Stopping the Information Railroad

Keynote Address
Winter 1994 USENIX Conference
San Francisco, California
January 17, 1994

by John Perry Barlow

Over the last four or five years since I left the cattle business I've
started to feel like my life has turned into a Thomas Pynchon novel. Weird
experiences are so numerous that I don't even keep track of them anymore.
But I've got to say, for somebody who's spent most of his working career
pushing cows around, this is a very weird audience for me today.

You people are great though. I honestly believe, and I say this without
hyperbole, that the people in this room are doing things which will change
the world more, in terms of what it is to be a human being, than anything
since the capture of fire. I'll try to justify that very broad statement as
I go along here.

I'm not entirely unqualified to think and talk about wild places. I'm from
a part of Wyoming. The county I live in is larger than the Netherlands and
has a population of 3500 at the moment. It is the focal point of the
history of the fur trade. The fur trade was an economic manifestation that
came into the west in the 1820's and 30's. Many of its constituents would
be familiar to you.

They were kind of a fractious lot of misfits and opinionated loners. They
were somewhat irregular in both their eating habits and their personal
hygiene. They were hairy and anarchistic. They were smart. They created a
society which was largely self organizing. And they were exploring unmapped
territory using tools they developed themselves for getting around. I would
try to draw a close parallel between them and the people in this room but I
think that will be unnecessary.

When I first started to put my head into Cyberspace, it was not as
unfamiliar to me as it is to a lot of folks who are now getting into that
area, because it had a lot of the characteristics that still remained
culturally in my odd little part of the world. I could see that a number of
things were going to go on in there. One of which, if history was to be any
guide, was that after a very free society had developed naturally in a very
free place, then another society would come and try to make money off of
it, and in the course of trying to make money off of it, would impose an
awful lot of control.

There has been a lot of unfortunate talk about the National Information
Infrastructure being a data superhighway. This is largely an artifact of
the fact that Al Gore's father was instrumental in creating the Interstate
System. So it's no mistake that Al Gore likes that metaphor. But in fact,
what has been going on lately reminds me a lot more of the development of
the railroad in this country. It is not a data superhighway so much as a
data railroad system that we seem to be developing.

There is a cautionary tale in there because the folks like Jay Gould and
his fellow barbarians who created the railway system in the west knew that
if they owned the roadbed, and the area around it, they also essentially
owned the society that was going to develop there. They could tariff
whatever products were going to be created in that society on the basis of
their own whim. The west today is still trying to get out from underneath
the burden of regulation and legal standardization that was created in
those early days by the railroad guys.

It was almost impossible for farmers in the upper midwest to make a living
for a while even though the Northern Pacific Railroad had asked them to
come in there and settle for nothing and had given them land. As soon as
they got established on that land they were charged usurious rates for
transporting their product to market. If we look at the history of the
railroad we can see exactly what kind of damage occurs when we give too few
people control over too much of the economy.

Actually I think it's far more useful to look at the development of the
Internet in biological rather than structural terms. The Internet seems to
me very much like a life form. It has all those characteristics. It is self
organizing. It adapts itself readily into the possibilities faced that it
finds. It is being created in an interactive way out at the margins rather
than in the center.

I've heard UNIX described as a virus from outer space, and it is very much
like a virus, I think, but it's more of a virus from inner space. The space
inside the cerebral cavities of many of the people in this room.

Among the notable characteristics of the Internet, outside of explosive
growth, is the extent to which it can naturally root itself around
problems. John Gilmore, who may be here and is known to many of you, said
something profound: "The Internet deals with censorship as a malfunction."
It really does. You see people trying to stop traffic in certain kinds of
intellectual material on the Internet. It simply routes itself around it
and gets that material distributed by some other pathway.

Unfortunately, the folks who are now entering the game ... to a more precise
extent, the organizations and institutions ... that are now entering the
game, are very different from what has previously has characterized the
development of digital networks in the world. They come at this with a
different paradigm of how the world works and how to create order. They
come at it with the notion that order is something that you impose and not
something that merges. They come at it thinking about their products as
something focused and centralized, that require large amounts of capital to
create, that are broadcast in a one-to-many medium.

