1
Point Counter-point with Bob LuckyGordon
BellMicrosoftMay 5, 1998

• The 1998 Bellcore Forum Competing in the
  Millennium
• Vision, Innovation, and Delivery
• Audience response was used for this
  presentation, and the attendees from Bellcore,
  LECs, CLECs, IECs/IXCs, Cable companies, and
  equipment suppliers were polled. The s indicate
  the results of these polls.

2
Predicting future telecom services or lack thereof

• Introducing the Internet and Web driver
• Cyberspace the quest and vision my opinion
• Laws of Prediction or how to understand the
  technology of our future
• The big telecommunication bets if we could only
  forecast the future
• Overthrowing the incumbents a plan
• Appended to the original talk.

3
By January 2001 there will NOT be 1 billion
people on the net.Bet Nicholas Negroponte
1KBet Nicholas Negroponte 1K5K it
happens by 2002.Also 1 T of commerce by 2001.
Bet Me? 77 NN? 23
4
Why this is the keystone bet!

• It determines the market
• for networks
• for access devices especially PCs
• It says something about the utility
• commerce
• communication
• entertainment
• Increased network capacity ubiquity enables
• phones
• videophones
• television
• serendipity

5
Interneters growth
World Population
10000 1000 100 10
TVs Phones
1 Gp by 2000Negroponte
PCs
Internetters
95 96 97 98 99 00 01 02 03 04
6
Interneters growth
12000 10000 8000 6000 4000 2000 0
World Populationextrapolated at 1.6 per year
Internet Growthextrapolated at 98 per year
95 96 97 98 99 00 01 02 03 04
7
(No Transcript)
8
Growth in hype
Data from Gordons WAG
9
Articles about security, privacy, fraud versus
commerce (M)
Data from Gordons WAG
10
Real growth in commerce?
11
The Virtuous Cycle that drives the BW quest
Userdemand
Application innovation
Internet(IP) ubiquity
Capac. (svc response)
Excess capac.--gtgtBW
12
Bob Lucky, Vice President Bellcore, 1995

• If we couldnt predict the Web, what good are we?

13
Exponentials change everything

• you cant see em coming!

14
Business Week even makes fun of Telecom because
you can see them coming
15
Business Week doesnt know Satellite isnt an a
B/W option
Phone CATV Satellite
16
Why phone guys hate computer guys

• 0. Where would they be without the transistor,
  RJ11 and UNIX that you de-standardized?
• 1. Theyre nouveau rich were old money.
• 2. We brought ISDN to the party, nobody came.
• 3. Theyre IPing on us including phones!
• 4. Packets dont make links fatter or faster.
• 5. All they want is free, high bandwidth, 24 hour
  a day phone calls. Why do they need bandwidth?
  They have nothing to show.
• 6. Dont talk service. I reboot my PC every day.
  Ever have to reboot your phone?
• 7. Just wait till the governments get on their
  case.

17
Agenda predicting and forecasting

• Introducing the Internet and Web driver
• Cyberspace the quest and vision my opinion
• Laws of Prediction or how to understand the
  technology of our future
• The big telecommunication bets if we could only
  forecast the future
• Overthrowing the incumbents a plan

18
Lets look at Cyberspace
19
Cyberspace A Network of ... Networks of ...
Body
Continent
Region/ Intranet
Home
Campus
World
20
Cyberization interface to all bits and process
information

• Coupling to all information and information
  processors
• Pure bits e.g. paper, newspapers, video
• Bit tokens e.g. money, stock
• State of places, things, and people
• State of physical networks

21
Platform, Interface, Network Computer Class
Enablers
The Computer Mainframe tube, core, drum, tape,
batch O/S direct gt batch
Mini Timesharing SSI-MSI, disk, timeshare
O/S terminals via commands POTS
PC/WS micro, floppy, disk, bit-map display,
mouse, distd O/S WIMP LAN
Web browser, telecomputer, tv computer PC,
scalable servers, Web, HTML Internet
Network Interface Platform
22
Everything will be in Cyberspace

