1
 2
 3
 4      New Frontiers on the Information Superhighway
 5
 6                 INTERNET TELEPHONY FORUM
 7
 8
 9                    September 4, 1997
10
11
12
13                        White Room
14                   National Press Club
15               529 Fourteenth Street, N.W.
16
17
18
19
20
21
22
23
24
25

                                                  2
 1                  P R O C E E D I N G S
 2              MS. BROWN:  Good morning.  Can everyone
 3    hear me all right?  Okay.
 4              I'm Kathy Brown.  I am head of the
 5    Office of Policy Analysis and Development at
 6    NTIA.  I'm delighted to see you all here.  This
 7    will be the first in a series of forums that NTIA
 8    will hold on emerging technologies.
 9              In the long tradition of NTIA and OPAD,
10    we thought it timely to start discussing some of
11    the new technologies that are emerging, and some
12    of the policy implications that arise with the
13    convergence of various networks, technologies,
14    new applications on old technologies, and this,
15    we're delighted, is our first opportunity to sit
16    together and to talk.
17              And we hope that today will be an
18    opportunity to stir the pot a little bit, to
19    understand what the emerging issues are, to
20    understand the technologies a little better.
21              For commercial advertisement here, our
22    next forum will be on November 6th and that will
23    be on wireless local loop.
24              So today, though, we're going to talk
25    about Internet telephony.  And I'm delighted you

                                                  3
 1    could join us.
 2              Larry Irving is the Assistant Secretary
 3    for Telecommunications.  He has been head of NTIA
 4    for now almost five years.  He has been a
 5    proponent, as you all know, of new technologies,
 6    the access to new technologies for the
 7    availability of these technologies for small
 8    business, large business and residential
 9    consumers.
10              As we talk about the deployment of
11    these new technologies, issues arise with respect
12    to the role of the industry, consumers and
13    government.  He has been a long-time advocate of
14    standing back and allowing innovation and
15    investment to happen as these things are going
16    forward, and I'm delighted that he will open
17    today's session.  If I can just locate him, I
18    think we're in good shape.  Larry.
19              MR. IRVING:  Good morning.  Let me
20    start by thanking Kathy, because really, these
21    forums, particularly today's forum, are the
22    result of her efforts and her concept of trying
23    to make sure that we understand it and that we
24    bring the experts in to help us understand it.
25              I also want to thank several other of

                                                  4
 1    my colleagues, but particularly Fred Lee and
 2    Becky from NTIA who worked so hard to pull this
 3    together.  They're just a few of the folk who
 4    have been working diligently night and day.  I've
 5    been watching them as they've been pulling this
 6    together.  And I want to thank all of you for
 7    coming to today's forum on Internet telephony.
 8              As Kathy has noted, I've been in this
 9    job about four and a half years.  My mother keeps
10    saying, Well, when do you get a promotion?  And I
11    keep trying to tell her, I like the job I have,
12    and the reason I stay is because it's absolutely
13    the best time in the world to do
14    telecommunication policy and telecommunication
15    technology.
16              Every now and then you think about what
17    else you could do in life, and I just can't
18    imagine anything more exciting than learning
19    about issues like telephony and getting to get an
20    understanding of what's happening out there in
21    the world.
22              This convergence is amazing, and I had
23    a chance to go to the Voice Internet's April
24    conference in San Francisco and just looking at
25    the technology and meeting the people, smart,

                                                  5
 1    young, focused, international, who are really
 2    trying to build something new and creating a new
 3    word, a good friend of mine, Jim, coined a few
 4    years ago, a paradigm of what the
 5    telecommunication communications world can look
 6    like.
 7              It's the innovators, the entrepreneurs
 8    and the policymakers who have to have a
 9    continuing dialogue on these issues to make sure
10    that we're promoting new technologies while
11    ensuring to advance traditional principles that
12    are important to Americans.  We can and should do
13    both, and I think that our belief, like today's
14    forum, will foster such a dialogue.
15              It's interesting because three years
16    ago, no one in this country was using the words
17    Internet and telephony together.  They weren't
18    used in the same sentence.  No one thought that
19    was really a concept that we would ever have
20    today.
21              Two years ago, no one in Washington was
22    using those words together.  And until this year,
23    really just this year, Internet telephony was
24    very much more of a hobby than a commercial
25    enterprise.  It was certainly of no great

                                                  6
 1    interest to the large telephone companies except
 2    as to something that might happen eventually.
 3              But as all of us know, things are
 4    changing fast.  Whether you look at Telecom
 5    Corporation of New Zealand, Daycomm of South
 6    Korea, France Telecom or GTE, telephone companies
 7    across this globe are testing this new
 8    technology, trying to figure out how to
 9    understand it, and trying to figure out what it
10    means for them.
11              And indeed, just last week Dutch
12    Telecom announced that it will invest in VocalTec
13    communications and use its products and services
14    to offer its customers Internet telephony
15    services.
16              Internet telephony is clearly gaining
17    momentum.  It is leading analysts to predict that
18    shortly after the turn of the next century, 50
19    percent of the world's communications will be
20    carried over the Internet.  Voice revenues will
21    exceed  $63 billion in five years, up from
22    $741 million in 1997.
23              So Internet telephony is unique from
24    other Internet applications in that people do not
25    need a computer to use the service and

                                                  7
 1    increasingly do not need a computer to use any
 2    Internet service, but certainly for Internet
 3    telephony you don't.
 4              Internet telephony can certainly work
 5    PC-to-PC, it is also working using phones at both
 6    ends of the call.  It will raise complex
 7    regulatory issues in this nation and issues that
 8    are going to be confronted on a global scale.
 9              We have already experienced how it is
10    sprinting ahead of the regulations.  It is
11    challenging the underpinnings of our telephone
12    network's distance pricing regime.  As I say
13    that, I can sense that a lot of folks who are
14    entrepreneurs and innovative are beginning to get
15    worried because two words you almost never use in
16    Washington -- three words never used together in
17    Washington are innovative, entrepreneurial and
18    Washington.  And so, I'm trying to figure out how
19    to calm the fears, because I know just a forum
20    like this raises the hackles on some people's
21    necks.
22              I have a friend out in Palo Alto I went
23    to school out there, who often says the best
24    thing, he likes to treat Washington like a
25    mushroom.  Keep us in the dark and throw

                                                  8
 1    fertilizer on us.
 2              Well, there are people in this industry
 3    who believe that the best thing they can do with
 4    regard to Internet telephony is keep Washington
 5    out of it and we agree with that.  Washington at
 6    this point has no need to regulate.  We do have a
 7    need to understand it, and that's what the
 8    purpose of this forum is.
 9              We're not -- let me repeat, we're not
10    holding this forum as any initial steps toward
11    regulation or oversight.  We're holding this
12    forum because the Clinton administration will
13    lead in regulatory forbearance and believes when
14    people are afraid of something, they're more
15    likely to regulate it.  When they understand
16    something, they're more likely to let it grow.
17              The Internet has grown faster than any
18    other communication means in the history of the
19    world and it's grown precisely because the
20    government has forbeared regulation, because we
21    stayed away from regulating this great new
22    opportunity.
23              Fortunately my bosses, President
24    Clinton, Vice President Gore and Secretary Daley
25    even more strongly agree on the need for us to

                                                  9
 1    stay out of this, to let this thing grow.  They
 2    have spoken eloquently, articulately and
 3      incessantly about the need to let new telecom
 4    communications grow and they've talked about
 5    looking to the private sector for these
 6    problems.
 7              There is no question that regulatory
 8    and policy questions will arise and there's
 9    absolutely no question that there will be some
10    who seek to involve Washington in the solution to
11    those problems.  We have to understand these
12    issues so we make the right decisions.  We have
13    to understand these technologies.  And that's why
14    we invited all of you here today.  We're going to
15    let you hear from the experts and I'm going to
16    stop talking now.  Thank you very much.
17              MS. BROWN:  I'll ask the first panel to
18    come up.  I'll just give you a couple of
19    administrative tidbits.  The rest rooms are down
20    the hall.  There are some phones down there and
21    there's a water fountain that way.  The VON
22    Coalition has graciously provided some
23    refreshments down in their demo room, so there's
24    some coffee and doughnuts or Danish down the
25    hall.

                                                  10
 1              We have a number of panels today that I
 2    hope will be worthwhile for all of us in terms of
 3    a dialogue, in terms of being able to talk to
 4    each other about the technology, and about the
 5    policy issues that arise as this technology comes
 6    to market.
 7              We hope that this day will be
 8    interactive.  We are -- we have planned this so
 9    that folks will talk for a bit, do a
10    presentation, and then we will have plenty of
11    time for questions and answers and dialogue, not
12    just questions and answers, but comments back and
13    forth.  So we hope you will take the opportunity
14    to do that.
15              We are also broadcasting over the
16    Internet and have the ability for folks to E-mail
17    in comments, questions, and so we will work that
18    also into the presentation for those of you who
19    are listening on the Net.  We have someone here
20    who will take your questions and will present
21    them to the group.
22              With that, I'd like to introduce Fred
23    Lee of my staff.  Fred is a senior advisor at
24    NTIA, is one of my closest senior advisors on
25    technology and policy issues.  He has been

                                                  11
 1    instrumental in pulling this forum together, has
 2    spoken to many of you in the industry.  I'm
 3    delighted that he will moderate the first panel.
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

                                                  12
 1              INTERNET TELEPHONY AT A GLANCE
 2              MR. LEE:  Thank you, Kathy.  Good
 3    morning and welcome.  To open our meeting this
 4    morning, we have assembled an interesting and
 5    informative group of people who will provide us
 6    with a brief overview of the Internet telephony
 7    market.
 8              I'd like to move quickly to the subject
 9    matter by introducing our three panel members.
10    To my left -- excuse me, to my extreme right is
11    Daniel Berninger who will speak first.  As a
12    communications systems consultant for VocalTec,
13    Dan advises corporate clients on the deployment
14    of his company's gateway products.  Prior to
15    joining VocalTec, Dan was a member of the
16    technical staff at Bell Labs.  In his
17    presentation, Dan will review the development
18    history of Internet telephony and offer his views
19    on why Internet telephony is different from
20    conventional telephony.
21              Seated directly to the left of Dan is
22    Larry Fromm, who will be our second panelist.
23    Larry is a vice president of business development
24    for Dialogic.  Larry is playing a major role in
25    Dialogic's Internet telephony efforts and he will

                                                  13
 1    talk about factors driving the development of
 2    this market.
 3              Finally, we will hear from Dr. Nicholas
 4    Economides, seated to my immediate right, who is
 5    a professor of economics at the Stern School of
 6    Business at New York University.
 7              Picking up from the brief overview of
 8    Internet telephony provided by our first two
 9    speakers, Nicholas will introduce a policy
10    perspective by suggesting a range of policy
11    issues the industry could face as it continues to
12    grow and mature.
13              With this brief introduction, I'd like
14    to turn the mike over to Dan to kick off our
15    first presentation.
16              MR. BERNINGER:  Hello.  Okay, thank
17    you.  I've been asked in the 20 minutes to get
18    started with the education process, to
19    essentially make you a 20-minute expert in the
20    technology issues of Internet telephony.
21              So far, so what am I starting with?
22    Who knows what the Internet is?  Not everybody
23    raised their hand.  Let's try that again.  Who
24    knows what the Internet is?  Okay.  And is there,
25    you know, any particular -- everybody here works

                                                  14
 1    for government.  Who works for the government
 2    here?  Everybody, right?  Is there any particular
 3    branch of government, like the Executive branch,
 4    raise your hand?  Legislative branch?
 5    Judiciary?  Nobody is missing, so let's get
 6    started.
 7              First thing is what is Internet
 8    telephony?  So let's get into that.  Now,
 9    Internet telephony of course has existed for a
10    long time, even before, you know, the three years
11    ago where it first came on the scene in the sense
12    that data networks have existed for a long time.
13              The story that was told to me was
14    that -- maybe somebody in this room can verify
15    it -- is that, you know, back when they were
16    originally putting together the data networks
17    that eventually became the Internet, you know,
18    the military was behind it and they said, Well,
19    the reason we need voice on this data network is
20    because we don't trust somebody pressing the
21    button to drop the bomb.  We want to hear the
22    president's order, we want to hear the voice.
23    And it doesn't have to be, you know, completely
24    great audio quality, but we have to be able to
25    detect this is the President.  So that the first

                                                  15
 1    message was -- you know, the test message over a
 2    voice over data network was, okay, drop the
 3    bomb.  And that -- but back then, in the early
 4    '70s, it was, you know, barely recognizable, at
 5    least according to the people that tell me about
 6    that.
 7              But the problem then, in order to do
 8    Internet telephony in those days required a
 9    mainframe computer on each end of the data
10    network.  But around February 1995, VocalTec
11    introduced a product that allowed people with
12    multimedia computers to talk to each other across
13    data networks.
14              And the interesting thing here was, you
15    know -- Wham, changed things.  All of a sudden,
16    it wasn't a mainframe computer at the end of this
17    data network, it was a multimedia computer and
18    just a PC that people had in their homes.  And
19    there was sort of a magic point around the
20    486 33 megahertz kind of, where you could
21    actually do this.  And now, of course, as things
22    have continued to improve, you can do it even
23    better.
24              Back then it was just basic audio.
25    Since that time in the two years or two and a

                                                  16
 1    half years, we've added a whole list of
 2    additional capabilities to just the multimedia
 3    computer talking to a multimedia computer.  But
 4    once somebody did that, they said, Uh-oh, there's
 5    only 30 million multimedia computers out there
 6    but 600 million phones.  So what was next?
 7              We got to talk to a phone.  First it
 8    was just the techies talking to the people,
 9    talking to their mother, grandmother that didn't
10    have a computer.  So they said it was computer to
11    telephone.  First it was just hop off.  Jeff
12    Pulver is in the room and he was behind the
13    effort called the Free World Dial Up that
14    initially set that up.  Those were the so-called
15    hobbyists that made that happen.  At the time it
16    was just audio.
17              Today we can do a lot more of things
18    with that.  We can do DTMF and we actually have
19    service providers providing that kind of
20    service.  Well, we got to the phone, so forget
21    it, let's just go phone-to-phone.  And about a
22    year and a half ago, we started out with
23    phone-to-phone, and then it was just audio.
24              But as we add features to that, we
25    improve the audio, now we have fax and different

                                                  17
 1    kinds of connections to the PSDN and sort of
 2    professional administration features.  You keep
 3    going and we're adding stuff to this and you get
 4    something very complicated and called
 5    communication over data networks and we'll get
 6    into more of that today.
 7              But really, it's not that complicated.
 8    Essentially what we have, and this is my block
 9    diagram I was asked to give.  We can slide
10    different things into this diagram.  But one --
11    the first thing you have is you got something
12    that someone's using, the end user device, the
13    premise equipment.  Maybe it's a telephone, maybe
14    it's a computer.  But there's really -- I mean,
15    it's just some terminal equipment, and the
16    difference in the terminal equipment is how I use
17    it.  Maybe has audio capabilities.  A computer
18    has a screen, so it's just a different user
19    interface.  It's just terminal equipment.
20    Various kinds of terminal equipment.
21              The next thing that had to happen to
22    make all this work was network access.  There's
23    lots of different ways to get access to my
24    terminal equipment.
25              In the case of a computer, I'm already

                                                  18
 1    on a data network, so I can get direct access.
 2    But in the case of a telephone, there's something
 3    between me and that telephone and today that's
 4    the public switch telephone network.  Of course
 5    there are lots of scenarios whether it's the
 6    analog local loop or we heard there's going to be
 7    a forum on the wireless local loop, so there's
 8    lots of options there.
 9              Number three is the edge equipment.
10    Now, traditionally the edge equipment is a large
11    central office but in Internet telephony we're
12    talking about just simply computer servers,
13    similar to what you would have in a typical ISP
14    that enable Internet telephony.
15              And so as we go around searching for
16    Internet telephony, it turns out that all these
17    things existed before Internet telephony, so
18    whether it was the network or the premise
19    equipment, the only thing that we added to enable
20    Internet telephony and which makes it awfully
21    interesting in terms of flexibility is we put a
22    little bit of software in -- in our computers at
23    the edge of the data network, and that turned all
24    these things into a telephony service.
25              Okay, so why do we care?  I mean, on

                                                  19
 1    the one hand, you know, why do we need to change
 2    anything, because the telecommunications network
 3    presumably already works.  And presumably it --
 4    the price is coming down and everything's getting
 5    better, so why bother to change anything?  Well,
 6    I guess because it's there.  No, I mean, the
 7    bottom line is that no, maybe it's not cheap
 8    enough and maybe it's not flexible enough, and
 9    maybe it's not reliable enough.  How do we know?
10    Because I mean, as it stands, we spend, you know,
11    more on telecommunication costs than we do on
12    oil.  And what do we get for it?  Okay.
13              Essentially the telecommunications of
14    today is essentially telecommunications of my
15    youth and if Alexander Graham Bell roamed around
16    the telecommunications network, he would kind of
17    feel pretty much at home.  So we're a little bit
18    suspicious that if things are so much the same
19    and we love consistency, maybe there's something
20    else that could happen.  And one of the
21    fundamental things that I'll make in terms of our
22    common goal of universal access is this issue of
23    cost.  How do I make it, the price performance
24    more reasonable?
25              Okay, so why do we care?  What's

