HyperNetwork conference report part 2 of 2

[David Farber <farber () central cis upenn edu>]

controller or filter for such a broadband information base easy enough
to use. This is something that could be looked into in field tests. But
the emphasis of commercial systems is to provide the *current* type of
service on *demand* -- not a fundamentally different type of service
that is truly interactive.


All this means that there is little awareness or interest in two-way
interactive systems that are in fact two-way: as much room on the output
channel as the input channel.


There is some growing awareness that such real interactivity has not yet
been seen as a major business opportunity. This is unfortunate. One
thing that is changing is that the public interest in the Net is
changing. There's always something in the news these days about the Net.


Indeed, the Net has gone through big change. Nowadays, the Net is really
just the interconnection of many separate commercial nets. Hundreds of
businesses are now offering online services through the Net. Here are
some dimensions about this growth. There are now dozens of guide books
available about the Net for first-time users, containing all sorts of
information previously known only by the specialists and hard-core
users. There has been a rapid growth in the number of public-access
systems. These systems are now selling access to the Net at great
prices. There are new Net protocols, such as WWW and Mosaic and Gopher,
making easily available great depositories of information. And finally,
major services like America On-Line (AOL) are planning to offer full Net
services and access this year.


So, in the U.S., there are two *different* visions of the Infobahn. One
is put forward by commercial providers, and it envisions something along
the lines of 500 channels, with high-bandwidth delivery, and low-
bandwidth response. The other is put forward by the Net as it exists
today: truly equal two-way interactive communication.


The EFF supports the Net's vision, because it believes that this vision
is better for the public interest compared to the 500-channel approach.


What are the lessons of the Net? What can it teach us? First, something
about centralization: central control of the system is not necessary.
Technically, effectively, and socially, control is given to the local
providers and is enforced at the local level. Second, there's a
fundamental difference between fixed providers with controlled content
as compared to point to point communication, which is controlled by no
one, really. Thus, the interests of democracy are served by the very
architecture of the Net. People's interest will propel the development
of the system.


A question: how do we achieve this? Gore outline several principles: 1)
privatization: let the private sector build and run it; 2) competition:
local monopolies should change into regulated competitive markets; 3)
universal service: every citizen should have some way to get to the
network, that is, income and geography should not be barriers to access;
and 4) open access: all content providers should have the right to
connect to the network, and everyone should have the right to be a
content provider (i.e. the system should work along the lines of a
common carrier).


Another question: will these principles be turned into law? There are
still major business, technical, and political uncertainties. For
example, is this a real market? Will these technologies work? Can it be
pulled off? So, really, an activist, optimistic approach is needed.


Complications: the government is helpful in many ways in sorting through
these issues. Certainly, it has an important role to play. But as far as
privacy issues go, that's another story. And the EFF opposes the
government on many things in this regard, like Clipper.


Finally, it's important to understand that what we're talking about here
is a global information infrastructure. It's not just national. So,
international competition is necessary and Japan has a major role to
play.


3.8. Harry Saal, Smart Valley


Now is not the time to define a global system. We must experiment with
things locally, to find out about the global issues. Here's a list:


   Top-Down --> BAD!      Bottom-Up --> GOOD!
   ------------------     --------------------


   push                   pull
   pick winner            market decides, through competition
   central decision       distributed trials
   supply side            demand generation
   pork barrel            open process
   government funded      private investment
   risk averse            innovative


The reality: bottom-up is much messier, but it also works much better!
Here's the mess:
   - too many players
   - chance and likeliness of market and technological failure


With top-down, there's the uncontrollablity of changes in leadership.
And this isn't good when we need to define a social agenda and leave it
up to a system that can change unpredictably.


We must recognize, accept, and use to our advantage the reality that we
aren't good visionaries. We learn through experience. So, the system we
hope to achieve as a bottom-up approach is a good souffle':
   - high quality
   - industry-wide support
   - socially acceptable
   - market tested


[Some questions from audience.]