• Is this a challenge? goal? quest? fate? or
• Cyberization enables new computing platforms
  thatrequire new networks to connect them
• Infrastructure supports the content
• Three evolutionary dimensions

23
All the platforms we have and will inevitably
build have to be totally interconnected to have
Cyberspace.Thats your job!The price has to
be right you cannot count on voice revenue
forever!
24
Connecting all these plain old computers into
Cyberspace an opportunity
25
Mobile videophone
26
Honda Robot
27
People surrogates
28
Steve ManninCyberspace
29
MedtronicsImplantedCardioplastic
30
For openers audio, pix, T, P, ECG, location1 GB
31
Not shown ECG GPS
Compass altimeter
PCS Pilot
Libretto, .5mm pencil
Libretto PS, Ricoh Camera Swiss Army Knife
32
Cyberspace one, two or three networks? in
2005, 2010, 2020
33
The Worlds of TV, Telephony Datacom a.k.a.
Computing Internet
Game Cons.
Telephony world
VCR
Television world
DVD gt97
Wire- less
Cable
Cable phone
Broad- cast
LEC
CLEC
IEC
xDSL TV
DBS
Cable Inet
PBX
Inet Phone
ITV?
The Internet
LAN PBX
Inet RADIO
RADIO
ISP
Pvt. WAN
IntranetExtranets
Wire- less
LAN
Datacom world
34
Cyberspace A spiraling quest in 3D real space
Computation
Cyberization
Communication
35
Agenda predicting and forecasting

• Introducing the Internet and Web driver
• Cyberspace the quest and vision my opinion
• Laws of Prediction or how to understand the
  technology of our future
• The big telecommunication bets if we could only
  forecast the future
• Overthrowing the incumbents a plan

36
On predictionand betting why historically have
we been right or wrong?
37
I think there is a world market for maybe five
computers.


Thomas Watson Senior, Chairman of IBM, 1943
38
Harvard Mark I aka IBM ASCC
39
Predictions require some history.

• The computer hadnt been invented.
• Watsons prediction held for 10 years.
• You need history to predict.

40
Moores Law vs Morons Law for prediction
differences in Cyberspace perception
41
Moores Law 60/yr. Memory -- 4 x size every 3
years
10 G 1 G 100 M 10 M 1 M 100 K 10 K 1 K 0.1 K
1970
1980
1990
2000
2010
42
Extrapolation from 1950s 20-30 growth per year
43
(No Transcript)
44
Moron's Law (cf. Newton)

• the residual effects of regulation ensure that
  telecom carriers will never provide what the
  customer wants and it comes five years too
  late.--venture capitalist Roger McNamee.
• Computer folks ... fear that Moron's Law will
  trump Moore's and stunt the revolution's growth
  over the next 30 years.

45
Newton's First Law. Bodies at rest ...

• LECs, PUC and FCC are at rest, supplying
  assuredly profitable (by definition) local
  service.
• Cable TV raises prices to cash out
• CLECs and IECs cherry pick lucrative corporate
  (not consumer) local service
• LECs see reduced revenue and profit
• LECs, PUCs, and FCC raise prices
• LECs downsize to remain at rest, buy cable and
  CLECs to remain at rest maintain monopoly

46
Electricity and LECs to Merge

• Wires to the home
• Guys and trucks
• Poles and holes
• Same customers
• Geo-monopolies
• PUC interaction
• Core competency lobbying

47
Why computer guys hate phone guys

• 0. Ads say someday you will. Just do it!
• 1. All they give us is POTS to IP on.
• 2. Its not the price of bandwidth (that they
  said would be free) its the availability
• 3. They wont buy packet switching. Computers let
  everyone be telecoms. Theyll pay.
• 4. The net is the net. Services are services.
  Content is content. Stop trying to own ALL.
• 5. Telephone quality is an oxymoron. With 64
  kbps, we should be getting CD quality.
• 5. Name any monopoly that has love respect
• 6. Its taken 35 years to learn that computer
  calls last 24 hours/day, not 5 minutes.