I don't think these are necessarily bad people. But hey have a very hard
time understanding the modern digital environment. Most of the folks that
I've talked to from the television industry think that interactive
television consists of putting a "buy" button on your channel clicker. I'm
not kidding! I wish I were. They fail to understand that there is a
profound difference between information and experience. They are trying to
sell non-interactive stored information as though it were experience. And I
think that they actually believe that they are accomplishing that task.
They are going to try, in many ways ... some of them overt, some of them
unknown even to themselves ... to impose their culture and their metaphors on
this environment.

There are all sorts of ways in which their immune response system is
already working and I'll give you just one example. I was recently talking
to somebody from Viacom about the importance of creating interoperability
between whatever set top boxes Viacom was sponsoring, and other kinds of
networking ... specifically the Internet. I talked about TCP/IP with this
fellow from Viacom. He said, "Well we would love be able to incorporate
TCP/IP, but really, it's just too slow, the packets are just to big, they
can't be made to be isochronous. We really don't think that it has a place
on top of your television set."

While that may seem like an irrelevant factor to many of you who probably
don't even own a television set, if we are to create a society on the
Internet that is genuinely inclusive, and doesn't consist of its present
large band of wild geese, we are going to have to make it so that you can
get Internet connection from that electronic device that is your principle
access point into Cyberspace. That may well to be the set top box.

The folks in the television and entertainment business are also intrigued
by the possibilities that the railroaders first confronted, which is that
they are going to build the road bed essentially, and they're also going to
be in the information business. It's not lost on them that if you own the
rails and you're also shipping the cargo, you can get a really good deal on
your rates. Other people may not get such a good deal, especially if they
feel they are in competition.

This has given rise to a whole set of concerns and problems which the
Electronic Frontier Foundation is now dealing with. EFF did not start out
to be a traffic cop on the data superhighway. That wasn't our objective. At
the time Mitchell Kapor and I started EFF we had a very narrow set of
concerns. We thought that there were actions taking place on the part of
the Government that made it clear that they didn't quite realize that
speech was speech whether it was expressed in bits or ink on the page.

We felt that all we had to do was hire a few really scary civil liberties
attorneys from New York, kick hell out of the Secret Service, dust our
hands off in satisfaction, and go back to whatever it was we were doing.
But we hadn't been at this very long when I got some electronic mail from a
young fellow who was in what was still at that time the Soviet Union. He
said, "I applaud what you and Mr. Kapor are doing in trying to insure that
your First Amendment extends to Cyberspace, but what about us?" And we
realized for the first time that in Cyberspace the First Amendment is a
local ordinance.

That was a revelation to me, and also to Mitch. We started thinking about
what had to be done on a structural rather than a legal level to make
certain that people who connected to one another electronically would do so
without fear of reprisal for the things that they might think and say.
Mitch said something profound at one point: "Architecture is politics."

Now when I said that to something like the Academy of Television last week,
they didn't have the slightest idea what I was talking about, but I'll bet
the people in this room know that very well. It's a message that I think we
need to start carrying to the world in a much more forward and pro-active
way than is your natural bent.

I know that when I've talked to computer audiences in times past I've had a
continuous question and complaint from people in the room who say, "Well,
you know, you want me to behave as though I were a social philosopher and,
actually what I do is bus architecture." Well, exactly. I don't think you
can expect the social philosophers to understand bus architecture very well
for a while either. So the job falls to you and the people who understand
the basic nature of this very different environment.

When I was down in Los Angeles late last week, I attended something that
some of you may of heard about. This was  "Superhighway Supersummit." You
have never saw such self-importance in your life. It was unbelievable.

This next anecdote has nothing to do with anything, except it's more
evidence that I'm in a Thomas Pynchon novel. As a guest of the White House
I had a packet which included a discrete little part which read, "Those
persons who will be accompanied by a personal security assistant are
reminded that their assistant may not carry his weapon while in the
building." There were also two parking passes. One for regular cars, and
one for limousines. So they knew a fair amount about the culture that they
were pitching to down there.