Q: What's the investment so far in Smart Valley? Where does the money
come from? How is it split up?


A: There are two parts to this answer. First, there's the actual
investment in Smart Valley, which is after all a non-profit coordinating
organization. Then there are the projects. For the first part, there are
very many member companies, and the membership per year is about $1
million. For the projects, there have been several hundred million
dollars invested so far. Most of this has been private investment, with
only a small portion coming from the government.


Q: Bottom-up is okay when things are going well, but when things aren't,
like in crises, where industry might rather flee, the government can
muddle through. So what about then?


A: There aren't only the two rigid choices between government and the
private sector. Smart Valley has a certain view of things, and we feel
many in the government agree: partnership between industry and
government can be successul. And let me say that people *do* feel there
is a cirsis *now*, and they're working very hard, and in partnerships.


3.9. David Farber, University of Pennsylvania


Let me start off by talking about Clipper. I think it's the
constitutional equivalent of an earthquake. Earthquakes destroy parts of
highways, making them unusable and difficult to rebuild. Also, they
destroy people's faith. Clipper would do that to the Infobahn: if it
goes through, people will lose their faith. They'll just not use the
thing.


Now, let me talk about Liberty Net. Gigabit networks are my usual topic.
Liberty Net is at the opposite end. It's being driven by community, not
industry. While the Delaware Valley, with Philadelphia at its core, has
a high concentration of universities, information centers, libraries,
etc., the area has seen hard times.


The role of Liberty Net is to build information streets into the city
centers of Philadelphia -- and elsewhere -- to create and improve a
better sense of community. It's being built by community leaders, such
as teachers, public radio, etc., who don't yet see the relevance of the
NII to public life.


The result over the last year has been the creation of a real network.
There have been contributions from area universities, for technical
know-how, and local businesses, who are trying to improve the community
among themselves.


For example, on an experimental basis, new entrepreneurs can communicate
with Liberty Library and Science Centers to consult and get support on
technology. The universities offer computing power, with such things as
community BBS's and event and offerings rosters. It's really "power to
the people" thinking.


Here's the current state of Liberty Net. It's been launched, there are
services, there's a staff with a director, and it has a budget, though
it's a small amount of money. In fact, this is what the goal is: nothing
extravagant, keep it small and responsive. In a real sense, Liberty Net
is a small-town network, similar in sense to the small-town meeting-
house metaphor. Liberty Net should be fully launched by 1996, as a fully
independent community network. Interestingly, it would then be fifty
year ago, in 1946, when the first digital computer, the Eniac, came
online at the University of Pennsylvania.


3.10. Edmund Tham, Singapore IT 2000 Project


[Mr. Tham presented a report on the creation of the "intelligent island"
using advanced network and computer technologies. Example uses include
connecting shipping companies with the docks, customs offices, and other
concerned parties to make the port system work much more efficiently,
without the need to put things on paper and route them around manually
or through the mail. More in the proceedings.]


3.11. COARA members


[Several reports were presented in a round-robin, whirlwind fashion on
the usages of networking and computers by owners of a construction firm,
a hotel, a store, and others. It was quite an impressive demonstration
of the breadth and depth that networking and computing have penetrated
in the community. The value of COARA in general, as a new community and
means of doing business in Oita, was really made clear. More in the
proceedings.]


3.12. Masamichi Tanabe, VP, Multimedia Planning and Promotion, NTT


> From early January, NTT decided to work on the basic ideas of its
approach to the Infobahn. But don't get the impression that this was
sudden: we've been preparing for this for some time.


In our view, cable and the assorted technologies associated with
convergence will soon become like a household appliance. The competition
in the US in this area is much more developed, largely because of the
breakup of AT&T 10 years ago. And now, cable companies there are
wondering what to do. Because of the technology, we see many
opportunities. First, we will have to continue being able to provide
normal telephone service perfectly. But after that, think of arealess
and borderless companies. We've been thinking about services, but we
need competition. So, we need to change our attitude towards competition
as well as affordable services.


