal form of switching still made sense because 
bandwidth remained limited. The phone companies couldn't afford to waste 
a scarce resource sending information anywhere it wasn't needed. The same 
is true in packet switching networks like the Internet. Expensive as they 
are, switches prevent wasting even more valuable bandwidth.

But imagine a world where bandwidth is virtually unlimited, a world 
where we can waste gigahertz of the spectrum sending data to millions 
of unnecessary destinations simply to avoid the bother and expense of 
processing it through expensive, time-delaying and bandwidth-limiting 
switches. That is the potential that fiber-optics offers. Fiber optics 
isn't just another phone line or microwave link. One thin fiber has a 
potential bandwidth about 1,000 times greater than the entire usable 
portion of the radio spectrum from VLF to upper microwave. (And unlike 
radio, it offers virtually error-free transmission over any distance.) 
When that one fiber fills up, simply add another. Like computer chips, 
they're made of silicon, one of the most abundant substances on earth.

Control depends on scarcity. When bandwidth is like oil, others control 
the price and availability. We pay their price and live with their 
conditions.  Bringing the government into the situation, as many Roundtable 
participants seem to want to do, merely changes the master; we are now 
dependent on a government policy which may or may not favor us. What the 
government can give, it can take away. Not very bright. (In the usual 
economic terms, monopoly is artifical scarcity, competition is limited 
abundance, and government regulation an attempt to declare by fiat that 
scarcity will not exist.)

But eliminate scarcity, making bandwidth like sea water in the ocean 
or sunlight in the desert, and the dynamics of control change radically. 
Technically, some may still "own" the resource--telecommunications 
companies--but when what they own is easily available in enormous 
quantities, it creates a virtually irresistible temptation to sell at 
any price. That is what you and I ought to working toward--making 
bandwidth cheap and abundant--not arguing about how to dole it out like 
beggars around a dumpster.

Take television. Creative people with good ideas have every right to 
be upset by a system dominated by "gatekeepers" who can deny access. 
Mandating certain percentages for outside productions may limit the 
power of some gatekeepers, but it doesn't fundamentally alter the system. 
It's still a scarce resource that some control while the rest go begging.

But imagine that a mere one percent of a single, basic, nation-wide 
fiber-optic cable is dedicated exclusively to television programming. 
Give it extra bandwidth to handle high definition TV, multiple language 
dubbing, closed captions and so forth. How many channels would that open 
up? Roughly 25,000. Will we ever fill that up? Probably not, but if it 
does we simply increase to 10 per cent and 250,000 channels.

How would people find anything among so many channels and how would it 
all be financed? In short, how will we cope if the gatekeepers no longer 
limit our choices and mail us monthly bills? Enter the teleputer, an 
easy-to-use computer and television hybrid and the solution to the 
"can't program their VCR" syndrome. Another portion of the cable would 
be dedicated to interactive teleputer to computer communication. Anyone 
who wanted could put a computer on line and begin offering a package of 
services, some free, some based on a monthly fee, some pay-per-view. 
(Everything being digital, encoding and decoding would be relatively 
easy and require no proprietary set-top box.) One service might offer 
old French movies, another might be run by a fishing club. The result 
would be thousands of competing "virtual" cable companies, companies 
that anyone could start on a shoestring. (Technically, most would be 
gatekeepers, but we'd pick our own gatekeeper, not the FCC or the local 
city council.) Seventy-five per cent set asides would be pointless in 
the midst of such abundance and there is no danger that "sunset laws" 
would take away hard-won access.

Other features that Roundtable members want (computer data, person-to-
person video transfer etc.) would simply get their own band of 
frequencies on the basic fiber. Your computer or teleputer would simply 
request a block of spectrum and time from any of dozens or hundreds of 
competing suppliers and spectrum resellers (or use public domain regions 
roughly equivalent to the amateur radio bands). You wouldn't be stuck 
with a handful of neighborhood providers. Any of a vast array of 
services could reside on the basic fiber much as they do on today's 
Internet.