The idea that this particular set of hooligans was going to be in charge of
the future was terrifying to me, in spite of the fact that, to my surprise
and satisfaction, I found people like John Malone saying all the right
things. It was very gratifying to hear that the things that EFF was pretty
much alone in saying two years ago were now politically correct.

But there is a great distance between being able to mouth the politically
correct thing and the actually having the kind of consciousness that will
promote those goals in a serious way. These folks are in business, they're
not in it for their health.

I looked around that audience and I realized that what I was looking at was
perhaps the ultimate expression of contemporary civilization. Which made me
start to think. Mitch and I had always talked about the job of EFF as
civilizing the electronic frontier. I think that our job, and your job,
increasingly, is going to be frontierizing civilization.

I believe that as a species we have gone just about as far as we can go by
design. If we are going to try to design society from the top, we will
continue to have the sort of results that they had in the Soviet Union and
at IBM.

The world is simply too complex a place to figure out. It's pretty good at
figuring itself out as long as you have an extremely open architecture, or
ecosystem, which supports ideas in a fluid and nutritious kind of way.
That's one of the great geniuses of UNIX. I have a Next machine ... I expect
a boo or two ... but that's as close to UNIX as I've been able to get. It's
kind of like UNIX with training wheels by Armani.

I have no personal aspirations to write a lot of shell scripts. I still
feel like C++ must be an exceptionally mediocre report card. But when I
look at the development of UNIX over the course of its existence, I find it
truly remarkable how this critter has grown. I sometimes think of it as
being the 1990's equivalent of Chartres Cathedral, where thousands of
people worked for many years creating something that was amazingly complex,
and yet somehow worked rather elegantly to the purposes for which it had
been created.

I look at UNIX as it continues to develop and I think that it will continue
for a long time to be the genetic code of Cyberspace. You have to approach
your work, I think, with that in mind. Of course the Government and the
large entertainment and television bodies that are now getting into this
really don't have a sense of how important it is to have an adaptive
organism as your substrate. They are not approaching it from that angle.

I am pleased to say that among the things that Al Gore announced down in
Los Angeles the other day had to do with opening up information
infrastructure to competition. In the past, as you know, most of the
information infrastructure in the United States was designed on the basis
of a regulated monopoly. We had for many years a stranglehold on the part
of AT&T which up until very recently was still requiring you to fill out a
whole bunch of forms to put a suction cup on your telephone.

So I'm very grateful to Judge Green, who took a lot of flack at the time
for having the insight to see that this stuff was going to develop much
more rapidly and much more openly in the hands of a lot of different
companies rather than one. The same thing is now starting to happen with
regard to the impending train wreck between the cable industry and the
telcos and the wireless industry.

These various industries have been regulated in the past by completely
different regimes originating in completely different places. Most of the
broadcasters have been regulated by the FCC ... and poorly, I might add. The
telephone companies are regulated by state public utilities commissions,
and most cable operators are regulated by municipalities. What they are
trying to do is create a system whereby all these different media can come
into direct competition with one another so that the path by which bits can
get into your home or office are so repetitious and so open that
competition brings down prices and creates bandwidth.

There is going to be, as you folks know well, an enormous desire for
bandwidth, and a lot of different agencies are going to be engaged in
fulfilling that desire. There will be plenty of business for them ... 
Bandwidth is one of those things kind of like money and sex ... the more you
got the shorter it feels.

Demands for bandwidth will also grow as we start moving away from text.
(Personally, I can hardly wait. I have a text allergy at this point. I get
kind of an ASCII glaze at the end of the week after 5 days of 100 to 150
e-mail messages a day, each one of which I have to read in order to
understand whether or not it's important to me. I want to see a lot of
richer data that has the kind of semiotic format that tells me right away
whether or not I want to mess with it, but it's going to take a lot of
bandwidth to do that.)

In any case, there are several bills already in Congress which EFF has been
pretty involved in helping create. There's the Markey bill, which is HR
3636, the National Communications Competition Information Infrastructure
Act of 1993 which would make it possible for cable companies to provide
telco services and visa-versa. And also make it possible for the national
long distance carriers to compete with the RBOC's in local
telecommunications.