                                                  20
 1    different?  The first sort of fundamental
 2    difference is in terms of how we are going to
 3    transmit this media, and my slide says packet
 4    transport simply costs less, okay.
 5              And this is sort of -- the simple
 6    analogy to think of this is that if I'm going to
 7    move your house and the mover came to you and
 8    said I'm going to do this in the most efficient
 9    way possible, I'm going to take all of your
10    furniture and I'm going to run it through a
11    shredder, it's going to fit into my moving van
12    much more efficiently.  And I can use a smaller
13    moving van and I can use up less gas and get it
14    there.  So now I get to it the other side, and
15    you have a bit of a problem.
16              But in the case of Internet telephony,
17    that's essentially what we do.  We take this
18    continuous signal on the one end and we shred
19    it.  We break it into lots of little pieces and
20    we ship it off across this data network.  And at
21    the other end, we put it back together again.
22              And the reason that's appealing is that
23    again, it's an efficiency thing.  If I have to --
24    you know, if I want to -- everybody to
25    communicate, let's say I have eight billion

                                                  21
 1    people that need to communicate.  On the one
 2    hand, I could hook wires from everybody to
 3    everybody, and that will allow everybody to talk
 4    to everybody else.  But then -- but you know, if
 5    you're talking to one person, you're probably not
 6    talking to the other one minus eight billion, so
 7    maybe you don't need all wires to everybody
 8    else.  You just need some kind of switching
 9    mechanism to allow you to talk.
10              But if you're going to do a switching
11    mechanism, you've got to decide well, how long am
12    I going to make that connection?  Should I do
13    that connection, sort of change it once a month
14    or should I change the connection -- maybe I'll
15    change the connection once a call.  So I'm going
16    to dial a phone number and that's going to give
17    me this continuous connection for the duration of
18    that call.  And that is in fact, it's a cheaper
19    way to do it than have wires to everybody in the
20    world.
21              But in fact, why stop there?  Because
22    really, the cost of providing the service is
23    fundamentally related to the duration of the
24    connection that I'm providing.  So we have
25    evolved the telephone network that it's sort of

                                                  22
 1    based on the -- the cost is based on a dialed
 2    connection, okay.
 3              And in fact, there's parts of the
 4    telephone network that are -- an administrator
 5    will set up a continuous connection that will
 6    last for months and that's -- you pay more for
 7    that service.
 8              But ideally, we would like a shorter
 9    connection time and that allows us to fill the
10    network with more voice.  So today, to accomplish
11    what we have in terms of, you know, blocking it,
12    if I pick up the phone, I want to get to the
13    other side.  Our telephone networks run dry in
14    terms of an efficiency level, you know, if you
15    aggregate all of time running about 1 percent
16    utilization, okay.
17              So now I've built this entire network
18    to provide a certain quality of service and in
19    order to provide that quality of service, my
20    level of overall efficiency, depending on how
21    good your marketing is in terms of how you can
22    fill up your network, is around 1 percent.
23              Whereas the Internet, even if it was
24    using the same technology, can based on the fact
25    of breaking things into little packets, can move

                                                  23
 1    me up to 30, 40, 50 percent and if your Internet
 2    service provider, he might be running you at 90
 3    percent and then you've really got some
 4    problems.
 5              So the idea is we want to move up the
 6    efficiency, we want to shred that thing.  We
 7    don't want to keep it in one piece.  And the only
 8    problem with doing that, again, is putting it
 9    back together again.  And the reason we weren't
10    having this conversation a while ago was because
11    we didn't have the edge processing power to
12    enable us to do that.  So again, it didn't make
13    sense to have mainframes doing that.
14              Now, the other thing that's driving all
15    of this is the fact that this idea,
16    nonproprietary open solutions, costs less, the
17    idea of the PC computing versus the large
18    mainframe.  And there's really nothing magic
19    about, or special silicon, with respect to
20    nonproprietary versus open.  It's just how you
21    break up the problem.
22              So when we say open, that means for a
23    given problem, we allow different companies or
24    different entities to work on that same problem.
25    So rather than AT&T being vertically integrated

                                                  24
 1    and solving all of the problems with respect to
 2    technology -- telecommunications, we allow
 3    different kinds of companies to apply sort of an
 4    economy of scale to solve the problem.
 5              So we've had Intel, you know, working
 6    on microprocessors for reasons that had nothing
 7    to do with Internet telephony, and so we get to
 8    ride on their shoulders and benefit from that.
 9    And the same thing for operating systems and
10    et cetera.
11              So we would like to build our
12    telecommunications network out of things that are
13    built on -- built by other people for other
14    reasons and other people pay for those
15    development costs.
16              So Internet telephony did not pay for
17    the development cost of the Internet, we did not
18    pay for the development cost of computers, we did
19    not pay for the development cost of operating
20    systems and so we're riding on top of all that
21    and we would like to reuse that.
22              In the picture, there's something
23    called all this adds up to Moore's law, which
24    hopefully you've heard about.  It's certainly one
25    of those buzzwords.  It simply means that every

                                                  25
 1    18 months, we double price performance.
 2              So one of the sort of fundamental
 3    interesting things about what we're doing here is
 4    that you put all this together and every 18
 5    months we double the price performance and that's
 6    something that does not happen in traditional
 7    telecommunications.
 8              Now, the other thing, again, I'm trying
 9    to think about why it is we regulate and what our
10    goals are, and one of the goals is reliability.
11    So the situation here is the same thing.
12              Packet reliability or Internet
13    telephony reliability actually costs less than
14    sort of traditional.  And the traditional way of
15    doing things is sort of the switch is a
16    battleship approach -- that this thing is so
17    reliable it will not go down.  And in order to
18    achieve that, it costs us a lot of money.
19              Now, the thing is, we don't really care
20    if switches are reliable.  What I really care
21    about is if my end-to-end service is reliable.
22    So I want to pick up the phone and I want to be
23    able to get to the other end and that really has
24    nothing to do with switching and there's other
25    ways to do it.  So what packet telephony does is

                                                  26
 1    it provides route diversity, whereas in the
 2    switch network there is route diversity but I can
 3    count it on my hand.
 4              In a packet network, you have a very
 5    large number of route diversity and the
 6    reason that was sort of designed from the top
 7    down in the sense that the packet networks come
 8    again from the military application whereas
 9    they're saying we're going to blow up this part
10    of the country and we still want the thing to
11    work.
12              In fact, it was tested in the Iraq
13    war.  This is another story maybe somebody here
14    can confirm is that we could not take out Saddam
15    Hussein's command and control network because he
16    was running IP and we just kept on bombing it and
17    bombing it and it just kept on going.  And so it
18    did what it was supposed to do.
19              And there's some other things too,
20    aside from just at the router level where you can
21    work around problems, and that is at the
22    application level.  So now VocalTec, in deploying
23    our applications, we can control reliability.
24              And in a traditional network, I have a
25    phone and it's as reliable as it is.  It is as

                                                  27
 1    reliable as the wires between it.  At my
 2    application level, I can send out two packets for
 3    every one that I generate, and then I can collect
 4    at the other end -- I said well, I didn't get one
 5    but I still have this other packet.
 6              So now the application person, the guy
 7    that's just doing that little piece of software
 8    can have an effect on reliability.
 9              Now, we talk a lot about, you know, the
10    cost of things and as I said, sort of
11    fundamentally, the Internet telephony approach
12    does make things a lot less expensive.  But the
13    thing is that what actually drives it from the
14    other side is it actually gives us a lot more
15    reasons to use a telecommunications service.
16              So Internet telephony did not invent
17    this idea that new services increase revenues.
18    Heck, we knew that a long time ago.  And at Bell
19    Labs, that's what I did at Bell Labs.  We were
20    about figuring on new ways to use the telephone
21    network.  The problem we found is that this
22    network and this infrastructure was sort of
23    invulnerable to new services.  You just could not
24    do it.  So over a period of, you know, 25 years,
25    you end up with caller ID and some things that

                                                  28
 1    are interesting and we need, but somehow you feel
 2    a little let down.
 3              And so the question is Internet
 4    telephony all of a sudden and was one of the
 5    reasons why I gave up a perfectly good job at
 6    Bell Labs, is that all of a sudden now you have a
 7    platform at the edge of a network that will allow
 8    to you create services, okay.  And it's going to
 9    get a whole bunch of people working on that
10    thing.
11              For example, AT&T and the large
12    companies have -- and traditional telecoms have
13    lots of people to write software but nowhere near
14    as many people as that are working on software
15    for the Internet.  For example, I mean just JAVA
16    is something that came up not so long ago.
17    There's already 400,000 people working on JAVA
18    applications.  So what we need is these sort of
19    massive creative people working on applications
20    for telecommunications.  The easy part is we have
21    lots of people working on it.
22              The other important thing about
23    Internet telephony is that the network itself
24    really knows nothing about whether it's voice or
25    whether it's video or whatever, and that's really

                                                  29
 1    important.  Because I do not want to have to
 2    reinvent my network every time I want to change
 3    services, which is the case today.
 4              So the network today was designed for
 5    voice and that -- anything else you do on it is
 6    kind of a hack of some kind.  But the data
 7    network itself could care less as to what it is.
 8    Now, there's definitely performance issues there
 9    and there's some sensitivities but for the most
10    part, it's generic.
11              The other nice thing is that you
12    essentially can make mistakes in Internet
13    telephony, whereas you can't in traditional
14    telecommunication.  Imagine the AT&T executive
15    that's going to deploy a new service.  I am going
16    to have to deploy this to 90 million people and
17    make it work.  And you're going to make a
18    decision and decide about it for three years.  On
19    Internet telephony, I do it on one computer and I
20    try it out with a few customers and I'm done.
21              I'm getting the red light here.
22              The idea is there's no fundamental
23    obstacles.  In fact the fundamentals are on our
24    side.  The speed of light is not going to stop us
25    or anything.

                                                  30
 1              But let me just run through real quick
 2    some existing challenges.  We need to reduce the
 3    cost of doing it.  We need to sort of figure out
 4    how to charge customers for it.  We need to
 5    reduce the cost of IP transport.  We need to work
 6    on the platform to increase port density.  Offer
 7    a range of features.  Cost, how do we achieve
 8    universal access?  Again, it comes back to cost.
 9    We need to increase the ubiquity of IP.  Again,
10    these are all things that are happening
11    independent of Internet telephony.  Different
12    PSTN connection options.  Interoperability
13    standards, reliability.  We make it cheaper, we
14    can do redundancy.  Increase voice quality.
15              So there's lots of things that have to
16    be done, but fundamentally, these parts that
17    we're working with make us a lot more excited
18    than what we had in traditional telecom.  Thank
19    you.
20              MR. LEE:  I should mention that we are
21    going to be assembling a report containing the
22    materials from this paper, so if those slides
23    flew by fairly quickly, you will be able to see
24    hard copy at some point, and probably we're going
25    to put some of them up on our Web site later,

                                                  31
 1    too.  Larry?
 2              MR. FROMM:  I think I'll build on what
 3    has been talked about and use a variety of
 4    applications to set those.  To set the context
 5    for my remarks, Dialogic Corporation provides
 6    enabling components that our customers use to
 7    build a wide variety of applications that
 8    automate phone calls.
 9              The applications that have received a
10    lot of interest over the last year or so are
11    those for connecting IP networks to phone
12    systems.  We began working with VocalTec in 1995
13    to take their IP telephony technology and bridge
14    to it the telephone network so you can now use
15    the most common mobile, lowest-cost, easiest to
16    use terminal, the phone, to take advantage of all
17    that IP telephony has to offer, and worked with
18    VocalTec to introduce the world's first
19    IP telephony gateway last year and continue to
20    work closely with them.
21              In the true spirit of open systems, we
22    have competition that provides enabling
23    components.  VocalTec has competition at the
24    application level.  And as you'll see, the idea
25    of open systems as applied to communication

                                                  32
 1    systems will spur the same kind of rapid advance
 2    that we've seen when competition has been applied
 3    to open systems in the computer systems.
 4              I think the drivers for IP telephony
 5    boil down to two things.  The first is cost.
 6    Part of that is the fact that sending full duplex
 7    audio over data networks today avoids access
 8    charges and avoids settlement charges and that's
 9    clearly a topic of discussion but as Dan
10    mentioned, with settlement charges coming down
11    around the world, with access charges coming
12    down, that's not a permanent factor.
13              I think strategically the real driver
14    is that today, we have two networks for
15    communications:  We have a telephone network and
16    we have a data network.  And with IP telephony,
17    we can now converge those two networks into a
18    single network for all of our communication needs
19    with consequent savings in the cost of deploying
20    and probably even more importantly maintaining
21    communication networks.
22              Now we have a single communication
23    network that we have to spend our attention and
24    focus on, and that applies whether you're a
25    public service provider, whether you're a company

                                                  33
 1    deploying a network within your building, or
 2    whether you're a company deploying a network
 3    between your buildings.  And I'll show some
 4    examples that illustrate all of those.
 5              You might ask if it's so economical to
 6    combine these two different networks together
 7    into one, why hasn't it happened before?  Dan
 8    touched on some of those factors.
 9              I think a simple answer to that is that
10    people have tried to do that before, tried to run
11    data over voice networks.  But as Dan mentioned,
12    trying to take a network that's specifically
13    tuned to running voice and trying to put data on
14    top of that hasn't worked out that well.  You
15    just can't sent that much data across something
16    that's tuned for voice.  Flipping it around,
17    taking a data network and adapting that to carry
18    voice turns out to be much better and we can get
19    very high quality voice transportation on top of
20    those data networks.
21              So to me, the answer of why is it
22    happening today is the Internet.  The Internet
23    has been exploding, of course, and now for the
24    first time with IP technology, we have a global
25    worldwide ubiquitous data network that we can run

                                                  34
 1    voice on top of and so we are.
 2              Some people have projected that within
 3    five years, in the United States, we'll have 10
 4    times the capacity, the bandwidth for IP networks
 5    that we do for telephone networks.  In five years
 6    we can take the entire telephone infrastructure,
 7    if these projections are true, and put it on top
 8    of our IP networks and not even notice the impact
 9    that it has on our IP bandwidth capacity.
10              The second factor driving IP telephony
11    is the increased features and capabilities.  As
12    Dan mentioned, these new communication devices,
13    these new IP telephony communication devices are
14    fundamentally open systems, which means there's
15    competition at every level, which means that
16    there are now dozens or hundreds or thousands of
17    companies that are creating new applications for
18    the way that we communicate.  And as a result,
19    we'll see some companies compete on cost, some
20    companies compete on reliability, some companies
21    compete on quality, some companies compete on
22    features.  We're giving more choices and just
23    more competition and everything that that
24    provides.
25              A company like VocalTec, a start-up

                                                  35
 1    company, revolutionized the way we communicate.
 2    What I'd like to do is step through some of the
 3    applications that our various customers are
 4    building to illustrate some of these themes on
 5    how they play out in real-world applications.
 6              The first is what is called next
 7    generation telcos or Internet telephony service
 8    providers.  These are companies that offer
 9    worldwide telephone service using data networks
10    as a backbone.  Receiving a lot of attention when
11    people think of IP telephony, this is the
12    application that most people think of.
13              A couple comments here.  One is that
14    there are at least 60 companies today that offer
15    telephone services over IP networks.  A lot of
16    entrepreneurs and with the Dutch announcement
17    last week, you can see some of the major service
18    providers are moving into this network.
19              What this is allowing, though, is
20    really diversity of choice.  People can choose
21    the quality of service that they'd like and pay
22    accordingly, and we're seeing some new
23    interesting economic models for developing
24    telecommunications services.  There are some
25    companies that offer franchises as an

                                                  36
 1    entrepreneur.  You can buy a franchise, market
 2    telecommunication services in your local area,
 3    connect to other franchisees and the franchisor
 4    will set up a clearing system to allow you to
 5    offer service on a worldwide basis while you
 6    operate a market locally.
 7              Other companies are employing barter
 8    mechanisms and other economic models for
 9    delivering telecommunication services.
10              Another really interesting application
11    that illustrates this theme of where we used to
12    have two networks, we're now converging into one
13    is something that I called last mile
14    applications.  The idea here is that if you are
15    at home, logged on to Internet, if you're like
16    most Americans and certainly most people are
17    around the world, you've tied up your single
18    phone line to your home.  So anyone trying to
19    reach you would get a busy signal and likewise,
20    you can't dial out.
21              By applying AIP telephony technology,
22    we can now take that incoming call, convert it to
23    Internet telephony and send it over your Internet
24    link to your multimedia PC where a window pops up
25    that notifies you, you have an incoming call and

                                                  37
 1    give you the option of receiving that call,
 2    talking to the person that's calling you while
 3    you remain on line.  You've taken your voice
 4    conversation, your Internet session and put them
 5    over a single link, a single wire to your home, a
 6    single IP connection.  Likewise, you can dial out
 7    while you remain on line.
 8              The way this works is that your
 9    Internet service provider as an Internet
10    telephony gateway, forwards your central office
11    to do a busy forward, just like you would do for
12    central office voice mail or any kind of call
13    forwarding technology that you can get from your
14    local service provider today.
15              When it's terminated at the Internet
16    service provider, the gateway box answers that
17    phone call, converts it to Internet traffic and
18    sends it to you over your Internet link.  You can
19    now have both your voice conversation and your
20    Internet conversation over a single phone line.
21              This allows dual use of a single wire
22    to the home.  Instead of having to pay for two
23    networks to your house, in this case it would be
24    two separate phone lines, you can have a single
25    network for your voice and Internet connections.