48
(No Transcript)
49
Why no one likes cable guys

• 0. But you just increased my rates
• 1. They dont even own the poles or holes.
  Theyre just schmucks in trucks.
• 2. They are too close to the broadcast industry
  and that eliminated logic.
• 3. Wed all like to be unregulated too then we
  can raise rates faster and do whatever we want.

50
Telecom View of Cyberspace We will tell you the
services.
The Network and all the Sevices you would ever
want.
Subscribers
51
Silicon Valley View of Cyberspace provide a
switch and entrepreneurs will create the services
All the Services you might or might not want
Subscribers
Worldwide mostly IP Switch
52
Telecom Equipment Industry 1998
ATT
Nortel
Siemens ROLM
NEC
Solutions Applications OS Computers Switching/DSP
53
Future Telecom Industry

• Applications
• Applications
• Databases
• OS
• Switching
• Computers
• DSP
• Processors

Ericsson, Aspect, Nortel, Octel, others
Microsoft, Delrina, many others
Informix, Microsoft, Oracle, Sybase, others
Microsoft, Apple, Sun, Novell, LINUX
Ericsson, Nortel, Bay, 3Com, Fore, others
Compaq, DEC, Dell, IBM, many others
Dialogic, NMS, Rhetorex, others
Intel, AMD, Motorola, others
54
Computer vs Telecom

• Proc. mem 2x/1.5 yr Access B/W 2x/4yr Trunk
  BW 2x/ 9 mos.
• Prices decline Prices increase
• Email, web, audio, etc. Voice, voicemail, fax
• Packet switching Circuit switching...
• will too, do voice! is required for voice
• 64Kbps gives great audio Telephone Quality Voice
  is an oxymoron
• IP ATM, ISDN POTS
• Fungible bits for Unique bits and phone fax
  data tv ala carte pricing
• Connect bits go free. Billing lawyer costs
• Just give us any old net! We want it all
  well compute the rest net, nodes, content
• Technology-based its Lobbying-based its
• Information Services Telecommunications
• Moores Law Morons Law

55
The beginning of the end or a new beginning?

• You have to separate the hype from the reality.
  We have customers, real customers paying hundreds
  of millions of dollars to us today. When you
  look at Level 3 and Qwest, how many customers do
  they have? ---- R C Notebaert, Ameritech
• Were not going to wake up and find our voice
  business is gone--- M Turner, US West

56
A technology e.g. packet switching can come from
nowhere and wipe you out!

• MOS memories and micros were introduced in 1972.
• By 1976, MOS memories wiped out core memories.

57
There is no reason anyone would want a computer
in their home.

• Ken OlsenPresident, Chairman and founder of
  Digital, 1977

58
Equating yourself to the averageuser/buyer
is risky . . . unless youre an average user.
59
ISDN will be ubiquitousby 1985.

• Irwin Dorros, VP Long Lines ATT, 1981

Unfortunately were stuck with ISDN speed using
POTS for the next 5 years. Gordon Bell, ACM 1997
60
Network bandwidth becomes available slower than
the most conservative prediction.
61
ATT will not have screwed up its purchase of
NCR by 1996.

• Robb Wilmot, advising Robert Allen in 1992 (100)

62
We under-estimate the devastating power
ofcompanies planners, lawyers government,
to foul up predictions.
But to really foul up requires an econometric
and market models!
63
By 1997 Video-on-Demand will be available and
operating in six cities


When will this happen?2000 8, 2005 46, 2010
25, never 21

• Raj Reddy Ed Lazowska1992 (winners get fed)

64

By April 1, 2001 videophones will ship in 50 of
the PCs and be in 10 daily use.
Will this happen by 2001 25, 2005 50, 2010
18, 2015 3, never 4

• Gordon Bell vs Jim Gray1996 (one paper, loser
  gets fed)

65
Always bet against the optimistic researcher.
They are wrong in the short term, but ultimately
right.
66
Build it and they will come

• Vision Provide bandwidth and they will come to
  share supercomputers, do telemedicine,
  telescience, ship pictures around. Traffic
  continues to double.
• The web was serendipity!
• "A user ... would build a process and launch it
  into the network ... with explicit instructions
  about things to look for and what to do ."