Some bands would be local or regional, reused over and over again. 
Others would be national, going everywhere the fiber goes. Frequency 
rather than switching would decide where data goes. Divided into 
frequency bands (like radio) rather than broken into tiny packets, each 
kind of data could be sent in its most efficient form, not mutilated to 
fit a one-size-fits-all switching scheme. Broadcasting could be handled 
differently from person-to-person, interactive could be different from 
one-way. Encryption would be up to the sender and receiver not the 
service provider's Clipper chip. Standards could change without requiring 
the network to be rebuilt. Everything could be done in the cheapest and 
most efficient way. (Switching could even be used where it makes sense.)

In short, the network would be dumb, reaching virtually everyone but 
offering little more than raw bandwidth to all comers. The smarts would 
lie at the end of the cable with the users. And while someone would own 
the actual fiber in each community, it would be licensed as a common 
carrier, providing service to all (including cut-rate resellers). Anyone 
who wanted to lease or sublease a bit of that bandwidth could do so 
either for a few minutes or full-time.

You get the point. There's room for all. Virtually every imaginable 
use could get its own space.  Because spectrum is so abundant and the 
technology so flexible, fussing over what services the system will or 
will not provide and at what cost becomes unnecessary. Computer-hating 
couch potatoes wouldn't be forced to subsidize Internet wall-plugs for 
nerds. TV-hating nerds wouldn't have to subsidize 500 channel TV for 
idiots. If you want a feature, buy or rent the black box that lets you 
use that service. If you don't, you don't pay a penny. And doing things 
the cheapest, most efficient way would make it much easier to provide 
low-cost access for schools, libraries and public forums. (The more 
abundant a resource, the lower the "incremental cost" of putting them 
online.)

The most obvious way to do this would be to provide most people in the 
country with, not just fiber-to-the-home, but the same fiber carrying 
the same data. That's the dream, but for the moment economics suggest 
a different approach. Except for the more isolated areas, fiber itself 
isn't the financial bottleneck, the gadgets we put on the end of it are. 
Many of the gadgets to transmit and receive over fiber-optic lines are 
still being developed in laboratories and devices to take extract a 
single signal out of hundreds are still expensive, com parable to the 
minicomputers of the early 70s. They're not yet commodity items. But 
they will be.

For a time, interim solutions may be necessary. Until prices come down, 
the expensive gadgets will need to installed and owned by service 
providers with the rest of us sharing them by making requests from their 
black boxes and getting the data or video via wireless, coaxial cable, 
compressed data phone lines or even not-used-to-their-full-potential 
fiber-optic cable. But the costs will come down and when they do, the 
consumer will acquire enormous power, the kind of power only large 
corporations and governments command today. The only danger is that we 
may lock ourselves into an infrastructure dominated by an unwieldy 
switched digital system that we and our children will live to regret.

In the language we use, we've already recognized the wisdom of this 
approach. Switched digital is like the old system of stop signs, red 
lights, and low speed limits that were necessary for a world of two-lane 
highways. The information superhighway needs to be built like a real 
superhighway (or better yet, like the ocean) where data flows from its 
source to its destination with minimum outside intervention. That means 
a pervasive but dumb fiber not switched digital.

The Usual Disclaimer: I work on the high-technology project of Discovery 
Institute, a Seattle think tank.  I also work with the technological 
futurist George Gilder, one of our senior fellows. Many of the ideas in 
this statement come from the latter and I recommend reading what he 
writes, particularly in his series of Forbes ASAP articles and the book 
Life After Television (if you can wait, a revised and enlarged paperback 
of the latter will be out in June). But the statement itself and the 
responsibility for all this says are my own.

-Mike Perry, Discovery Institute, Seattle, WA   Internet: natlwtrgrp@aol.com

[CNI Home Page]