                                                  38
 1    Since these gateways are located close to your
 2    home, there's really none of the latency problems
 3    that are apparent when you use the public
 4    Internet for -- for audio traffic.  It has the
 5    quality of service very high and it's a very good
 6    application in the sense that long distance
 7    carriers tend to like it.  It increases their
 8    call completion rates.
 9              Local carriers in some cases like it as
10    well.  They don't need to deploy a second phone
11    line and in some places in the world, they can
12    charge you per minute to get -- to deliver those
13    phone calls.
14              ISPs have a new application to offer
15    their customers and of course as a consumer, you
16    now have the option of accepting those phone
17    calls while you're online without having to pay
18    for a second phone line.  This kind of
19    application is being deployed around the world
20    today by companies like Ericcson, by VocalTec and
21    by others in various stages of field trial and
22    deployment.
23              Another application that's receiving a
24    lot of interest is the idea of call center
25    integration.  Certainly more and more commerce,

                                                  39
 1    more and more activity is taking place on the
 2    World Wide Web.  You can now go to the Web to
 3    find out things that you might want to buy from
 4    a -- from some kind of retailer, and can you also
 5    call them if you want to buy something.
 6              Now by bridging the two with
 7    IP telephony, you can go to the Web to find out
 8    the information you want but still have an option
 9    to call the company to talk to a live agent and
10    find out more details about when it might be
11    delivered, give your credit card number over an
12    audio channel instead of over the Internet or
13    whatever else you might want to ask that agent.
14    It allows us to again converge what used to be
15    two separate activities, our data network and our
16    voice network, into a single network.
17              The World Wide Web page puts on a
18    button that says, Call Agent.  You can be
19    connected to the company's call center over
20    IP telephony.  Speak into your microphone, listen
21    over your speakers and via gateway technology.
22    Those calls are delivered to a call center
23    agent.  So the call center agent, those calls can
24    look like any other calls that might come in from
25    regular telephone users, but in this case they

                                                  40
 1    originated on the Internet.
 2              These gateways can be located near the
 3    call center or they can be located somewhere in
 4    the network or they can even be located closer to
 5    you at your ISP point of presence and we have
 6    different customers that are deploying them in
 7    different ways today.
 8              One of the reasons this is interesting
 9    to call center operators is they like to funnel
10    more and more of their traffic to the World Wide
11    Web.  Let customers go to the Web first to try
12    and find their information, also select what
13    services they want.
14              MR. PULVER:  We'll be demonstrating
15    this, this afternoon.
16              MR. FROMM:  Larry says they'll be
17    demonstrating this afternoon.
18              The call centers are hoping more of
19    their customers will go to the Web first.  They
20    can get to an agent quickly and easily when they
21    want to.
22              The next step then is to take this
23    IP telephony technology, take it from the wide
24    area, funneling calls into the call center and
25    apply to it the local area, delivering calls to

                                                  41
 1    the agent.
 2              Every call center in the world or at
 3    least every call center in America today has two
 4    links to every call center's desk: a data link so
 5    they can find out information about the customer
 6    over their computer network and a voice link so
 7    they can talk to that customer.  Again, the idea
 8    of two separate networks for communications.
 9              Using IP telephony technology, you can
10    combine those into a single network, offering a
11    single link to that agent's desktop.  The
12    advantages are reducing the cost of deploying and
13    maintaining those communication channels, but
14    perhaps even more importantly, it provides more
15    flexibility to the call center operator for how
16    they deploy those agents.  You can now think
17    about more flexibly deploying agents at various
18    locations around the country, even employing
19    agents at home for peak hours and now you only
20    have a single communication link.  It becomes
21    easier to coordinate the voice and data traffic.
22              I think when Shane from Netspeak
23    demonstrates later today, he may be demonstrating
24    this as well.
25              Taking it one step further, we can now

                                                  42
 1    virtualize switches more generally.  PBXs.
 2    Again, within every office building in America
 3    and most places in the world, there's two
 4    communication networks.  There's a local area
 5    network for our data traffic and there's a phone
 6    network for our telephone traffic.
 7              Again, by being able to combine both of
 8    those onto a single network, we save the cost of
 9    deploying and maintaining those separate networks
10    and in addition, we offer a new set of audio
11    communication features and capabilities by
12    linking the two communication networks together.
13              It's now easier to do things like share
14    applications with a colleague while you talk to
15    them, collaborate on a Power Point presentation
16    or a Word document while you're talking to
17    somebody because all that's happening over a
18    single network.
19              Being able to share and send video
20    along with the audio, it offers a richer form of
21    communication and also because all these various
22    applications can be desktop software deployed by
23    any number of thousands of entrepreneurs around
24    the world.
25              I predict and I think it's an easy

                                                  43
 1    prediction that we'll see more innovative
 2    communication applications that make our jobs of
 3    communicating easier just like we see in a vast
 4    diversity of computer applications given the rise
 5    of open computer systems since 1980.
 6              In all these virtualized switch
 7    applications there's the role of a gateway to
 8    connect the telephone network, the public switch
 9    telephone network to these IP networks, whether
10    the IP networks are providing wide area access or
11    providing our local area access.  These two
12    different networks that we have, voice and data,
13    aren't going to collapse and converge overnight.
14    It will be a gradual process, probably spanning
15    decades, and the role of the gateway is
16    in between those two worlds, allowing traffic to
17    convert from one to the other.
18              Another application for this is
19    corporate intranets.  Again, today most companies
20    have a data network between their branch offices
21    and a voice network between their branch
22    offices.
23              By deploying IP telephony gateways,
24    they can now begin sending some of their voice
25    traffic onto their data network and eventually

                                                  44
 1    eliminate the need for their voice network
 2    altogether, collapsing all their traffic, again,
 3    onto a single network.
 4              There are techniques and methods for
 5    doing this today, things like voice over frame
 6    relay.  We believe that voice over IP is very
 7    powerful because IP is the common denominator for
 8    companies' communication traffic today.  By using
 9    the same protocol, the same techniques, it is not
10    only more efficient, but it's also very familiar
11    to the people that operate and maintain those
12    networks, so there's less training and less
13    maintenance cost.  These kinds of applications
14    are starting to be deployed today with companies
15    like Motorola and others offering these
16    capabilities.
17              One of the advantages of deploying
18    IP telephony in terms of corporate intranets is
19    that those networks are typically well managed
20    and you can manage the bandwidth and latency so
21    you can deliver a high quality of service over
22    existing networks.  Companies are investing in
23    their IP networks anyway by putting voice traffic
24    on top.  It's just a natural extension of their
25    communication networks.

                                                  45
 1              It might be interesting when you go
 2    back to Europe offices, whether you work for the
 3    government or a private employer, to find out how
 4    much IP bandwidth you have at your place of
 5    employment versus how much phone bandwidth you
 6    have at your place of employment.  I bet if
 7    you're like most companies you have vastly more
 8    IP capacity than you have voice capacity and I
 9    think there's no doubt that the IP capacity is
10    what's growing the fastest.  At Dialogic, for
11    example, depending on what you're talking about,
12    we have between four and 50 times as much IP
13    capacity than we do voice capacity within our
14    internal networks.
15              Another great application and Dan
16    touched on this a bit for IP telephony is fax
17    over the Internet protocol.  Fax is a great
18    application for IP telephony for a couple
19    reasons.  One, is a minor amount of latency, a
20    half a second or third of a second, does not
21    affect the quality of a fax transmission like it
22    would the quality of a voice transmission.
23    Two, there's enormous amount of money being spent
24    to send faxes long distance, at least $30 million
25    a year, it's probably higher than that.

                                                  46
 1    And third, the whole issue of -- the regulatory
 2    issues with regard to fax are a little cleaner
 3    than they are with voice.
 4              Fax is technically an enhanced service
 5    which isn't subject to the access charges and
 6    settlement charges, and therefore it's a pretty
 7    cleaner regulatory perspective.  We've seen
 8    companies like UU-Net and others begin to offer
 9    fax over IP service as an extension of their ISP
10    services already.
11              Just to touch on standards and some of
12    the technical issues that the industry is
13    addressing.  There are a variety of forums for
14    companies that are developing technologies in
15    this area to work together to make sure we're
16    working on a common set of standards to ensure
17    interoperability.  Probably the most relevant
18    specifically directed towards this activity is
19    the voice over IP forum, which was founded last
20    year by VocalTec, Cisco, Microsoft, Dialogic,
21    U.S. Robotics and others.  It's now part of the
22    International Multimedia Teleconferencing
23    Consortium.  This group meets on a regular basis
24    to agree on interoperability agreements, so that
25    devices, both computer software as well as

                                                  47
 1    gateway devices from different vendors can work
 2    together.
 3              The next meeting of this is going to be
 4    in September in Boston at the tail end of the
 5    Voice Over Net Conference.  I believe there's
 6    more information on that conference in the demo
 7    room if you're interested.
 8              One of the things that this group
 9    sponsors is regular interoperability forums where
10    vendors of different kinds of products can get
11    together in a very noncompetitive cooperative
12    method just to make sure products work together.
13    There's fierce competition at every level, but I
14    think there's a good awareness within the
15    industry that having products work together is
16    very important for the growth of the industry and
17    the diversity of applications and services that
18    we can offer.  And so about every two months,
19    there is a convention of companies so that they
20    can test their products with each other to make
21    sure they work together.
22              Here's I think the current member
23    group.  It may have grown a little bit since I
24    printed this chart of various companies that are
25    part of the Voice Over PC forum.  There are

                                                  48
 1    computer vendors, networking vendors, some of the
 2    telecommunication vendors and many service
 3    providers are all part of this activity.
 4              I won't go into the technical issues
 5    but just address real briefly some of the topics
 6    that this group is addressing.  The first is call
 7    control, the ability to have different devices
 8    signal to each other and set up a communication
 9    channel between each other so that products from
10    different vendors can talk to each other.
11              The second is the ability to compress
12    the voice in a common method.  This is the
13    shredding that Dan talked about, the ability for
14    each device to shred the voice the same way and
15    to reconstruct it the same way so that if you
16    have different movers from different companies,
17    they can still get your house back together when
18    it gets delivered to the other side.
19              The interoperability events that I
20    mentioned and also directory services, this is
21    probably the area that has the most work yet to
22    be done in terms of industry agreement, which is
23    if I want to communicate with somebody, Jeff
24    Pulver, for example, how do I find out what his
25    address is on the Internet?  There are a variety

                                                  49
 1    of means that different vendors have deployed to
 2    do this.  Right now they do not all yet work
 3    together, but there is a discussion on how to
 4    make that happen.
 5              That's a brief overview from my
 6    perspective.  Thanks for the time and turn it
 7    over to Nick.
 8              MR. LEE:  Thank you, Larry.  Let me
 9    remind you that we are going to have a
10    demonstration right after the lunch break with
11    Jeff and Shane Mattaway showing off their call
12    center and gateway applications.  Also, we invite
13    you to visit the demo room if you want to get a
14    hands-on view of this.
15              MR. ECONOMIDES:  Hello.  I'll be using
16    a more old-fashioned method of presentation, that
17    is, transparencies rather than the high-tech
18    stuff with the computers.  So it will take a
19    couple of minutes for it to be set up.
20              My name is Nicholas Economides.  I am a
21    professor at New York University called Stern
22    School of Business.  And I didn't change my name
23    after I became an economist.  A lot of my MBA
24    students think I did.  They said well, you first
25    decided to become an economist, then you changed

                                                  50
 1    your name to Economides.  I didn't do that, I can
 2    assure you.
 3              What I want to talk about today is
 4    start putting Internet telephony into the
 5    perspective, how does it fit with other kinds of
 6    telephony and what exactly does it contribute and
 7    is it worthwhile in some sense.  Since the old
 8    times of the emergence of MCI as a competitor to
 9    AT&T, the question has always arisen, is this
10    something viable and something that should
11    survive, or is it something that is an
12    unnecessary bypass and something that is set up
13    there because of the peculiarities of regulation
14    and pricing and so on.  So it's very important to
15    understand if this is going to be viable or not.
16              One thing I want you to write down from
17    this whole talk is this bottom line, the server
18    on economics of networks that I have set up.
19    It's HTTP:\\RAVEN.STERN.NYU.EDU/NETWORKS and it
20    talks about telecommunication networks and other
21    networks and I think it will be very useful.
22              So I want to start by discussing
23    fundamental features of networks and in
24    particular communication networks, and go on from
25    there to see how net telephony fits in the other

                                                  51
 1    kinds of telephony.
 2              So there are three fundamental features
 3    of telecommunications networks and let me go
 4    through them one by one.
 5              First, in any networks there is
 6    complementarity among the various parts of the
 7    network.  Think of this very simple star
 8    network:  You have a switch in the middle, A, B,
 9    C and customers and so on.  To make a call from A
10    to B, you need to go from A to S and then from S
11    to B.  These are complementary parts of the
12    network.  And if you think of it in some
13    abstraction, to be able to make any phone call,
14    you need many pieces, many pieces, many links,
15    many components going in to together to be able
16    to make that.
17              Okay, second thing, externalities.
18    It's very important that the network is large to
19    be able to be valuable.  A large network, a large
20    telecommunication network is much more valuable
21    to any user than a very small one.  If I had a
22    network in one building only, it's useful.  If it
23    doesn't go outside, it's not very useful.  If I
24    had a network that just had a few blocks of
25    Greenwich village in New York, it's useful but

                                                  52
 1    most people want to call outside, as well.  So
 2    this is very important in the context of Internet
 3    telephony and how it interconnects with the rest
 4    of the networks.
 5              So it's important to understand that
 6    large networks are very useful.  Third feature of
 7    networks is critical mass.  I won't go in very
 8    deeply into it, but because of network
 9    externalities, small networks don't usually
10    survive.  Because of the big value that you get
11    from large networks, small networks don't
12    survive.  You need big, significant critical mass
13    to be able to survive.  Next slide, please.
14              So where does this get us?  Let me show
15    you the next slide, the traditional networks and
16    how does the Internet change the world.  Okay,
17    think of a customer at A and a customer at B, and
18    they want to talk to each other, okay.  Usually
19    they talk in the past through a telecommunication
20    network.  So you had a telephone, a local part, a
21    long distance part, the local was the blue, the
22    long distance was the green and the local on the
23    other side was the red.  And the solid lines
24    showed the ordinary telephone lines.
25              Okay, when Internet telephony became

                                                  53
 1    available, it became available through
 2    computers.  So the dashed lines -- so a computer
 3    to computer link from A to B.  So a regular
 4    alternative.  Now, in the back of your minds you
 5    should think that this dashed lines also go
 6    somewhere on top of telephone lines, but at least
 7    the communication was all through computers,
 8    okay.  That was the first step.
 9              Then we got to the stage in which the
10    computer of A could go through the Internet, but
11    when it reached some switch out here, D, it got
12    thrown to the regular telephone network to reach
13    B and everybody else.
14              Why does this make a difference?
15    Because there are not so many people at B who
16    have a computer here and there are many more
17    people who have telephones, okay.  So it makes --
18    it makes a big difference if a phone, wherever --
19    a phone call, no matter how it comes, gets
20    given -- gets distributed to many customers on
21    this side.
22              Then the next step was to start bidding
23    on phone calls, or the telephone part, the solid
24    part up here, then switch them to the Internet,
25    and then throw them back again to the public