Finally, there are five points I want to make. 1) We want to strengthen
our promotion of multimedia. We've been studying multimedia in many
areas, but we need to combine our efforts. 2) We need to promote the use
of the backbone network. We started this in 1993, and by the end of the
summer of 1994, we expect a network of 165 Mb/s for customers. We
already have an optical fiber network all over the country. But we need
to test this network using some distributed systems. 3) We need to look
at the services available now to households, using current telephone
lines. We plan to look at this in more detail in a year or so. 4) By
2000, we should get all this figured out, but by 2015 we hope to have
all homes connected to the network. Last, 5) having the network is not
enough. We need to have the interface. We need to have the users on
board as well.


3.13. Katsuhiro Onoda, VP, C&C Systems Market Dev. Div., NEC


We've been working on the Oita Digital House project. Talking about
multimedia, NEC has been thinking about it since 1973. That's when we
first thought of computers and communication. Since then, we've seen
many changes. The computer can now work on many things, like multimedia,
and it can work to be sensitive to humans.


Now, in our studies, we need to pay attention to the individual's power
and interest. But this is why we need something like B-ISDN. We need to
move to presenting information. Not just text, in order to explain, but
also images and sound, in order to *convince*. The network is important,
but it isn't so important. It's what will be *on* the network that's
important: the services.


We have an experiment in Kansai. We are looking at two things, one,
information and broadcasting, and another, sending characters over B-
ISDN. For B-ISDN, there will be business and other uses. In the autumn,
we'll use things like HDTV and electronic libraries in a set of 15
experiments to take place in the Kansai Science City. The MPT will have
a conference and observe and discuss the experiements. This is the
dawning of the ISDN world.


By 1997, the public version of this network will be created. What will
the services be? We need to arrive at some collective wisdom of what
will be created. But we can only start with a small network. The key
question is then, how can we make this grow? Perhaps this: start with a
set of points, add new points to make a line, then planes. So, we should
start next with a focal point, and then branch out. Maybe this means
moving to business next.


The U.S. NII is an example. Is it being driven by Internet? The focal
approach is bottom-up. But who fights the battle? The people who are
involved. The job now is to make a business out of information exchange.
And who has the information in the local area? Business and the
government! So, maybe the way to move is to cut them out of the picture,
connecting people together, to make new businesses.


We heard the example of Liberty Net: this is not a government effort,
but a community effort. In Japan, this effort would likely not work. So,
we have to work together. We need leaders, with fire in their mind, to
kindle the needed efforts. Let them fight the battles and pave the way
for what is needed. And we need financial support: government efforts
won't be enough, so it must also be private investment.


For example, the "Telecottage" efforts: 60 in the UK and many, perhaps
40, in France. Also, there are 10 in Australia. A telecottage is based
in the community, and it connects people together with each other. It
promotes new business, it is community based, with business services
handle by people on the network. A new community can be created through
cooperation with others.


3.14. Hirofumi Takanashi, Manag. Director, General Manager,
      Multimedia Systems Lab, Fujitsu Laboratories


[Presentation of the Monster workstation. More in proceedings.]


3.15. Koichi Utsumiya, Professor, Oita University


[Discussion of connecting together local, regional, and national
networks, using the Internet as an example. More in proceedings.]


3.16. Other


[Three other sets of meetings followed. The first session focused on
trying to understand what multimedia is, what is means for society, and
how it is being used *today*. Some examples: Kevin Kelly from WIRED
magazine talked about the "four facets of distributed being." Hajime
Nonogaki from Fujitsu talked about user behavior in Cyberspace, and
Fusako Hirabayashi addressed a similar theme with his talk about
"evangelizng and evangelism on the Net." A very interesting presentation
was given by Yoichiro Kawaguchi of Tsukuba University on combining
artificial life and networking to make art. This session went very late
into the evening.]