There's another fairly similar bill in the Senate, the Telecommunications
Infrastructure Act, which is being promoted by Inouye. As Vice President
Gore announced on Tuesday, the administration is currently drafting an
amendment to the Communications Act which would include a whole new section
code called Title VII.

Title VII, if promulgated, would do something that seems pretty enlightened.
It's a promotion of a lot of the principles that EFF has been talking about
in Open Platform. It would essentially make it possible for
telecommunications providers enter into a fairly non-regulated regime if
they were willing to ensure complete openness of whatever channel they were
creating to whatever service or server might want to attach itself to it.
There is a lot of emphasis being placed on making certain that the Data
Superhighway is not 500 lanes in one direction and a foot path the other.

I can't tell you how important it is that these communications structures
are designed to be full-duplex. This does not resonate with the culture of
their builders, however. They don't know very much about getting that bit
back from the consumer. They are understandably a little afraid of what
will happen when the couch potatoes actually start to speak up about what
has been smothering them from their glass tubes all these years. It may
turn out that they don't really like this stuff very much and that they are
not going to be pleased by 500 versions of the air channel for men or the
ability to watch "My Mother the Car" at any hour of the day or night.

In spite of these fairly enlightened activities, I think that you will see
that Congress is even more inclined than ever to act "in loco parentis."
There are impending bills which would impose the necessity of having some
kind of technological switch on your set top box would sense incoming
violent content and would just circumvent its entry into your home.

This is obviously pretty bone-headed, but these are the kinds of things we
have to deal with. We have to make Congress and the various communications
providers recognize that the best way to assure family values, for whatever
family might be having those values, is to tag information in ways so that
people can make their own choices. There are ways to do that that are not
particularly demanding from the technical level. We do not need a society
which protects us from our own words.

One of the great things about talking to you guys is that I don't have to
go through a detailed history of the EFF or what we are doing. I know that
many of you, who had a natural affinity for the work we did defending the
freedom of speech in the very beginning, were baffled when we were suddenly
became something that looked like a telco trade organization, started
pushing ISDN, and dealing with telecommunications regulation. I think that
we did that for prudent and sufficient reasons even if we didn't
communicate those reasons very well to the outside.

I want to run down some of the fundamental aspects of Open Platform. We are
trying to promote the idea that there needs to be common carriage, much as
there has been through out the history of the telephone system. The phone
company certainly didn't try to regulate content over its lines.

The problem here is that common carriage under the telephone model was
protected by a regulatory regime which essentially gave a monopoly the
right to go on being a monopoly ... and a lot of incentive to go on being a
monopoly ... if they were to keep those lines open. Now it's a whole new ball
game trying to come up with a model for common carriage which does not
involve a regulatory overburden or monopolistic practices.  It's going to
be a very significant challenge. We don't have all the answers by any
means.

We have to work on interconnection and interoperability. When the fellow
from Viacom told me that TCP/IP had too much overhead, as a non UNIX
weenie, I didn't have a good response to him, except that it sounded to me
vaguely like a religious, rather than a technological statement. Which I
think it is. The people who know that and have sound evidence to prove it
will have to work hard on these people for whom  such beliefs are a canon
of faith.

You also need to be thinking about set top box and video architecture
protocols that will make it very easy for the telcos and the information
services to provide video fairly cheap bandwidth. You need to help the
Internet ramp up in a wise and orderly fashion. I know it's hard to do
under the current system. We all saw what happened to the Internet as soon
as Mosaic got out there. I'm somewhat concerned that if a lot more of this
goes on it's going to be very difficult to get traffic across the Internet.
There are some serious technical challenges. But there are things we can do
to solve them.

There is a new initiative that EFF is just starting up. We're trying to
work with the companies themselves, and these include some of the new
Internet based companies, to convince them that there is a business
advantage in allowing the people who connect to their system to use those
connections for whatever purpose.