                                                  54
 1    switch network, and that's where we are now.
 2              So you see, through these two critical
 3    steps, suddenly, Internet telephony matured.
 4    Why?  Because it managed to use the externalities
 5    of the public switch network.  So now you can
 6    pick up the phone, reach the other side, and your
 7    phone call could go either through the solid
 8    lines or through the dashed lines.  And you might
 9    not know it as a consumer, except for a variety
10    of quality features that might exist up there.
11    But you don't know.  I mean, it's just in the
12    background that this happens.
13              Okay, so once we got to this stage, the
14    crucial question became an issue of quality, has
15    become an issue of quality, reliability and
16    money.  That is, is this application of the same
17    quality or comparable quality as traditional
18    telephony?  Does it have the same reliability?
19    You get to talk to the guy you want to talk to?
20    Do other people -- is it sure that other people
21    don't listen to it, the conversation?  And does
22    it -- can you hear it well, okay, and so on?  And
23    does it cost less?  Does it -- is it more
24    efficient, inherently more efficient?  And Dan
25    and Larry have made already the point that it's

                                                  55
 1    inherently more efficient, cheaper and so on, so
 2    on and so on.
 3              But ultimately, because it runs,
 4    because the dashed lines run on the underlying
 5    telecommunications lines, the regular
 6    telecommunications lines, whatever cost is going
 7    to be there has to rely to some extent, has to be
 8    built up on the cost of the ordinary telephone
 9    network.
10              Okay, so let me try to go now through
11    some issues that arise.  Okay, when we look at
12    the traditional telecommunications networks, you
13    see first many services are provided, often the
14    same component of -- or services used in
15    combination with many other services.  I already
16    said this before.
17              The third I didn't talk very much
18    about, there is price discrimination.  There are
19    all kinds of price discrimination schemes in
20    selling telecommunications services.  You can get
21    your ten cents a minute or you can get a
22    25 percent discount or this or that, okay.
23              Because of network externalities and
24    critical mass that I mentioned before, it's
25    crucial that we have network interconnection,

                                                  56
 1    that even if I am in New York and my service
 2    provider is NYNEX, NYNEX is interconnected with
 3    AT&T and AT&T's interconnected with Pacific
 4    telephone so I can make a phone call from New
 5    York to San Francisco.  Okay, you need a
 6    conformity of standards.  So that's taken as a
 7    given in a telecommunications network but it's
 8    really an achievement of AT&T deploying standards
 9    before divestiture and after divestiture that
10    have been used by everyone.
11              Third, you need somehow the people
12    who -- you need to make sure that nobody in this
13    chain of the three red, green and blue pieces of
14    the network that I was describing before doesn't
15    charge an outrageous price, because even if NYNEX
16    and AT&T were charging a reasonable price, if on
17    the last end of the call Pacific was charging an
18    outrageous price, I wouldn't be able to make a
19    call, you see, because everyone has to collect
20    something and I have to pay the whole thing one
21    way or the other.
22              Okay, next slide, please.
23              So it's important to have those
24    objectives in mind.  But on the other hand, if
25    you have entrenched competitors and in this

                                                  57
 1    industry there are -- this is a grown-up industry
 2    so there are entrenched competitors, some of them
 3    seeing a new application coming might be, one,
 4    against the connection, two, against common
 5    standards, in support of high termination prices,
 6    and an more devious way of thinking, in support
 7    of raising rivals' costs, finding ways to make
 8    the life of opponents more difficult.  And these
 9    are reasonable things to expect under some
10    circumstances if there are significant entrenched
11    interests.
12              Now, in various markets, the diverse
13    interests get resolved in various ways in
14    different markets.  In telecommunications, things
15    are particularly complicated because
16    telecommunications is a regulated industry.  So
17    it's complicated.
18              Now, regulation, in particular the very
19    positive reform of the Telecommunications Act of
20    1996, imposes some very reasonable and useful
21    rules on the way the Net -- the telecommunication
22    network is run.  It imposes mandatory
23    interconnection among the various networks.  It
24    imposes compatibility standards.  It imposes
25    social objectives, such as reaching everyone,

                                                  58
 1    universal service objectives.  And it imposes
 2    specific pricing rules, so in particular in the
 3    example I was saying before, Pacific cannot
 4    charge an outrageous price in a phone call that
 5    comes from New York to San Francisco and I'll be
 6    able to complete the call at a reasonable price.
 7              Okay, now, price discrimination is very
 8    common, as I was saying a minute ago, and for
 9    reasons that I won't go into detail right now,
10    regulation may encourage or favor price
11    discrimination of various sorts.  On the other
12    hand, in markets that are unregulated and I think
13    the telecommunications market is fast going in
14    that direction, competition typically eliminates
15    price discrimination through arbitrage.  That is,
16    if the same thing was sold at different prices
17    with different people, arbitrageers or other
18    mechanisms would arise which would eliminate
19    these differences through recontracting or
20    trading between people.
21              So given these two features, one has to
22    ask this important question:  Is Internet
23    telephony an artifact of regulation?  Is it just
24    because the telecommunications network is
25    regulated presently, has not been totally

                                                  59
 1    deregulated yet and probably will not be in the
 2    next five years, is it because of this that
 3    Internet telephony arises?  Is it because of this
 4    that it's cheaper to make an Internet telephony
 5    phone call rather than a regular phone call?
 6    That's an interesting question and I won't be
 7    able to give you a very long explanation why, but
 8    I believe that it's not, because of -- it's not
 9    an artifact of regulation.
10              And the reason behind it, the reason
11    why I say no, is because of the answer to the
12    next question.  And the next question is:  Would
13    Internet telephony really manage to survive and
14    prosper under competitive conditions?  Suppose
15    that we didn't have so much regulation.  Suppose
16    we get out of regulation over time.  Would
17    Internet telephony manage to survive?  I think it
18    will, and because of that, I think that it's not
19    an artifact of the present regime.
20              On the other hand, for Internet
21    telephony to survive, presently, it needs
22    regulation.  It needs these things that I was
23    listing before, for example, low termination
24    prices and all kinds of other stuff.  And it
25    needs to some extent the compatibility.

                                                  60
 1              Now, what are the problems that
 2    Internet telephony has or, rather, had?  One,
 3    they were mentioned before.  Larry mentioned the
 4    multiplicity of technical standards.  It's very
 5    important for the network that is interconnected,
 6    although many different companies are providing
 7    some components of the network, it's important
 8    that we somehow have a common standard so that
 9    the phone call goes through from New York to San
10    Francisco okay.  And it's important that if I am
11    a subscriber, let's say to VocalTec, and there is
12    this other company, ABC which my cousin
13    subscribes to in San Francisco, I'm able to make
14    this phone call go through, even though the
15    components are completely different.
16              So if we have multiple technical
17    standards, that might not work.  This might not
18    look so conceivable given the nice world that
19    AT&T built before, we have common standards and
20    after the breakup, the standards got preserved
21    and so on.  But this can really happen here.  And
22    this can create the lack of critical mass.
23              And there's another problem that the
24    Internet telephony used to have but doesn't have
25    anymore, which was the lack of termination to the

                                                  61
 1    public switched network.  Now, since you can
 2    terminate the calls to every phone, that problem
 3    of critical mass has been resolved.
 4              The problem of quality still remains.
 5    The quality of Internet calls is not as good as
 6    the quality of regular phone calls.  I think that
 7    that's an issue that can be fixed over time.  So
 8    I don't see any inherent reason why the quality
 9    of these calls has to be lower than the quality
10    of traditional calls.  It's just a matter of
11    getting more bandwidth, improving the technology
12    and so on.
13              I guess I already gave the answers to
14    these questions in the previous slide.
15    Multiplicity of compatibility standards is a
16    problem.  Oh, yeah, I didn't say about this.
17    This is important.  Some Internet service
18    providers, ISPs, want to run a network on top of
19    the telecommunications network with different
20    standards than the telephone companies.
21              So for the first time since the breakup
22    of AT&T, a serious conflict on standards has
23    arisen, and I don't know what the outcome of this
24    is going to be, but some people -- some ISPs
25    claim that the local telephone companies don't

                                                  62
 1    run the system efficiently and they want to run
 2    it differently.
 3              (Telephone interruption.)
 4              MR. ECONOMIDES:  Now, we have to ask if
 5    this is an Internet call or a regular one.
 6              Okay, I already said no more critical
 7    mass problem because of interconnection with the
 8    PSTN.  Many levels of quality available.  I think
 9    that quality can improve.  But still, it's lower
10    than traditional phone calls.
11              And the whole thing boils to the
12    conclusion that at this point in time, Internet
13    telephony appears as a legitimate competitor to
14    traditional telephony.  It might not be
15    equally -- of equal quality.  But all cars are
16    not Mercedes, so it's okay.  People buy other
17    cars, so it's a legitimate competitor.
18              So I see this -- I wanted to close with
19    some ideas of how should regulators see or try to
20    deal with Internet telephony.  And regulators
21    thus far have not really gone that deeply into
22    regulating their network, Internet telephony.
23    Okay, so the first thing I want to do, I want to
24    make sure, is that price discrimination is not
25    imposed artificially.  Second, it's very

                                                  63
 1    important, since there are still monopolies at
 2    the local level in the ordinary provision of
 3    ordinary telephony, it's important to limit the
 4    power of those monopolies with respect to
 5    provision of Internet telephony.  Like this power
 6    is limited by the Telecommunications Act of 1996
 7    with respect to provision of services of regular
 8    telephony.
 9              And three, I think that Internet
10    telephony is by now a legitimate competitor, so
11    it should shoulder responsibilities, like paying
12    for universal service, like any other ordinary
13    telecommunications company does.  But given the
14    fact that it hasn't really matured yet, I think
15    it's not unreasonable to give them a year or two
16    before imposing universal service requirements.
17    But at some point in time, they have to pay, too,
18    like everybody else.
19              Thank you very much.
20              MR. LEE:  Thank you, Nicholas.  I think
21    we have about 10 minutes for questions from our
22    audience.  I'd like to invite our audience
23    members from the Press Club to step to either of
24    two mikes that we have sitting in front of the
25    columns if you have questions, and also to invite

                                                  64
 1    our audience who are joining us via the Internet
 2    to E-mail questions to us.  I think our E-mail
 3    address, you can find on our Forum Web page.
 4              AUDIENCE MEMBER:  Good morning.  John
 5    from GTE.  In voice transmission over packets,
 6    what kind of latency are you getting, for
 7    example, in the satellite call it's like 270
 8    milliseconds.  What kind of latency do you get
 9    now and what do you see in say five years in
10    terms of reducing latency?
11              MR. BERNINGER:  I'll take that.
12    Actually, VocalTec's not responsible for
13    latency.  It's not our fault.  Actually, so where
14    does latency come from, okay, for example?  Does
15    it happen in the phone?  Does it happen on local
16    network?  Does it happen in the servers, that
17    number two item, the edge server?
18              Basically right now latency comes --
19    the reason -- just in case nobody knows, the
20    reason the question was asked is that sort of the
21    dominant quality impairment that exists for
22    Internet telephony today is a latency one.
23              So I don't know if you've ever made
24    voice calls a number of years ago, international
25    calls.  They would go over a couple of satellite

                                                  65
 1    hops and it's a little harder to talk when
 2    there's a -- there's a long latency there, and
 3    latency being delay.
 4              And the way it works is that if it
 5    takes about 250 milliseconds to say a word, if
 6    that latency is longer than 250 milliseconds,
 7    then all of a sudden you're uttering a couple
 8    words before the other guy -- person hears it on
 9    the other end and then you can sort of bump into
10    each other in the conversation and ultimately you
11    end up just saying, you know, over, are you
12    there, or -- okay.
13              So as it stands today, we see in and
14    the implementation, 50 percent of the latency is
15    showing up in the network itself.  I should
16    clarify, for the most part people who are buying
17    our products do not buy them on the Internet
18    itself, in the sense if I'm a service provider
19    and I want to provide a certain quality of
20    service, I'm going to want some control over the
21    data network I'm using.
22              So the people that we sell to are
23    putting them on private data networks that run on
24    latency anywhere from about 60 milliseconds to
25    150.  And on the high end, if I'm going from like

                                                  66
 1    New York to Hong Kong, just for the network
 2    itself, about 250 milliseconds.
 3              Now, so you have as a telecommunications
 4    industry, you essentially have a range of places
 5    to go for networks that will drive what your
 6    latency is.  If you actually want to put this on
 7    the global Internet and you're on a single
 8    service provider, maybe you can get decent
 9    latency there.  But if you're going to make a
10    call across multiple service providers, you're
11    going to end up with multiple seconds.
12              So one answer is the Internet and the
13    data networks, the latency performance is
14    improving.  Today it's anywhere from 60 to 500
15    milliseconds and it's improving like about 30
16    percent every six months in terms of price
17    performance.
18              Now, on the server side, there's
19    actually very -- some fundamental limitations.
20    Very little that needs to be done to decrease
21    latency.  As things stand today, we've taken
22    products that were designed for other things and
23    turned them into Internet telephony.  So we've
24    taken products for voice mail where half a
25    millisecond to get my voice mail, you really have

                                                  67
 1    to be intense, if that matters.
 2              So we've had to work to get rid of
 3    those buffers and reduce the delays for these
 4    things that were designed for other purposes.
 5    I guess Larry didn't mention it, but Dialogic has
 6    some dedicated products coming out that were
 7    designed for Internet telephony and these things
 8    have latencies around 30 milliseconds for the
 9    contribution of the server itself.
10              So today, latency in the servers are
11    around 220 milliseconds, so you combine that with
12    whatever your network latency and that's going to
13    give you your end-to-end latency.  But we don't
14    see any sort of fundamental limitations to drop
15    that way down in short order.  Sorry for the long
16    answer.
17              AUDIENCE MEMBER:  I'm a commissioner on
18    the D.C. Public Service Commission here in town,
19    Ed Meyers.  I wonder if you could address the
20    question of complexity versus simplicity from the
21    standpoint of the residential and business
22    consumers over the next few years, somebody.
23              MR. FROMM:  Complexity versus
24    simplicity.
25              AUDIENCE MEMBER:  In using these

                                                  68
 1    variety of services that you have been describing
 2    for IP.
 3              MR. FROMM:  I'll take a first shot at
 4    it.  I think one of the first things that
 5    competition allows is really a range of choices.
 6    I think there are some services that will be very
 7    simple.  It will be the same as using your
 8    standard telephone.  It will be the same as you
 9    use today with the consequent cost savings that
10    are available that way.
11              I think there will be other options
12    where you can have complex applications, where
13    you can share Power Point presentations while you
14    talk over the phone line, where you can
15    configure, increasingly complex applications to
16    do a wide variety of powerful communication
17    means.  I think it really ranges, and both ends
18    of the spectrum are available and different
19    service providers, different equipment vendors,
20    different application vendors will be offering
21    different aspects along that range.
22              AUDIENCE MEMBER:  I guess my real
23    question is, from the standpoint of the urban
24    consumer, let's say low to moderate income
25    neighborhoods, will they be increasingly shut out

                                                  69
 1    over time or what?
 2              MR. BERNINGER:  Actually, in terms of
 3    simplicity, one of the things I haven't
 4    personally figured out is the pricing schemes of
 5    the existing providers.  And so I'm actually
 6    probably paying a lot more for my service than I
 7    need to just because I don't have the time to
 8    figure out all the pricing schemes.
 9              So on the one hand if your question is
10    complexity for pricing schemes, there's going to
11    be a range of what is out there.  As far as
12    simplicity -- complexity keeping people out of
13    the market, I think one of the nice things that
14    will happen is that today, telecommunications is
15    kind of like one flavor fits all, and everybody
16    sort of pays that sort of same range of prices.
17              But what Internet telephony will allow
18    is a range of performance that -- and I'll pay,
19    you know, -- it'll simply get a lot less
20    expensive at the bottom range, the most simple
21    service, and then more expensive at the top
22    range.  So the net effect is that you'll get more
23    people into the telecommunications market and
24    just communicating because they can access sort
25    of the lower-rung services.  And then on the top

                                                  70
 1    end, people will be spending a lot more because
 2    they're sort of the high-end option.  So it as
 3    not one size fits all.  So I think it brings more
 4    people on.
 5              MR. ECONOMIDES:  If I can answer a bit
 6    of that question as well.  I think that Internet
 7    telephony at present allows for lower prices
 8    because of lower quality, because of the quality
 9    of the phone call is a bit lower, of lower level
10    of quality.  But ultimately, to answer the
11    essence of your question, to be able to, for
12    these prices to remain low or to even be lower
13    than traditional high-quality long distance
14    service, it's crucial that the -- that the role
15    that local regulators play in the states stay the
16    same, to make sure that the pricing of unbundled
17    network elements and of termination remains at
18    cost.  That's the crucial element that will keep
19    these prices and cost low.
20              MR. LEE:  The gentleman to my right.
21              AUDIENCE MEMBER:  Steve from the
22    Organization for the Promotion and Advancement of
23    Small Telecommunications Companies.  I understand
24    that VocalTec, in addition to working with
25    Dialogic, has recently entered into several

                                                  71
 1    agreements with AT&T, Motorola and another firm
 2    whose name escapes me in Germany.
 3              I was wondering what sort of products
 4    and services you see coming through these
 5    alliances and what sort of effects they'll have
 6    on the development of IP telephony?
 7              MR. BERNINGER:  So the question is how
 8    it affects small telecom providers or --
 9              AUDIENCE MEMBER:  I'm mostly interested
10    in how it affects the development of the
11    technology in bringing it to market as a whole,
12    especially Motorola being a hardware developer
13    but AT&T being still the biggest kid on the
14    block, what sort of implications that has for the
15    development of IP.
16              MR. BERNINGER:  Okay.  Well, just for
17    clarity, kind of the stage we are now in the
18    Internet telephony market, is we're sort of in a
19    market education mode.  And from VocalTec's
20    perspective, we need everybody in this game to do
21    the market education.  We need -- and once we get
22    to the market education part and everybody knows
23    what it is, like everybody knows what Internet
24    telephony is, then, you know, we can actually
25    make some progress and we'll have sufficient

                                                  72
 1    economies of scale, et cetera.
 2              So you know, as far as the alliances
 3    that VocalTec is putting up, working on today,
 4    it's in that mode of building the market.  And so
 5    I don't know if there's specific outcomes of
 6    that, but just to the extent that everybody
 7    benefits.
 8              Now, the one thing that I will say is
 9    that in existing telecom today, the big three or
10    four have 90 percent of the market, and the other
11    3,000 companies work in that 10 percent, which is
12    not so small, maybe 10 percent is $10 billion.
13              But the expectation is in Internet
14    telephony, the big guys won't have as firm a
15    grip.  But let's say they get 60 percent of the
16    market and the little guys work with the 40
17    percent.  So I mean, we can't change the
18    economics of big versus small, but I think
19    overall it improves the opportunities for
20    everybody.
21              MR. LEE:  I think we have time for a
22    couple more questions.
23              AUDIENCE MEMBER:  Yes, I'd like to
24    address this to Mr. Economides.
25              MR. LEE:  Could you identify yourself?