[The next session talked about the realm of the citizenry, corporations,
industry, and networking. Really, this was about the evolution of
citizens' interaction with each other through such things as virtual
communities, and about the evolution of traditional media. Howard
Rheingold of the Whole Earth Review and the Well talked about his ideas
of building virtual communities, as presented in his new book.
Representatives from Asahi Broadcasting, NHK, and Nikkei Newspaper also
talked about how the mass media will likely change as cyberspace becomes
more pervasive. ]


[In the wrap-up session, the talk became more rarefied, as discussion
led to discussion of global networking. High-level visions of the
development of the information infrastructure were related by a member
of COARA and a researcher from the Nomura Research Institute. This
latter presentation was noteworthy for its broad, national scope,
focusing on an "Information New Deal Initiative" for Japan, tying in
again with the theme of a social infrastructure. Finally, somewhat
baffling and pesimistic vision about the nature of cultural evolution
and the loss of distinction was presented by Fumihiro Nonomura of
Knowledge and Experience. But the idea at the heart of this particular
talk, that the Infobahn will tend to flatten cultural diversity,
possibly at some harm, is something that is worth serious
consideration.]


[Note that more on these sessions can be found in the proceedings.]


4. Conclusion


   Let me begin my conclusion with this recent news article:


TOKYO, JAPAN, 1994 MAR 8 (NB) -- The Japanese Posts & Telecommunication
Ministry says it will start an experiment involving high-speed
multimedia data transmission in cooperation with private firms. It is
called the Global Giga Network and will be completed by March 1995.


The Global Giga Network project aims to support a data transmission rate
of 10 gigabits per second. The network project will use supercomputers,
ATM (asynchronous transfer mode) switching devices, and various
databases.


It will be used in a variety of fields, such as videoconferencing,
remote medical diagnosis, and the exchanging of data on environmental
pollution.


To begin with, the network will be connected between computers of the
Posts & Telecommunication Ministry, NTT, KDD, and NHK (Japan
Broadcasting Cooperation) via NTT's fiber optics. The Ministry, NTT, and
KDD have a research institute for telecommunications. NHK also has a
research institute for broadcasting.


The Ministry intends to request the participation of research
laboratories, universities and private firms in Japan and abroad. It is
expected that Japanese computer makers, including NEC, Fujitsu and IBM
Japan, will also participate in the project.






   There can be no doubt that such an effort represents a sincere effort
by the government to help in the development of the Infobahn in Japan.
It is also sensible to think that many important technology and policy
lessons will be learned from this project. But the last paragraph
indicating an expectation of major companies to join in on this effort
gives rise to some concern about government intentions. It seems likely
that once becoming participants, companies in this effort will be guided
by a prescribed architecture -- rather than being guided by market
forces. As many of the speakers at the conference commented, such a top-
down approach may be the wrong one to take if the greatest success is to
be achieved. Instead, the markets of the Infobahn should be allowed to
develop naturally, using the bottom-up approach, even though, as some
speakers acknowledged, such a method may be difficult to apply in the
Japanese system.


   Government can play a pivotal role in promoting such competition. For
example, with respect to the above Global Giga Network, rather than
promoting a particular architecture or even network per se, the
government could develop through deregulation and selective funding an
open, level playing-field for information services. An invitation to
companies to participate would be unnecessary, since the chance to
standarized proprietary architectures coupled with a general profit
motive would be sufficient.


   Another role that government can play is demonstrated by the
successes in Oita. These, as highlighted by the conference, show that
community networking does work in Japan. Moreover, these successes are
being derived from partnership between government and industry at the
local level with national support. Valuable experience is being gained
and new markets are being developed and explored in Oita. People, as
individuals at the local level, are feeling the impact of such efforts
in their daily lives. The project also shows that through such efforts,
many important lessons about the particular nature of the needs for
computers and communication in Japan can be learned.


   The debate about Infobahn issues in Japan is healthy, and the success
of the conference demonstrates that finding a winning policy approach
for successfuly leading the convergence of networking, computer, and
media technologies is possible and worthwhile. Especially, vigorous
promotion of local networking throughout Japan may be the most promising
avenue for achieving the greatest overall success.