I don't want to pick on Rick Adams, whom I assume is here, but I think it's
unfortunate that AlterNet and PSI and other commercial Internet providers
have forbade their customers from using their facilities for such
commercial services as bulletin boards. This is a debate that's going to
have to be carried on among you folks who are on the Internet and use those
providers. It may well be that you have to start looking to other IP
providers are willing to open up their lines to real communication and not
impose unnecessarily restrictive barriers to competition.

There is another set of issues that I think are going to be particularly
troubling and difficult to solve. I refer to cryptography and the private
protection of your data. The Government is really not even on the chart
with this yet.

Parenthetically, I have to insert another Thomas Pynchon moment here. I had
a really weird experience the other day. I managed to schmooze myself onto
Air Force 2 and ride back up here to San Francisco with Al Gore. A great
sense of unreality pervaded that experience for me, and the conversation
with the Vice President which took place along the way. Al Gore is a good
guy and a smart guy, but, when it comes to designing the National
Information Infrastructure, he has been strictly focused on those issues
around regulation and competition.

What he has apparently been thinking about very hard is cryptography. When
I started to open this line of discussion he went all blank and said, "We
have national security interests at stake." End of discussion.

I think we need to think long and hard about whether or not our national
security interests are actually addressed by trying to impose export
embargoes on cryptographic code. This strikes me as being like trying to
impose export embargoes on wind.

You can get MacPGP or PGP from FTP sites all over the world in seconds, so
I'm not quite sure what they are accomplishing, except that they are
accomplishing a chilling effect on the ability of American corporations to
incorporate robust cryptography into software and hardware which they might
design. Obviously it doesn't make a lot of sense to build a system that
incorporates levels of cryptographic protection that the NSA won't let you
to ship overseas. You don't want to have to build one system for the United
States and another system for overseas sales, especially in a business that
exports as much of its product as we do in the hardware and software
industries.

We have to get the Government to recognize the futility of crypto embargo.
It would be nice if they could also recognize that the Cold War is over,
but that may take some time. Even when they do figure that out: we still
have to deal with their Four Horsemen of the Apocalypse: kiddie
pornographers, drug lords, terrorists, and unnamed foreign enemies.

These monsters are rattled out at me every time I suggest it would be a
good idea to free up cryptography. I think they are all fairly illusory at
this moment. Assuming that we have to shut down privacy in America because
of terrorists doesn't make a lot of sense to me, when we only lost6 of our
citizens to terrorism last year. This is not quite the threat that the
Government portrays it to be.

Really what we have going on, I think, is the NSA acting as a stalking
horse for the FBI and other domestic law enforcement interests. They are
scared to death they are going to loose their ability to wiretap as analog
communications become some kind of digital fruit salad. They don't
see ... and there may be a fortunate quality to this ... they don't see yet the
technological opportunities that digitization will present them.

I think we need to analyze the problems and opportunities on both sides
here and deal with them accordingly. We may be hurtling toward a future in
which everything we do will be visible to the Government. As it is right
now, any time you make a financial transaction you smear your fingerprints
all over Cyberspace. This does not need to be the case, but it's going to
take a lot of consciousness changing to have it be otherwise.

For example, when I was talking to Gore the other day, he boasted about how
Government services were going to be a lot more efficient as the result of
a centralized card that people could use to get any money that was owed
them by the Government in disability payments, social security payments, or
whatever. They could simply go to a kiosk and insert their card to get
their payments. I asked him if there weren't some privacy considerations
that went along with this. I drew a complete blank. So we have a serious
problem.

By the same token, there are also serious problems to reckon with in giving
cryptography to everybody. I'm not certain I'm completely sanguine with the
idea that the advent of digital cash may create an economy in which taxes
become voluntary. At first blush that seems immensely appealing to me. I'm
sure it does to you too. But the problem with simply buying only as much
government as you think you need is that the people who can afford
government get it and the people who can't don't get it. You can see what's
happening already in the delivery of a lot of vital services.

Education has become privatized at the top. Mail has become privatized at
the top. I don't know anyone with an income of more than $50,000 who uses
the Postal Service when they want to send a package. They use Federal
Express or UPS.

The rich even have their own police. If you go down to Los Angeles, which I
guess you won't be able to do for a few days [there had been an earthquake
there that morning], you will find that in the wealthier parts of Los
Angeles, the local established government supported police force is not a
major element. The real police come from Westec. It's like something out of
"Snow Crash."