                                                  73
 1              AUDIENCE MEMBER:  Star 67.
 2              I'd like to just comment on a notice
 3    sent out by Bell Atlantic to all customers.  Dear
 4    customer.  What it does is essentially say that
 5    under the current optional wire maintenance
 6    agreement, they won't -- didn't include voice
 7    calls and data calls, that if you have a problem
 8    now with a data call, they reserve the option to
 9    apply an additional service charge.
10              Specifically it reads: ...description
11    has been revised to indicate that service charges
12    may apply when a repair person is dispatched and
13    a problem is with the transmission or receipt of
14    data or signals, which is beyond the operating
15    capabilities of the dial tone line.
16              I'd just like your comment on a large
17    corporation doing that.  And what it effect it
18    has on this industry.
19              MR. ECONOMIDES:  I'm a NYNEX customer,
20    I guess Bell Atlantic now.  I haven't received
21    this, so I'm taking your word for it.
22              It sounds to me like companies who have
23    responsibility for dial tone and some data
24    transmission, but I would -- I would be hard
25    pressed to define the extent of responsibility

                                                  74
 1    that anybody should have for the network.
 2              I think that I better pass on this
 3    question and then let regulators talk more
 4    specifically about the standards, let regulators
 5    define the standards and responsibilities of
 6    telcos as far as data transmissions.
 7              I'm not a technical expert so I won't
 8    be able to help you too much about the specifics
 9    of the technical standards that should be there.
10              MR. FROMM:  Just as a consumer and
11    observer, I guess I'll add my thoughts, which is
12    that the Internet's a good thing.  Internet
13    telephony is a good thing.  It increases the
14    range of communication, increases competition,
15    increases the range of options people have,
16    lowest cost, and I would say that if a regulated
17    monopoly degrades or inhibits access to that
18    communication means, that I would hope the
19    regulators would look at allowing competition in
20    to allow someone else to fill that void.
21              MR. ECONOMIDES:  I would totally agree
22    with that.  Don't misunderstand my previous
23    answer.
24              MR. BERNINGER:  The other thing to be a
25    little careful about, VocalTec, these guys are in

                                                  75
 1    every day of the week, all kinds of large, small,
 2    telephone companies.  They're all interested in
 3    doing Internet telephony.  Absolutely it's crazy
 4    to run data over that network.  And I can
 5    understand their frustration.
 6              On the other hand, in theory, the
 7    company's supposed to do what the customers
 8    want.  If things were stable for 125 years, I'd
 9    quit and say okay, that was good, but now I'm
10    going to have to maybe change something.
11              But it's going to go both sides.  The
12    large telcos as far as we can tell from where
13    we're sitting are in fact implementing solutions,
14    they're taking data off of their local loop and
15    sending it right directly to ISPs, so it cuts off
16    half of it.  And to the extent there's not
17    monopolies and there isn't competition, the
18    market will figure it out.  If Bell Atlantic or
19    whoever sets up these obstacles, it's going to
20    accelerate people working around them.
21              AUDIENCE MEMBER:  My name is Mark.  I
22    work at NTIA.  This is for Nicholas.  You said
23    earlier -- I think we met before.  You said
24    earlier that you felt that Internet telephony is
25    not a product of regulatory forbearance or

                                                  76
 1    something along those lines.  Can you expand on
 2    that, particularly since you later made the
 3    observation that you felt that the universal
 4    service obligations should not be levied upon
 5    these sorts of service providers in the immediate
 6    future.
 7              MR. ECONOMIDES:  Yes.  Well, my general
 8    statement was that one should look at Internet
 9    telephony and see if it would survive under a
10    general competitive conditions.  Of course given
11    that still in the telecommunications networks we
12    have bottlenecks controlled by local exchange
13    companies, even local markets, but would it
14    survive given these circumstances, and my answer
15    is yes.  Because of that, I claim that it doesn't
16    look like Internet telephony is a product of just
17    regulation.
18              AUDIENCE MEMBER:  The --
19              MR. ECONOMIDES:  Just one second, let
20    me finish.  Now, it's important in my opinion for
21    every part of the telecommunications industry to
22    shoulder responsibilities for universal service,
23    and that's what I said.  The only thing I -- that
24    slide said was let them -- let them become real
25    telephone companies and then impose the charges

                                                  77
 1    on them.  Maybe you shouldn't impose them before
 2    they actually become real telephone companies.
 3    So that's the essence of what I was saying.
 4              And presently, there is an exemption
 5    for ISPs, but typically ISPs don't do Internet
 6    telephony only.  They do all kinds of other
 7    stuff, I can get E-mail and all kinds of other
 8    reasons and that's why they get the exemption.
 9    If ISPs become essentially telephone companies,
10    then they should really bear directly the
11    responsibility for universal service.  That's my
12    feeling.
13              MR. BERNINGER:  If I can just amplify
14    that.  Let's be careful now.  The goal is
15    universal service and I maintain that the
16    obstacle to universal service is price, okay.
17    And so -- and that's the case.  We're going to
18    move some money around and make certain things
19    cheaper than in theory and in cost and
20    otherwise.
21              My point is that what Internet
22    telephony brings to bear on that issue is a price
23    performance that's going to increase or decrease
24    the cost by half every 18 months, okay.  And I
25    think one of the things we need to do is sort of

                                                  78
 1    recalibrate our view of price.  We think of price
 2    in terms of per minute.  But as I said, I mean --
 3    so it's a penny per minute is very little money
 4    and 26 cents a minute is a lot of money.  But in
 5    fact, you know, a penny per minute is perhaps a
 6    ton of money.  Because who uses this
 7    telecommunications communication for one minute a
 8    month?  The question is what is your monthly
 9    bill?  And if I am someone that doesn't have a
10    ton of money and I want telecommunications
11    service, what is that sort of base monthly bill?
12              And what Internet telephony offers is
13    that price -- the price of that monthly bill will
14    come down by half every 18 months, and that's
15    what's going to get you universal service, not by
16    setting some rules that allow, you know, shifting
17    things around.
18              MR. LEE:  Let me suggest that's
19    probably food for thought for our next panel and
20    perhaps later in the day.  Could we take one last
21    question?
22              AUDIENCE MEMBER:  Thank you, my name is
23    Paul.  I'm with Teleport Communications Group.
24    I'm just wondering are we wasting time with voice
25    over Internet until we get to voice over ATM?

                                                  79
 1    Just want to get your comments on that issue, the
 2    ATM versus IP controversy.
 3              MR. FROMM:  I'll take it first off.
 4    That's a good question.  In my view, I think IP
 5    is the foundation for at least the next several
 6    years.  It is ubiquitous.  It's everywhere, at
 7    our homes, offices, worldwide.  What we will see
 8    as ATM as underlying backbone at various parts of
 9    this network.  And in talking to a number of
10    service providers, we are seeing them build out
11    large ATM networks and putting IP on top of them
12    because they all interconnect.  Once ATM is
13    everywhere, they we might go right to voice over
14    ATM, but that's not in the foreseeable horizon
15    that I see.
16              MR. LEE:  I think we're going to wrap
17    this panel up.  Could we, before I turn this
18    meeting back to Kathy Brown, could we show our
19    appreciation to the panelists and thank them.
20              (Applause.)
21              MS. BROWN:  It sounds like we have much
22    grist for the mill for the next panel.  We're
23    going to take about a 15 minute break.  VON has
24    graciously provided refreshments in the demo
25    area.  We'll be back here in 15 minutes.  Thank

                                                  80
 1    you.
 2              (Recess.)
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

                                                  81
 1                   POLICY PERSPECTIVES
 2              MS. BROWN:  Having listened this
 3    morning to what I think were very excellent
 4    presentations on what this technology's about,
 5    and some of the issues about how it fits into the
 6    current and evolving market structures, seems
 7    appropriate, particularly with the kinds of
 8    questions that were being asked and tossed about,
 9    that we now move on to our Policy Panel.
10              And we're privileged today to have a
11    group of people who both inside and outside
12    regulation, but certainly in the policy sphere,
13    are giving thought to these new technologies,
14    Internet telephony being one of them.  The panel
15    today brings together folks from the Federal
16    environment, from the state environment, and from
17    the academic environment to sort of tee up the
18    issues and to engage with you in a conversation
19    about what policy implications these -- this new
20    technology, this breakthrough, this experiment
21    presents.
22              So let me just briefly introduce who
23    the panel members are.  We will again have some
24    presentations from the panel and they we will try
25    to leave a good 25 to 30 minutes at the end for

                                                  82
 1    us to be able to talk together.  We have Michael
 2    Spencer here somewhere, who is keeping track of
 3    the kinds of issues that are raised so that this
 4    afternoon we might highlight them a little more
 5    and think about where or what we want to do
 6    beyond today.  There's Michael.  And Michael's
 7    keeping notes to me as sort of the issues that
 8    are coming up during the day.
 9              Diane Cornell is the chief of the
10    Telecommunications Commission of the
11    International Bureau.  You know internationally,
12    this issue is starting to surface in any number
13    of ways.  The OECD is looking at it, the OEU is
14    looking at it.  There are overarching
15    telecommunication sort of market access issues,
16    of course, that the Commission is looking at
17    carefully with WTO implementation and I think we
18    are very pleased to have Diane here with us today
19    to talk about those issues.
20              Next to her is Carl Johnson from the
21    New York State Public Service Commission.  In my
22    days at the New York State service -- Public
23    Service Commission, Carl -- Carl is the man
24    behind the curtain.  He's the fellow who has
25    thought about and written much of New York's

                                                  83
 1    competitive, pro-competitive telecommunications
 2    policy.  We worked together on New York's broad
 3    paper on interconnection.  He is very schooled in
 4    these issues and has given some real thought
 5    about this technology as a competitive entrant.
 6              Next to him is Tom Kalil.  Tom is from
 7    the National Economic Council.  He has been a
 8    very visible and very helpful point person for us
 9    in terms of Internet issues from the White
10    House.  He's been very much involved in next
11    generation in net issues.  This whole application
12    is just yet another intriguing facet of the kinds
13    of things that folks are thinking about.
14              And next to him is William Lehr, who is
15    a consultant to the MIT Telephony Consortium.
16    He's an associate faculty member at Columbia
17    University's Graduate School of Business and a
18    research associate at CITI.  He will help us when
19    the policy folks have -- the government-type
20    policy folks have talked, to think about this
21    from an academic policy point of view.
22              So with that brief introduction,
23    Diane.
24              MS. CORNELL:  Kathy's a little taller
25    than I am.  Thank you, Kathy.  Good morning,

                                                  84
 1    everybody.
 2              Well, it's normal for we government
 3    types to start out with a disclaimer that we're
 4    not speaking officially for our agencies, but
 5    this time I have to actually put that disclaimer
 6    in kind of bold face, because first of all, the
 7    FCC has not spoken officially on a lot of these
 8    Internet related issues, as I think you all know,
 9    but also because four out of five commissioners
10    are going to be new starting in, you know,
11    whatever the time frame will be in the next
12    couple of months, subject of course, to the
13    vagaries of Congress.
14              So these folks are going to have to be
15    dealing with new issues and uncharted territory.
16    So basically, you know, I am not speaking
17    officially for the FCC, but having said that, I
18    still am very pleased to be here talking,
19    focusing particularly on international issues and
20    particularly on Internet telephony, because I
21    personally feel that in the international sphere
22    is where we're going to see the greatest growth
23    potential for Internet telephony.
24              The main reason for this is because
25    obviously international telephone calls are way

                                                  85
 1    overpriced.  Here in the United States, a
 2    domestic long distance call averages about 13
 3    cents a minute.  An average international call
 4    from the United States to overseas is priced at
 5    about 88 cents a minute.  And the same story is
 6    much worse in other countries where international
 7    calling prices are dramatically higher than
 8    here.  Of course, the underlying costs do not
 9    correlate to this price difference at all.
10              The advantage of Internet telephony is
11    that it will push the price that consumers pay
12    for calls much nearer to the actual cost of
13    international calling.  And you can bet that this
14    is a lot closer to the 13 cents than it is to the
15    88 cents.  So we at the FCC enthusiastically
16    welcome the arrival of the Internet telephony on
17    the scene and hope that it will over time provide
18    some real price competition for traditional
19    high-priced international telephone service.
20              Unfortunately, it's apparent that not
21    all foreign regulators feel the same way.  But we
22    are working on bringing them around.  And we've
23    all watched Internet telephony evolve at a rather
24    astounding speed.  You've just been hearing about
25    this on the first panel, going from niche markets

                                                  86
 1    with computer-to-computer communication since '94
 2    and computer-to-phone since '96, and now to the
 3    beginnings of mass marketed services with
 4    phone-to-phone and video conferencing this year.
 5              And of course, it's not just phone
 6    service but it's fax service also, which is
 7    extremely important in the international market,
 8    particularly in Asia.
 9              In this new forum, it has -- the
10    Internet telephony has significant potential to
11    put pressure on accounting rates and on the
12    excessive profits of international telephony
13    providers, primarily because the Internet's flat
14    rate, non-distance-sensitive pricing bypasses a
15    settlement system.
16              Internationally, there remain
17    frustrating obstacles to Internet becoming a
18    substitute for the circuit switch network,
19    including things like network congestion, and
20    delays of packet delivery, poor quality and
21    reliability.  Some of this primarily, it's even
22    more true in the international context than
23    domestically.  This presents both a real
24    opportunity and a challenge for the FCC and for
25    other regulators.  I hope that we can foster

                                                  87
 1    market oriented prices that -- excuse me,
 2    policies that will encourage the global
 3    development of alternative competitive platforms
 4    like Internet telephony for the delivery of
 5    traditional communication services while at the
 6    same time, assuring the reliability and the
 7    quality of the telecom network.
 8              As Chairman Hundt observed just
 9    recently, Washington can't make the Internet
10    succeed, but it can be an obstacle to its success
11    through unwise action and unwise inaction.
12              Now, looking at the overseas regulatory
13    and market conditions front, there's a lot going
14    on in terms of unwise regulatory action.  Some
15    European countries have already banned Internet
16    telephony.  For example, the Czech Republic has
17    put in a total ban.  Hungary has said you can't
18    connect Internet to the public switch network.
19    Iceland has said only its own telecom authority
20    can provide voice services.  And Portugal has
21    said that voice services may not be permitted on
22    the Internet.  Several countries are now
23    contemplating treatment of Internet voice
24    applications under the same rules as traditional
25    long distance service and of course the active

                                                  88
 1    petition here at the -- in the United States at
 2    the FCC seeks a similar result domestically.
 3              Some ISPs have complained that foreign
 4    governments are imposing burdensome licensing
 5    restrictions as a condition of operation.  For
 6    example, the telecommunications authority of
 7    Singapore has announced that it will would
 8    require ISPs to have a basic telecom license if
 9    they wanted to provide telephone-to-telephone
10    service.
11              ISPs providing international service
12    may not be able to access the necessary elements
13    of the Internet backbone in some foreign
14    countries because of a variety of local
15    restrictions.  Some countries have restricted the
16    connection of computer devices to the public
17    network.
18              Even where competition in
19    telecommunications has been introduced, many
20    types of approval regimes have operated to slow
21    the ability of consumers to keep pace with rapid
22    technological developments.  But there are some
23    positive developments, as well, on the sort of
24    more pro-competitive side of the scale.
25              For example, the European Union has

                                                  89
 1    tentatively concluded that Internet telephony is
 2    not basic telephony.  But even there there's a
 3    cautionary note because it's a tentative report
 4    and it's still tentative proposal on voice over
 5    the Internet failed to recognize that
 6    phone-to-phone Internet telephony exists today --
 7    sort of put it in a future tense and implied that
 8    such services might be regulated like traditional
 9    telephony services.
10              And another positive development is
11    that some foreign carriers themselves have begun
12    offering or announced that they will offer
13    Internet telephony.  For example, Telecom
14    Finland, KDD and Deutsch Telecom have all started
15    Internet telephony trials, very promising
16    development.
17              There are a few company regulatory
18    policy issues that I see evolving as Internet
19    telephony emerges internationally as a viable
20    commercial offering.  First and foremost are
21    issues relating to what Internet telephony may do
22    to the cost and prices of international telephone
23    calls and what foreign regulators and telcos will
24    do in response.
25              As some of you may know, international

                                                  90
 1    telephone calls are carried under a convention
 2    known as the International Accounting Rate or
 3    settlement system.  This simply means that the
 4    originating carrier pays a terminating carrier a
 5    negotiated settlement rate to terminate the
 6    call.  It's conceptually similar to the access
 7    charges which long distance carriers pay local
 8    carriers to complete a call here in the United
 9    States.
10              The problem is that these international
11    settlement rates are absurdly high, far higher
12    than our domestic charges.  And U.S. carriers pay
13    much more of these above cost settlement rates
14    than any other carriers internationally because
15    Americans originate more calls than we terminate,
16    primarily because of price competition here in
17    the United States but also because we talk a
18    lot.
19              The fact that these settlement rates
20    are so far above cost is in large part
21    responsible for the traditional international
22    calls being so expensive.  As many of you know,
23    the FCC is taking aggressive action to try and
24    get these settlement rates down over time,
25    including an order that we just put out in August

                                                  91
 1    setting benchmark settlement rates that would be
 2    phased in over time.
 3              We have also adopted a policy which is
 4    particularly relevant and important for the
 5    Internet community, which encourages U.S.
 6    carriers to negotiate flexible, or alternative
 7    settlement arrangements, something different than
 8    the standard settlement negotiations, with their
 9    foreign correspondent carriers which either
10    bypass the settlement system altogether or cut
11    the prices of these -- of the settlement rate
12    that's used for call termination from the current
13    high levels.  We expect these measures will have
14    a very dramatic effect on international
15    settlement rates over time, particularly as the
16    implementation of last February's WTO basic
17    telecom agreement introduces competition in basic
18    telecom services starting this January.
19              But having said that, I do feel that
20    the settlement system and the familiar cartel
21    will continue to be the prevailing mode of
22    delivery of international telephony for some
23    years to come.  And in this context, the
24    advantage of Internet telephony is that it puts
25    very healthy pressure on inflated rate

                                                  92
 1    structures, especially international settlement
 2    rates because it bypasses the settlement system
 3    altogether.
 4              We at the FCC are so supportive of
 5    Internet telephony in part because we hope it
 6    will break down the artificially high cost and
 7    pricing scheme that has helped to keep
 8    international calling prices so expensive for
 9    consumers, not only here in the United States but
10    in other countries, as well.
11              But it's not clear how the underlying
12    cost structure for international Internet will
13    evolve.  For example, there may have to be
14    changes in the cost allocation and pricing
15    arrangements among the backbone and service
16    providers for international telephony traffic,
17    international Internet traffic, in fact.  We've
18    heard complaints from foreign carriers that the
19    cost of international leased private lines used
20    for Internet traffic are not being fairly
21    allocated.  They would argue that the U.S. pays
22    nothing while the foreign end picks up the entire
23    bill for the cost of the international private
24    lines.
25              And then there's the difficult issue

                                                  93
 1    which will surely arrive internationally, and I
 2    know you talked about it a bit in the domestic
 3    context already, which is how to factor
 4    international telephony into universal service
 5    obligations.  If a large part of international
 6    traffic moves off the circuit switch network
 7    and/or the volume of Internet telephone traffic
 8    grows to a certain level, should ISPs be required
 9    to contribute to universal service in each
10    country and what would that do to the service's
11    underlying costs?
12              Congestion attributable to Internet
13    traffic will also be a challenging problem
14    internationally because of the coordination
15    difficulties in ensuring adequate capacity and
16    interoperability when communicating between
17    countries.
18              ISPs and carriers will have to address
19    congestion of Internet backbones and switching
20    points.  This congestion can compromise the
21    quality of Internet telephony traffic with
22    congestion.
23              Of course, there is delay of delivery
24    of packets.  You get echo and deterioration and
25    quality and reliability, in comparison to a

                                                  94
 1    circuit-switch call.  And this quality issue can
 2    sometimes reduce the Internet telephony's
 3    acceptability as a viable competitor to circuit
 4    switch network, particularly in the telephony
 5    context, obviously.
 6              In the last year, we have seen an
 7    increasing concern with congestion of
 8    international transport, especially in
 9    transoceanic submarine cables, the transmission
10    medium of choice for the Internet.  Internet
11    usage is gobbling up rather large chunks of
12    submarine cable capacity beyond projections.
13    This has resulted in a temporary shortage of
14    Transatlantic capacity in particular, which has
15    not yet affected operations but has limited the
16    amount of capacity that's available for new
17    entrants in the international market.
18              Hopefully this will be relieved in the
19    next year by new cables coming on line and by new
20    multiplexing techniques creating more capacity
21    but it is in fact a problem we're looking at very
22    closely along with regulators on the other end of
23    the ocean for the short term.
24              As the pressure on Internet -- on the
25    Internet backbone increases and as Internet

                                                  95
 1    becomes relied upon for more serious traffic, put
 2    that serious in quotes, industry and regulators
 3    must together consider how best to protect
 4    against shutdown of the system.
 5              In light of the potential for problems
 6    in one country to spill over into others, we at
 7    the FCC and other foreign regulators should work
 8    with industry and with industry taking the lead
 9    to proactively ensure the reliability of the
10    Internet and the public switch network.  One
11    possible option would be that there should be
12    perhaps a global version of the U.S. Network
13    Reliability Council, something like a global
14    Internet, interoperability council led by the
15    private sector but with public sector involvement
16    necessary.
17              If Internet telephony becomes a
18    substitute for the provision of international
19    telephone calls, the big question would be,
20    should regulators treat voice over the Internet
21    like traditional voice or whether we should
22    change the regulation of traditional voice to
23    treat it like the way we don't treat -- don't
24    regulate the Internet today.  We and foreign
25    regulators have to decide which approach is best

                                                  96
 1    for our consumers in terms of price and choice of
 2    service.
 3              I, along with colleagues at the FCC and
 4    throughout the U.S. Government, will spend a lot
 5    of time in the next year or two talking to our
 6    foreign counterparts to try and influence their
 7    regulatory approach to Internet telephony.  In
 8    fact, I went with Ira Magaziner and the rest of
 9    the delegation, I guess about a month ago -- two
10    months ago now, on a swing through some of the
11    European capitals to do just that, and the idea
12    is to try to get ahead of the curve, to try and
13    get them thinking in terms of minimal regulation
14    before they get too far down the road towards a
15    knee-jerk approach to try and regulate Internet
16    telephony like regular telephony.
17              We here in the United States, and I
18    hope our counterparts overseas, will have to do
19    our best to develop and promote a market-oriented
20    and regulatory approach to emerging Internet
21    telephony issues and I look very much forward to
22    doing that within government but also working
23    with you folks in industry and you folks -- and I
24    should say in a variety of industries.  The
25    Internet industry, the carrier industry, to try

                                                  97
 1    and come up with the best approach to make these
 2    services actually flourish actively rather than
 3    sort of being shut down by too much regulation.
 4              So thank you very much, and I will turn
 5    the panel over to Carl, I guess.
 6              (Applause.)
 7              MR. JOHNSON:  The good news, that's the
 8    speech.  Good morning.  I'm happy to be here this
 9    morning.  Thank you, Kathy, for the kind
10    introduction.  I think you may have overstated my
11    importance here.  What Kathy didn't tell you is I
12    spent most of the last 10 years fighting the FCC
13    in comments.  But I didn't realize until this
14    morning I had so much in common with my
15    colleagues in Washington.
16              I, too, have to issue the standard
17    disclaimer, which none of you will pay any
18    attention to, but my comments are my own and they
19    do not reflect necessarily the views of the New
20    York State Department of Public Service or the
21    Public Service Commission.
22              I suppose the big question that you
23    would ask me if you had the opportunity, is do I
24    think the Internet and Internet telephony in
25    particular should be regulated.  I can tell you

                                                  98
 1    I'm not going to give you a legal answer on
 2    that.  My general counsel would have my hide if I
 3    did.  But I will try to offer a few thoughts from
 4    a policy perspective.
 5              As a state regulator, I have basically
 6    three goals in regulating any service, and in
 7    particular telecommunications.  First is high
 8    quality, second is we'd like to have prices that
 9    are competitive or services that are
10    competitively priced, and third, we'd like to
11    maximize customer choice and control.
12              Now, if I look at Internet telephony,
13    the specific application, I don't think we have
14    enough experience to judge it on the basis of
15    those three criteria.  If I look at the Internet
16    in general and all Internet applications, I think
17    with the possible exception of the service
18    quality issues that arise from time to time that
19    it seems to be doing a fine job on its own
20    without any help from regulators like me.
21              There is another issue area, obviously,
22    and it's been alluded to this morning, and that
23    is the question of regulatory parity,
24    particularly with reference to an Internet
25    telephony application as a potential competitor

                                                  99
 1    to traditional voice telephone networks,
 2    particularly in this case at least at this point
 3    the interexchange networks.
 4              From what I heard this morning and from
 5    what I knew before I came, which wasn't a lot, I
 6    don't get the impression that we've gotten to a
 7    point where regulatory parity is a critical
 8    concern.  Maybe we will someday, but it doesn't
 9    seem that way right now.
10              But even if we do get to a point where
11    Internet telephony is a significant competitor
12    for the traditional voice networks, it seems to
13    me that's exactly what we've been looking for and
14    we've been saying for years that as we get more
15    competition, we should be looking for ways to
16    reduce regulation of the incumbents rather than
17    add regulation to the new entrants, and I
18    certainly hope this is an opportunity to follow
19    through on that philosophy that we've been
20    spouting all these years, and I hope, Paul,
21    you'll appreciate that.  Thank you.
22              The other issue area that is of great
23    concern, of course, to a state regulator is the
24    regulatory treatment of the use of the public
25    switch network, and in this case, primarily the

                                                  100
 1    local exchange networks for accessing the
 2    Internet, either for Internet telephony or any
 3    other Internet application.  As many of you are
 4    probably aware, there are proceedings going on
 5    here in Washington at the FCC on those kinds of
 6    issues, and what you may not be aware, some of
 7    you are, we also have a proceeding in New York on
 8    the issues surrounding that regulatory treatment
 9    of that access.
10              As a consequence of that proceeding,
11    since we're right in the middle of it, I can't
12    give you any of my own thoughts on what the
13    answers might be to these issues.  I'm somewhat
14    restricted here.  I will, however, try to say a
15    few words about the issues and maybe suggest
16    something at the end here.
17              Basically there are a number of issues
18    raised before us and before the FCC.  One of them
19    is cost coverage.  It's the question of whether
20    or not the local exchange company and the
21    argument is principally voice by the incumbent
22    local exchange companies.  I'm not sure why it's
23    not also a concern for the new entrant local
24    exchange companies.  I have an idea but -- I'm
25    not totally naive -- but the question is, are the

                                                  101
 1    companies covering the cost that they incur to
 2    provide this access?
 3              The second issue, which you've just
 4    heard also with respect to international Internet
 5    traffic, is a congestion issue, at least
 6    allegedly there are congestion problems at
 7    various nodes in the local exchange network,
 8    occasioned by a change in the traffic pattern
 9    because we tend to use our Internet connections
10    differently than we use our traditional voice
11    connections.
12              The third area which sort of answers
13    part of the question that I posed before about
14    cost coverage and the competitive prices, is the
15    question of intercarrier compensation arising
16    from interconnection agreements that have been
17    signed either since the Act or before the Act
18    where a pair of locally competing LECs agree to
19    pay each other for terminating traffic on each
20    other's network.  This causes a problem,
21    obviously, if you're delivering more traffic to
22    your cooperating competitor than they are
23    delivering to you and that tends to happen when
24    one of the local exchange companies signs up an
25    Internet service provider, particularly is a

                                                  102
 1    large one.
 2              And related to covering perhaps all of
 3    these issues is a jurisdictional question posed
 4    as should that connection or that use of the
 5    public switch network, the local network to
 6    access the Internet be treated as a local call
 7    and be subject to state regulation, or should it
 8    be treated as -- well, as an interexchange call
 9    and at the same time, as an interstate -- or a
10    matter of interstate jurisdiction.  That's an
11    issue before the FCC right now.  And I can say
12    that we -- my department has filed comments
13    asserting that it is obviously a local call.  At
14    the risk of angering my general counsel, I'll say
15    at least it's a debatable point.
16              But I'm not really sure how important
17    the jurisdiction really is to the fundamental
18    questions because the answers to the fundamental
19    questions could be the same whether they are
20    answered by Diane's organization or by my
21    organization.
22              All of these questions really sort of
23    reduced the rate design issues and
24    increasingly -- I should say that I started my
25    career, very early on in my career with the New

                                                  103
 1    York State Department of public service, I was a
 2    tariff and rate analyst and that's what I did.
 3    It was rate design.  It was a lot easier when you
 4    were regulating a few basically monopoly
 5    telephone companies and their services, because
 6    when you design a rate, the real issue is what's
 7    that do to another customer of this company?  If
 8    I use usage rates and lower monthly subscription
 9    rates, I get one effect.  If I reverse it, I get
10    another effect.  But the effect is on the
11    customers of that company.
12              If, as we have more players in the
13    field, it becomes much more complicated because
14    now the question becomes if I change this rate,
15    not only what happens to the customers of this
16    company, but what happens to the market and the
17    other companies that are involved in this market
18    and what are their responses going to be?
19              I don't know what the answer to this
20    specific rate design issue is and I'm not even
21    sure there is a one right answer.  And I guess if
22    I had one thought I'd like to leave you with
23    today, it's that here we are in a transition,
24    theoretically, at least.  And I think actually,
25    to a more competitive telecommunications

                                                  104
 1    marketplace, it would be real nice to start
 2    seeing market-based solutions to questions like
 3    this rather than being asked as a regulator to
 4    come up with the answers to these questions.
 5              And I'm not naive enough to believe
 6    that my role in setting prices or encouraging the
 7    direction in which they are set is going to be
 8    eliminated tomorrow, but it would be nice when
 9    I'm asked to make decisions like this and make
10    recommendations to my Commission and to others
11    that if I -- it'd be nice to be given a range of
12    options, a range of different rate design
13    solutions to these issues that have been candled
14    against what would a competitive marketplace do?
15              Is applying access charges to ISPs what
16    a competitive marketplace would do?  I don't
17    know.  I'm not answering the question.  I'm not
18    suggesting I know the answer.  But I think we
19    ought to be looking at the question from that
20    standpoint, and to the extent that we have to
21    impose or continue to seek to impose regulatory
22    and judicial solutions to the issues of this
23    industry, that we do it with an eye toward what
24    would the marketplace do if it were competitive.
25    And with that, I'll thank you.