I think that we have to do something to detach financial transaction from
identity, or we are going to be in a serious mess. While I believe that the
current Government is, for all of its ineptitude, relatively benign, I'm
not sure I trust it with the power network-wide transactional analysis
might endow it with. As Lord Acton said, "Absolute power corrupts
absolutely." When the Government can see every single thing that we're up
to, I think that conveys to them a level of power that I'm not going to be
comfortable with they're having.

I don't think you should be.

There is also a whole set of extremely knotty questions about intellectual
property we're going to have to deal with. Again I'm pleased that I don't
have to explain to this audience that the digitization of everything
presents us with certain intellectual property challenges. You know a lot
about this. Believe me, the people in traditional media do not. We are
entering a situation where the primary article of commerce looks a hell of
a lot like speech.

Given the ambiguity of property law in this area, I think it's almost
impossible for us to say that free speech is assured when proprietary
interests will try to control its transport for their own economic
purposes. There is going to be a lot of that. There already is a lot of
that. There are other aspects of this that are more inconvenient than
threatening, but I don't see how we're going to avoid a complete collapse
of technological progress if we continue to put patents on things like
cursors.

I'm sure that many of you work for companies that now feel obligated to
patent every thought that happens to gel up in your head. I believe you
have to think about how you can serve Caesar, and at the same time, serve
the collective good of humanity. I am convinced that a lot of those
thoughts really are the collective property of humanity. Somebody once said
that art is what happens when God speaks through a human being. I believe
there's some truth to that. so it may be bold and arrogant to claim
whatever appears in your head. I think it's there for everyone. That's my
own personal belief and I'm sure I can get a good argument out of somebody
on that point.

When I discussed the intellectual property dilemma with Gore on the plane
the other day he said, "You're a songwriter and you must know that there is
already a system operating that deals in intellectual property that doesn't
have some physical manifestation, and that is BMI and ASCAP."

I said, "I'm a member of ASCAP, and if you think that's the solution, I
invite you to write some songs." ASCAP and BMI have a system for extracting
royalty payments from radio and television stations and distributing to
their members which is so disorganized and disorganizable that I look at it
ASCAP payments as being kind of like manna.

When I get a check from ASCAP I think, well, that's nice. I wonder if it
reflects anything real about actual radio play or television broadcasts. I
suspect not because they way they monitor air play is to have people
walking around with randomly selected tapes of radio broadcasts on their
Walkmans, writing down every song they hear and coming up with some very
crude statistical reckoning of these samples mean in terms of actual air
play. This is a sloppy system.

I think there are going to be other systems of intellectual property
protection which evolve, probably based on something more like a
performance model or a service model than instantiation in some physical
widget.

Those things are all going to go away and we're going to have to figure out
how to sell the wine of intellectual property without any bottles we've
been putting it in. I think that we probably will come up with something. I
can't imagine we're going to get very far into the Information Age without
any way of being paid for the work that we do with our minds.

We're going to have to change our perceived relationship with intellectual
property from one based on ownership to one based on performance and
service. We're going to have to look at ourselves in a continuous
relationship with the people who use our work rather than saying, "Alright,
I put my work in this box and the next time I give you this box it's going
to be a whole new transaction."

All of this is going to involve some fairly profound economic and social
changes. About the only thing I'm willing to say about them today is that
any projection made from this vantage point is liable to look ridiculous in
ten years. Everything is going to change very much. I really feel that what
we are essentially doing here is roughly like what the French theologian
philosopher Teilhard de Chardin was talking about when he started to write
in the thirties about the Omega point ... that point at which human beings
became so good at communicating with one another that they would create
what would amount to a great Collective Organism of Mind.

I think we are going to become such a creature. Perhaps we already are. It
is a very different kind of creature than has ever been seen in the
universe before. It will be enormously powerful and intelligent. And you
folks are helping it be born. Thank you very much.

[Transcribed, mostly, by Jeff Davis and, a little at the end, by Stanton
McCandlish]