                                                  105
 1              (Applause.)
 2              MR. KALIL:  Good morning.  Like my
 3    colleagues, I have to start off with the usual
 4    disclaimer.  I'm going to indicate when I'm
 5    reflecting official administration policy and the
 6    rest of the time you should just assume that I'm
 7    making it up as I go along.
 8              I wanted to start off by telling you my
 9    favorite Internet telephony story.  I saw this on
10    the Internet, so it must be true.
11              One of the world's major telecom
12    companies with significant long distance revenues
13    recently held a conference call among their top
14    technical people to try to figure out how
15    significant a threat the Internet telephony
16    technology was to their long distance revenue,
17    and this conference call went on for about 20
18    minutes with no conclusion in sight until one of
19    their participants admitted that he had called in
20    using Internet telephony.  So I really think that
21    this is happening and it's real and it's
22    something that we have to take seriously.
23              I think that at least from my point of
24    view, as we're looking at these issues of
25    regulatory parity between Internet telephony and

                                                  106
 1    the public switch telephone network, I think what
 2    we have to ask ourselves is which is the more
 3    desirable end state?
 4              And as far as I'm concerned, there are
 5    really four important characteristics of the
 6    Internet as both an economic system and a
 7    technological system.  One is obviously the low
 8    barriers to entry and the great deal of
 9    competition that that creates.  The second is its
10    incredibly rapid growth.  The third is the little
11    or no government regulation, at least in the
12    United States.  The Internet works on rough
13    consensus and running code, so Internet service
14    providers or software companies don't have to ask
15    for a notice of proposed rulemaking or file a
16    tariff when they want to do something new.  And
17    the fourth is an architecture that really permits
18    an incredible amount of technological
19    innovation.
20              As Dave Clark at MIT has said, the
21    Internet has an hourglass infrastructure, an
22    hourglass architecture, namely that the Internet
23    protocol is the thin waist.  You can run the
24    Internet protocol over anything that you want
25    to.  It's going -- Internet packets are being

                                                  107
 1    transmitted over everything from, you know, X25
 2    circuits running in a developing country at 2.47
 3    kilobits per second to 622 megabit packets per
 4    second over Sun.  Then you can run any sort of
 5    application that you want to over IP.
 6              So in the telephony market, for
 7    example, people are now developing voice codex.
 8    They are integrating telephony into Web
 9    browsers.  They're developing multicast
10    applications.  And although there are a lot of
11    problems with today's Internet, those are things
12    that are being addressed.
13              People talked a little bit about
14    capacity.  There are now a number of start-ups
15    that are being aggressively funded by venture
16    capitalists and others to develop terabit
17    routers.  You have carriers, carriers that are
18    springing up in the United States that are
19    building out national networks with a hundred to
20    200 fiber pair and each fiber has OC 192 with
21    8 to 16 wavelengths using dense wavelength
22    division multiplexing.
23              So I think when you see those sorts of
24    investments and that sort of rate of
25    technological innovation, you have to believe

                                                  108
 1    that some of these capacity issues are going to
 2    be worked out.
 3              Another major limitation of today's
 4    Internet is that right now, you can only provide
 5    one quality of service.  That's being addressed
 6    in a number of ways.
 7              The administration is supporting
 8    something called the next generation Internet.
 9    One of the goals of that is to in a research
10    environment start experimenting with new
11    applications, like tele-emersion, which is
12    network virtual reality, shared virtual
13    environments, distributive computing, distance
14    learning, telemedicine, a whole lot of very
15    high-end applications that are going to require
16    that users have the ability to say I need so much
17    bandwidth, I need so much latency, I need a set
18    of services that goes beyond best efforts.
19              In the commercial area, you have
20    customers that are starting to get their act
21    together and say I want the following quality,
22    reliability, up time.  The best example of that
23    is an extranet that the big three are using to
24    connect their suppliers and they are developing a
25    methodology for spelling out what their

                                                  109
 1    requirements are, certifying Internet service
 2    providers that are willing to meet those
 3    requirements and then using Bell core to actually
 4    make sure that the Internet service providers
 5    live up to that.
 6              They anticipate, by the way, that
 7    they're going to have to pay more hand that for
 8    that sort of service.  They might have to pay a
 9    premium of 50 to 100 percent.  But the point is
10    that although we can identify a lot of
11    shortcomings to today's Internet, we can also see
12    solutions on the horizon.
13              So as we compare the Internet with the
14    public switch telephone network and the
15    accompanying technological, economic and
16    regulatory environments and we ask questions
17    about what is a desirable end state, I think at
18    least from my point of view, I would like to see
19    the public switch telephone network move in its
20    regulatory environment, move in this direction
21    rather than the other way around.  I think that
22    one area where PSTN certainly has a big advantage
23    right now is on issues related to reliability and
24    robustness.
25              So let me talk about some of the

                                                  110
 1    specific regulatory issues that have been
 2    raised.  One was that -- that we ought to start
 3    treating software companies as common carriers
 4    and it's official administration policy expressed
 5    in a letter by Larry Irving, our NTIA
 6    administrator that this is a bad idea and would
 7    be unworkable.  So you've got a pretty clear
 8    administration policy on that.
 9              Second issue is that -- the second
10    issue that people have raised is that the
11    Internet will collapse because there will be too
12    much voice traffic.  Again, I think that this is
13    something that market forces can solve.  The
14    Internet service providers can offer more
15    capacity or start charging different amounts for
16    differentiated services.
17              I think that there is one technical
18    issue that needs to be looked at.  A lot of the
19    voice applications use a protocol called UDP as
20    opposed to TCP, which does not back off
21    gracefully in the face of congestion.  So you may
22    have a situation in which all the TCP-based
23    applications are acting like a good citizens in
24    an Internet environment and the UDP applications
25    aren't.

                                                  111
 1              The third issue that has been raised is
 2    whether this increase in Internet traffic that
 3    might be triggered by a move to voice telephony
 4    would degrade the public switch telephone
 5    network.  I think that we don't have a whole lot
 6    of evidence for that except in a few central
 7    office switches around the country in places like
 8    Northern California, for example.  And I think
 9    that there are also mechanisms for people to be
10    able to address this.  There are already products
11    on the market, for example, that take the packet
12    traffic off of the PSTN before it hits the
13    central office switch.
14              Fourth issue, regulatory parity.  I
15    think that one of the issues that we have to
16    start thinking about is that what would happen if
17    Internet telephony became a real substitute for
18    POT service in terms of quality, reliability,
19    availability, and when given that Internet
20    telephony became a real competitor to traditional
21    POT service, when we could start viewing them as
22    a true substitute in the economic sense and when
23    that would justify the FCC using the forbearance
24    authority that they have under the '96 Act.  I
25    think that's an important question.

                                                  112
 1              A fifth issue that's been raised is
 2    whether -- this is really just sort of regulatory
 3    arbitrage and a mechanism for avoiding access
 4    charges.  I think that's another area that
 5    deserves more empirical work because I think
 6    it'll indicate one of those elements of our
 7    current regulatory structure that are going to
 8    come under a lot of pressure.
 9              I don't think that this is just about
10    regulatory arbitrage, although there may be a
11    component of that.  I think that the more
12    exciting part of this is genuinely new services
13    that you couldn't introduce before, number one,
14    and number two, some efficiency and productivity
15    gains that come along associated with statistical
16    multiplexing and those things like introducing
17    new compression algorithms.
18              The sixth issue is what does the United
19    States do about other regulatory entities, either
20    at the Internet or state level that have -- that
21    have the view that, you know, we should regulate
22    the Internet because it's starting to offer like
23    services, such as voice and broadcasting, and if
24    you look at the administration's paper on global
25    electronic commerce, we took a pretty clear view

                                                  113
 1    on this and opposed international efforts to
 2    regulate the Internet on the grounds that they
 3    were offering like services.
 4              And finally, the seventh issue is, you
 5    know, what are we going to do about universal
 6    service?  And I'm gone into let Larry handle that
 7    one.  Thank you very much.
 8              (Applause.)
 9              MR. LEHR:  We're trying the high-tech
10    version here.
11              I'm the academic, which lets me off the
12    hook with respect to my other panel members,
13    because I'm not bound to propose anything that's
14    actually practical.
15              What I'm going to talk to you about

16    today -- I'm going to talk in sort of three
17    parts.  First I'm going to talk about what is the
18    MIT Internet Telephony Consortium which is the
19    hat I'm wearing here today.  Then I want to talk
20    a little bit about what I view as is Internet
21    telephony and explain that, and then move on to
22    the bulk of my talk, which is explaining what I
23    think the policy challenges are that are composed
24    by Internet telephony.
25              The MIT Internet telephony consortium

                                                  114
 1    is a joint industry/academic research consortium
 2    that was actually formed in August of 1996 and is
 3    housed in the Center for Technology, Policy and
 4    Industrial Development at MIT.  It's a
 5    multidisciplinary research group that's focusing
 6    on a wide range of issues, technical issues,
 7    economic issues, strategic issues, policy issues
 8    that are raised by the question of convergence of
 9    telecommunications and the Internet.
10              I'm an economist.  A number of the
11    other researchers that are involved are
12    engineers, political scientists, other academics
13    and industry participants who have been
14    interested in what's happening with the
15    Internet.
16              We have a Web site.  The Web site
17    address is up there for those of you that would
18    like more information about that.
19              With respect to what Internet telephony
20    is, I think it's important to think of Internet
21    telephony as much more than POTS, or plain old
22    telephone service, over the Internet.  You should
23    think of it as an evolving platform for
24    ubiquitous multimedia services.
25              We just heard a set of features that

                                                  115
 1    characterize the Internet.  And Kalil's talk,
 2    issues he thinks are interesting and attractive.
 3    I have a slightly different list.  One thing is
 4    it is decentralized.  It's lots of different
 5    networks talking to each other.
 6              Second, it's based on open standards
 7    that right now largely aren't controlled by any
 8    individual interest and that are freely available
 9    and all of that.
10              Three, and I think this is extremely
11    important, the Internet's inherently
12    media-independents and customer-premise equipment
13    independent.  In other words, all kinds of boxes,
14    computers, telephones, fast computers, slow
15    computers, can talk to each other over the
16    Internet set of protocols.  This makes the
17    Internet inherently scaleable and accounts for
18    some of the features that Tom mentioned when he
19    talked about how it's being used internationally,
20    in environments that have very different
21    qualities of underlying infrastructures.
22              Four, it started as data network.  It's
23    inherently digital.  If it's going to be the
24    ubiquitous platform for multimedia services, it's
25    going to have to be able to handle all the

                                                  116
 1    heterogeneous data: data, voice, video traffic,
 2    and in a sense, the Internet's here to do that.
 3              And fifth, the Internet is designed
 4    based on the ideas of significant end user
 5    control.  In other words, it's less of a
 6    network-centric view.  The idea is that you have
 7    pipes connecting boxes that are intelligent at
 8    the periphery of the network and to the extent
 9    you can, you leave as much control in those boxes
10    as possible.
11              The long-term view is that today, the
12    Internet runs on public switch telephone network
13    facilities but that tomorrow, the telephony
14    services could actually run on the Internet.
15              Now, that doesn't mean that we're going
16    to get from somewhere this wholly new network.
17    What it means is that today's owners of the
18    telecommunications facilities, the regional Bell
19    operating companies, the long distance carriers,
20    they'll certainly be part of this whole picture,
21    but so, also, if this works, will be cable
22    systems, will be hopefully electric power
23    companies, hopefully will be satellite
24    providers.  All of these other ways and channels,
25    for getting electronic signals into your home,

                                                  117
 1    will be able to compete.
 2              So in a sense, one of the things that
 3    this infrastructure promises is a more open
 4    environment that will foster much greater
 5    competition across, the different ways in which
 6    we can connect households to the electronic
 7    superhighway.
 8              Probably sort of my overall view -- and
 9    I probably should have mentioned a disclaimer,
10    I'm here talking again with my independent
11    views.  These are not the views of the Internet
12    Telephony Consortium.  I think that in the near
13    term, Internet telephony's forcing the policy,
14    the debate to confront a lot of the problems with
15    telephony regulation.
16              We're in the process of a transition in
17    this country and internationally, and Internet
18    telephony is making a number of the issues that
19    are being addressed much more poignant.  Today,
20    the bulk of telephone traffic, almost all of the
21    telephone traffic, is still traditional plain,
22    old telephone service traffic.  The actual impact
23    of Internet telephony today is still quite
24    limited from the point of view of minutes
25    carried.

                                                  118
 1              But in the much longer term, I think
 2    that there's going to need to be a new paradigm
 3    for industry structure and public policy if the
 4    Internet, if it does become the infrastructure I
 5    think it could potentially become, affects both
 6    the way we work and play.
 7              The Internet faces a number of
 8    challenges.  We have talked a little bit about
 9    some of these already.  There's a number of
10    technical challenges.
11              What's interesting about Internet
12    telephony is it's about putting voice end-to-end
13    communications on the Internet.  Although that
14    may seem relatively low tech, human factor
15    design, people go out there and decide what
16    quality of service humans can really tolerate.
17    It turns out delay, sensitivity and audio, voice
18    communications is very sensitive.  So it's
19    actually a very difficult thing to do, in many
20    ways more difficult than doing video.  Video, you
21    need a lot of bandwidth but you can tolerate a
22    lot more variation in delay, et cetera.
23              There are obvious problems of
24    congestion -- we've heard something about that
25    already -- if you do have a multimedia network

                                                  119
 1    where you have very heterogeneous services coming
 2    in with very different demands in terms of what
 3    they place on the network, but then also very
 4    different values for those demands.  So you have
 5    someone that's got an electronic transaction for
 6    a bank trying to clear a billion-dollar check,
 7    there might be a very few number of bits and
 8    you've got someone else who's downloading movies
 9    from Sweden of perhaps dubious quality that takes
10    a lot of bits.  So you have to handle that.
11              There are quality of service issues.
12    We've heard about reliability.  There are
13    questions of grades of services.
14              Then there are issues of directories of
15    services.  How do you locate individuals when
16    they may be across multiple different networks
17    and you have no sort of network-centric
18    controller basically saying how to get from one
19    point to another point.
20              There are a number of economic issues.
21    There are questions of cost recovery and
22    efficient pricing.  We're talking about putting
23    in big huge pipes that are very expensive to put
24    in.  Once they're in, they're not very -- the
25    costs aren't very sensitive to how much they're

                                                  120
 1    actually used.  And so that creates a problem
 2    that's extremely interesting to economists like
 3    myself and Nick Economides, but is something that
 4    I won't spend a lot of time talking about today
 5    because, one, I don't have enough time, and two,
 6    I'm sure it would bore the pants off of most of
 7    you.  But this is a real issue that will need to
 8    be worked out, and there's research going on
 9    trying to resolve some of these things.
10              These things will feed into the sort of
11    industry structures that are likely to evolve,
12    and tracing that, there's another area of where
13    we need to do a lot more research.
14              The regulatory policy issues, we have
15    both convergence and competition; convergence, in
16    that the Internet's facilitating the convergence
17    of multiple industries:  Media industries,
18    telephone industries, et cetera, and that's
19    blurring traditional regulatory boundaries and is
20    also even raising questions about, you know, what
21    is it that should be regulated?
22              And we're living in a world today where
23    the overwhelming view is that more competition is
24    preferable to more regulation.  So there's a
25    question about how to manage that transition.

                                                  121
 1              The Internet as an engine of technical
 2    and industry convergence is driving this.  This
 3    is the sense in which I think it's sort of
 4    forcing the issue.
 5              Just a few of the policy areas that are
 6    obviously affected.  You have telephony, and
 7    Internet telephony has its most direct effect
 8    there.  You have broadcast media where you worry
 9    about issues of content, programming diversity,
10    access to the airwaves, spectrum management and
11    then a whole separate thing for cable television,
12    you know, content issues, you have issues of like
13    obscenity, pornography, et cetera.
14              You have intellectual property issues,
15    copyright issues.  Once you've got multimedia
16    networks connecting everybody, it's a lot easier
17    to copy and distribute stuff across national
18    boundaries without respecting the traditional
19    mechanisms of protecting intellectual property.
20              You have questions of privacy and free
21    speech.  As we live more of our lives in
22    electronic world, more of ourselves are exposed
23    to this outside world.  There are problems of
24    flaming, receiving mail that we don't want to get
25    or connections we don't want to get, or, you

                                                  122
 1    know, kids being able to access pornography,
 2    et cetera.
 3              And then there's traditional antitrust
 4    and trade issues that come up when the Internet
 5    and Internet telephony perhaps changes industry
 6    structures and competitions and regulatory
 7    domains.  Currently we're in a process of
 8    significant telecom regulation.
 9              The Telecommunications Act of 1996 is
10    trying to introduce local competition and with
11    local competition, eventually, hopefully,
12    deregulation.
13              It calls for reforming the access
14    charges that currently today are a major source
15    of the implicit subsidies that fund universal
16    service in telephones and are one of the areas
17    where people are using Internet telephony to try
18    an arbitrage of those uneconomic rates.
19              There's also the World Trade
20    Organization agreement to lower international
21    barriers to competition that was signed last
22    February and the orders that Diane Cornell
23    discussed about international settlements.
24              Issues during the transition, I'm going
25    to pretty much skip this slide because I think

                                                  123
 1    these issues have been dealt with.  I personally
 2    look at the question of Internet service provider
 3    access and Internet traffic and the congestion
 4    it's causing for local telephone networks as a
 5    short-term solution and a solution about which
 6    the response should not be regulation at this
 7    time.
 8              The question about what happens in the
 9    international arena is you've got even more money
10    at stake because the international accounting
11    rates are so out of line with cost and it's an
12    issue that's very similar.  Again, it's a
13    transitional issue, an issue where I think that
14    existing regulatory structures are in the process
15    of changing and it wouldn't make sense to try and
16    loop the Internet into this traditional telephony
17    architecture.
18              Longer term, I think we're going to
19    have to confront the issues of industry structure
20    and vertical integration.  That does not mean
21    that there's a clear government role, but I think
22    the government needs to be concerned and sort of
23    watch what's happening.
24              There's a major drive towards one-stop
25    shopping, because customers demand it and because

                                                  124
 1    it allows you to offer a whole lot more services
 2    and because there are large economies of scope.
 3    If I'm already selling telephone service to
 4    someone, I'm a telephone carrier and customers
 5    want Internet service, why not sell them Internet
 6    services too?  If I'm selling Internet services
 7    and customers want telephone service, why not
 8    sell them telephone service, too?  The problem is
 9    if customers don't have a choice.  One-stop
10    shopping is a bad idea if it means you can only
11    buy the services from a single provider.
12              There's also the possibility that the
13    technologies that develop, the standards that are
14    developed to make it difficult for customers to
15    actually use their choice.  In other words, there
16    could be multiple carriers out there but we have
17    to be careful of what it actually costs consumers
18    to change from one carrier to another.
19              There's also the question of what are
20    going to be the bottleneck facilities.  There's
21    been a lot of focus on the local loops as owned
22    by the local telephone carriers.  One of the
23    things I hope the Internet will do for us is make
24    it more possible to use some of the other ways
25    that you can get into a home as access lanes and

                                                  125
 1    reduce the bottleneck aspect of those local
 2    loops, facilitate the emergence of local
 3    competition.
 4              But there's other potential bottlenecks
 5    that could emerge.  If we do have all these
 6    wonderful services we're going to have to start
 7    having smart homes in the future and the way
 8    these things will roll out is going to be
 9    incremental over time.  What we're going to need
10    is an interface to the home that doesn't mean the
11    customer has to replace everything on the home
12    side, all his electronics, because some network
13    provider upstream has decided that oh, we've now
14    tweaked to IP in some way you don't want to know
15    about and you can no longer talk to the network,
16    or that we have -- everybody has to upgrade their
17    home equipment simultaneously in order to take
18    advantage of a new service.
19              There's been a lot of talk about issues
20    like directory services and domain names.  If
21    you -- if there is this world of an Internet of
22    multiple networks all interconnected, all talking
23    to each other, then the question of your
24    electronic identity and how you keep track of it
25    becomes very, very important, and what is your

                                                  126
 1    electronic identity?  There are significant
 2    property rights there and these are difficult
 3    issues to sort out.
 4              Also, the possibility of backbone
 5    interexchange points.  We need to watch for the
 6    fact that the Internet doesn't become a small
 7    club of a few people who only share traffic with
 8    each other and in effect close out of it a lot of
 9    other participants in the process.
10              Internet technology standards.  We need
11    to open interface standards but we need to marry
12    those against the delicate thing of creating,
13    dealing with the fact that people have to have
14    incentives to innovate and those incentives are
15    they create a better mouse trap and then they
16    want money for it.
17              If the Internet is to actually become
18    infrastructure, then we're going to have to deal
19    with a lot of the traditional telephony issues.
20    We're going to have to deal with questions of
21    access and universal service.  Do we have to --
22    where do we draw the line in enhanced
23    definition?  We're going to have to deal with
24    questions of interconnection and then finally, if
25    we think of the Internet as global

                                                  127
 1    infrastructure, which it is rapidly becoming,
 2    then we have to deal with the fact that what has
 3    been a very United States-centric view and way in
 4    which the Internet has evolved is rapidly
 5    changing.  And so U.S. regulatory authorities,
 6    U.S. industries are not going to be able just to
 7    steer this themselves.  Thank you very much.
 8              (Applause.)
 9              MS. BROWN:  I think you have gotten a
10    lot of information.  We have about 15 to 20
11    minutes for questions and discussion, so I invite
12    to you the microphones and we might as well just
13    start in.
14              I'll take something from the back which
15    is also our Internet connection first.
16              AUDIENCE MEMBER:  This is a question
17    from Joel, Simon Strategies.  I guess it would be
18    for you, Diane Cornell.  And could you comment on
19    the development of Internet telephony in Japan.
20    Japan, traditionally very closed to competition,
21    seems well on its way to being a key player in
22    Internet telephony business.
23              Statistics suggest that by the
24    year 2000, Internet telephony will account for
25    about 25 percent of telephone use in Japan.  Any

                                                  128
 1    comments?
 2              MS. CORNELL:  Yeah.  I would say that
 3    Japan is a -- for a developed economy, is a
 4    country in which there have been very high
 5    telephone prices and I'm not at all surprised
 6    Internet telephony has great potential to grow in
 7    Japan.  Same thing with call back, which is very
 8    popular in Japan for that very same reason.
 9              I would think that, assuming that the
10    Japanese regulatory authorities allow it to
11    prosper, and that's going to be the challenge,
12    that it could be very, very -- a very big
13    industry in Japan.
14              KDD is experimenting with Internet
15    telephony as I mentioned earlier.  KDD has also
16    extremely high settlement rates in addition to
17    high calling prices.  So either KDD is going to
18    have to change and adapt to the new environment
19    and the fact that they are experimenting with
20    Internet telephony is a very good sign, or
21    Internet telephony is going to grow extremely
22    quickly in Japan and we'll see how that flows.
23              There are very -- have been
24    traditionally very restrictive international
25    market conditions, international telephony

                                                  129
 1    conditions in particular in Japan and I look very
 2    much forward to the Internet telephony providers
 3    helping to, shall we say, break that down.
 4              MS. BROWN:  Why don't we go to my right
 5    here.
 6              AUDIENCE MEMBER:  I'm George, with --
 7    I'm a partner in a consulting company called
 8    International Protocol.  We specialize in doing
 9    infrastructure projects around the world,
10    including voice over the Internet, terminations
11    in foreign countries.
12              I'd like to follow up on that last
13    question, actually.  I was going to do this
14    slightly differently.  But one of the things we
15    do is we help our clients terminate calls in
16    foreign jurisdictions, including Japan, and a
17    number -- in a number of situations, we have had
18    our circuits physically removed or have been
19    blocked from providing those terminations.  And
20    when I heard this debate about regulation, it
21    came to mind something that Ralph Nader said not
22    too long ago.  It was a quip.  He said that while
23    regulation may be the dirty word of the 1990s,
24    deregulation may be the dirty word of the next
25    century.

                                                  130
 1              And my question, maybe, Diane you can
 2    help answer this, is without some form of
 3    regulation, what prevents the PTT in foreign
 4    jurisdictions from shutting down this industry
 5    before it even begins?
 6              MS. CORNELL:  Yeah.  That's -- I think
 7    that's why Reed Hundt said that one of the things
 8    we have to look at is the question of inaction
 9    and action.
10              And I do think that we are going to be
11    facing, particularly as we see liberalization
12    internationally during the next year or two as a
13    result of the WTO telecom -- basic telecom
14    agreement, one of the things that we are focusing
15    on is making sure that along with that
16    liberalization, you have fair rules of
17    competition.  And fair rules of competition have
18    to be targeted at exactly these kinds of abuses.
19              And a lot of regulatory agencies around
20    the world don't have any experience in this at
21    all.  They don't -- they don't know how to
22    monitor these kinds of activities and the PTT's
23    obvious -- or the incumbent telephone companies
24    obvious incentive is to do exactly the kinds of
25    activities to shut down competition that George

                                                  131
 1    is talking about.
 2              So we are going to have to spend a lot
 3    of time, have already been spending a lot of time
 4    at the FCC, the other U.S. Government agencies,
 5    talking to the new regulators overseas and trying
 6    to explain to them exactly the kinds of steps
 7    that need to be taken in order to monitor the
 8    market conditions, to make sure that they're
 9    fair.
10              So in the short run, as was the case
11    here in the United States, you're not talking
12    about just a flat deregulation.  You're talking
13    about opening to competition but introducing fair
14    rules of the state of play so that the incumbent
15    telephone company is not sort of left free to
16    just do these kinds of activities.  And we have
17    been getting a lot of complaints about exactly
18    the kind of thing that you're talking about.
19              And let me say, all of the U.S.
20    Government agencies are going to be monitoring
21    implementation of the WTO agreement, and if you
22    all are encountering difficulties out there in
23    the marketplace on the ground, please come and
24    talk to us, because we can, through jawboning and
25    through potentially more formal actions, have a

                                                  132
 1    very great influence on what's going on out
 2    there.
 3              So you know, please let us know.
 4              MS. BROWN:  Thank you.
 5              QUESTION:  My name is Tom Gibson.
 6    I'm with MCI.  I'm here for my own personal
 7    edification.  My question should not be construed
 8    in any way as reflecting MCI policy.  I'll open
 9    this up to anybody.
10              The big, simple, difficult question:
11    What signals are you all looking for in the
12    marketplace that suggest the need to begin moving
13    the two networks towards some sort of regulatory
14    parity?  And if you could quantify those signals,
15    broadly quantify those signals.
16              MR. JOHNSON:  Absolutely, I can't
17    quantify them.  I'm not sure either what
18    regulatory parity might mean at the point at
19    which you decide to go there.  Regulatory parity
20    at that point might be no regulation at all.  I
21    don't think so, but -- obviously you regulate --
22    society regulates.  It's not my idea.  This is
23    what society does.
24              When the service or product in question
25    is vitally important and when the provision of

                                                  133
 1    that or delivery of that product or service can
 2    be constrained by primarily a single entity, if
 3    you don't have that situation, you have grounds
 4    for deregulating.  That's about as best an answer
 5    as I can give you.
 6              I don't think you'll find any regulator
 7    who's going to give you a score card and tell you
 8    when you meet all of these things, when all of
 9    these things happen in the marketplace, we'll
10    stop.  And I'm sorry if you would like something
11    more precise, but I can't give it to you.
12              QUESTION:  Yes.  I'm with Sure
13    Technologies, Fairchild.  My question is for
14    Ms. Cornell.
15              You mentioned a number of European
16    countries that have already placed restrictions
17    on Internet telephony.  I want to know, are you
18    finding that their stated reasons for these
19    restrictions -- are they phrasing it in terms of
20    concerns about network congestion or about rate
21    arbitrage or how are they doing this?  And if you
22    want -- if you want to make any comments about
23    their sincerity, then feel free, as well.
24              MS. CORNELL:  Far be it from me to make
25    any comments about their sincerity.  I think in

                                                  134
 1    the short run, I think it's primarily more of a
 2    market access issue.  It's more of a question of
 3    protecting the incumbent telephone carriers is my
 4    impression.  They have not necessarily based it
 5    on network congestion or technical issues as much
 6    as protecting the incumbent telephone companies'
 7    sphere of activity.
 8              That is not just a legal protection but
 9    it is also going to be an economic issue as time
10    goes on and it gets back to this whole regulatory
11    parity question and the pricing scheme.  And it's
12    going to be very interesting to see how that
13    evolves once we get into a situation where their
14    markets are liberalized under the WTO agreement
15    and they can't put up a formal market barrier.
16              At that point, it's then going to
17    evolve into a more regulatory issue.  How do you
18    regulate this, whatever it is.  You know it has
19    to be in, but how do you regulate it?  And that's
20    when we get into the regulatory parity type
21    issues internationally.
22              QUESTION:  So before liberalization
23    they don't feel any need to cloak their
24    restrictions in any sort of technical or economic
25    argument?

                                                  135
 1              MS. CORNELL:  I don't -- no, I don't
 2    think that they've been focusing on that.  In
 3    some circumstances, they might have cited it, but
 4    it's primarily a market access or sort of a
 5    protection of the incumbent telephony.  I'm not
 6    sure they would articulate it that way,
 7    obviously, but you know, I think that's what's
 8    motivating them.
 9              MS. BROWN:  Other questions?  Go
10    ahead.
11              QUESTION:  This is a domestic question
12    actually for Carl.  And knowing that you have
13    that disclaimer before, it's actually a question
14    about the state so you might want to use it
15    again.
16              I heard you say, I think, that the
17    State of New York is asserting that Internet
18    access and Internet telephony is primarily a
19    local service -- is local in nature, and if that
20    is the case, is the state also then supporting
21    the payment of reciprocal compensation payments
22    to ISPs as provided for under the
23    Telecommunications Act?
24              MR. JOHNSON:  If I said what you
25    understood me to say, then I misspoke.  What I

                                                  136
 1    referred to was the use of the public switch
 2    network to access the Internet, the debate being
 3    whether it was essentially a local call, which is
 4    the way it's been treated now for however many
 5    years since the so-called ISP exemption.  I don't
 6    know if you could convince a customer who gets
 7    charged a local call charge for those calls to
 8    his Internet service provider that that in fact
 9    is a problem of interstate jurisdiction and
10    something that the FCC does.  I'm sure they would
11    come looking to me if they didn't like it.
12              But I said that our department had
13    filed comments in an FCC proceeding to the effect
14    that we thought those calls looked like local
15    calls and were within our jurisdiction.  Now,
16    what was the second half of your question, which
17    I'm not going to answer, by the way.
18              QUESTION:  If the access portion of it
19    is local, in other words, the call placed from
20    the consumer to that ISP, then that's a local
21    call.  It's a local call and it's local in
22    nature, then under the Telecommunications Act,
23    there is a reciprocal payment that could be made
24    from the incumbent LECs to the ISP for that
25    interconnection.  Some of the ILECs are refusing

                                                  137
 1    to make those payments, saying these are not
 2    local calls, that they are actually interstate in
 3    nature and I was wondering where New York State
 4    will come out on that when this comes in front of
 5    them?
 6              MR. JOHNSON:  Where we will finally
 7    come out, I can't tell you because I don't know.
 8    Back when it was NYNEX, our largest telephone
 9    company decided to tell all its interconnected
10    LECs that it was no longer going to pay the local
11    termination -- for local termination on that
12    traffic, and the Department politely told the
13    company to continue doing so until we had made a
14    determination on that and we do have a proceeding
15    addressing exactly that issue.
16              MS. BROWN:  Looks like you're up.
17              QUESTION:  Paul on behalf of Teleport
18    Communications Group.
19              In light of the advances made on
20    Internet telephony that we're hearing about
21    today, has that caused any second guessing or
22    further reflection upon the wisdom of recent
23    policies which essentially tax the incumbent
24    telecommunications industry to subsidize the
25    network builds to allow broader Internet access

                                                  138
 1    to schools and libraries?  That sort of folds
 2    into the issue of regulatory parity and so
 3    forth.  Any further thoughts on that?
 4              MR. JOHNSON:  I'm assuming you're
 5    asking the FCC.
 6              QUESTION:  Anybody.
 7              MR. JOHNSON:  That's their role,
 8    right?
 9              MS. CORNELL:  Actually, Tom, do you
10    want to comment on that as the Administration
11    person?  Nice try.  You haven't spoken yet.
12              MR. KALIL:  The short answer is no.
13    I think the Administration is very much committed
14    to getting technology into our schools, not only
15    network connectivity, Internet access, but also
16    increasing the number of computers, providing
17    high quality educational software and providing
18    teacher training so that they know how to use the
19    technology.
20              So the implementation of
21    Snowe-Rockefeller and the discounts for schools
22    and libraries is only one piece of the initiative
23    but we think that is a very important piece.  And
24    we think that this initiative will be good for
25    the telecommunications industry broadly because

                                                  139
 1    once kids have multi-megabit access in the
 2    schools, they're going to want the same thing in
 3    the home.
 4              MS. BROWN:  With that, I think there is
 5    much more to discuss, but I'm sure our people are
 6    starting to feel a little bit of hunger, so let
 7    me wrap this panel up and thank each of you for
 8    your comments and all of you for your questions.
 9    Let's give a round of applause to this panel.
10              (Applause.)
11              MS. BROWN:  Let me just echo Diane's
12    invitation for any of you to talk with us, the
13    United States agencies who are involved in WTO
14    implementation.  Larry's very much involved, and
15    some of his international staff, some of our
16    international staff from NTIA are also here,
17    Susan Settle is here.  Larry's here.  I know he's
18    always around the world talking with ministers
19    all over the place and he's been listening very
20    hard today so I think you should catch his ear
21    while he's around.
22              Let me just say that lunch is on your
23    own.  There's a list of restaurants, some
24    selected menus on the table outside.  For those
25    of you who don't know this facility, you're in

                                                  140
 1    the shops area.  There are lots of places you can
 2    catch some lunch.  There are also demos that we
 3    will have in the demonstration room that the
 4    VON Coalition is supporting and is hosting.
 5              There will be Web-based call center
 6    types of demonstrations, phone-to-phone through a
 7    gateway, and net-to-phone applications in the
 8    demo room.
 9              At 1:30, there will be -- when we
10    reconvene, there will be a demonstration here.
11    I'm hoping you will all come back this
12    afternoon.  I think we have some more interesting
13    things to talk about.  And with that, thank you.