67
(No Transcript)
68
If you do the right thing, serendipity works!
69
NCs will NOT outsell PCs 91 by 2000.

• NCs include those embedded in TV sets, phones,
  and used as PC alternatives.
• Bet with Larry Ellison, CEO Oracle While the
  devices connected to the web (e.g. instruments,
  cameras, appliances, printers, phones, and
  television sets) may be greater than PCs, the
  number of personal access devices that are NOT
  PCs will be less than 91.

70
Dont bother betting where the goal is just PR.
It is safe to bet against Larry Ellison even if
he has lots of money.Elllison bought Ncube. It
failed as an MPP computer, failed again as a
database engine, and finally failed as a
video-on-demand server.
71
Observations on predicting

• Existence proofs are essential, otherwise its
  faith and luck.
• Numbers and data are our friends to use.
• Bet against the optimistic researcher. They are
  ultimately right, while initially wrong.
• Bet on predictors who are grounded, intuitive,
  imaginative, and lucky.
• Because it could, doesnt mean it will.
• Its usually just the economics, stupid look
  there first!
• Dont count of Newtons First Law because greed
  (an outside force) usually beats fear (inertia).

72
What percent of biz traffic by contacts, (not
traffic) will be

• Item Bell/Lucky 1994 2000
• Fax [phone]
• Voice 15 50 15 30
• Voice/Video confer 2 2 10 10
• Overnite mail 2 3 1.5 5
• Snail mail 20 25 10 25
• Email 50 10 56 30
• Newspaper?
• Web?
• Television?

73
Agenda predicting and forecasting

• Introducing the Internet and Web driver
• Cyberspace the quest and vision my opinion
• Laws of Prediction or how to understand the
  technology of our future
• The big telecommunication bets if we could only
  forecast the future
• Overthrowing the incumbents a plan

74
Some of my recent bets!Would you bet with me?
75
PC growth is not stalled...

• Bet (Walt Mossberg, WSJ) There will be
  continued growth of PCs (I.e. things that run
  Windows, NT, or CE) at a double digit rate for
  until 2001.
• Bet with me? 87 With Mossberg? 13

76
Web will reach 50 of the U.S. households...

• Negroponte predicts 1Bp and 1T/yr of commerce
  by 1/2001.
• Bet with Dave Nagel, ATT one dinner The web
  will reach 400M world-wide users OR 50 of U.S.
  households by end of 2001.
• Do you bet with me? 57 Nagel? 43

77
One million people will access the web via TV
sets...

• BetAt least one million users will access the
  web/internet via their television sets by the end
  of 2001.
• Via phone line, 16xDSL 19 cable
  modem 17 settop box/cable 39not at all 9

78
Average bandwidth to homes will not soon be huge

• High bandwidth links are being delivered today
  in trials of 10K using cable xDSL.
• Bet The average bandwidth to all U. S. homes
  with PCs will be lt56 Kbps in 1/2001.
• The availability of bandwidth coming to consumers
  and small businesses by 1/2001 via Cu twisted
  pair aka LECs or IECs will be comparable to ISDN
  today.
• lt56Kbps 71 gt128Kbps 29

79
A gigabit net will not be in place for research
users

• Bet Raj Reddy et al fine food drink10K
  users at 10 sites of gt 500 users/site in 3
  states will NOT inter-connect through a gigabit
  path by the end of 2000.

Do you bet with me? 65 Or with Raj? 35 Based
on market need, this service will be available
by 2000 10, 2005 53, 2010 29, 2020 5, never
4.
80
Telepresentations will be a well-defined app by
2001.

• ACM97 aka its 50th on March 3-5, 1997 was the
  first telepresented conference with Mbone
  multicast and servers
• Bet with Denise Curruso, NY TimesMore people
  will view conferences that are made telepresent
  from Cyberspace than those attended it in
  meatspace by 2001. Note this bet was won within
  6 months.
• Bet Even this meeting will be telepresented by
  2001 37, 2005 35, 2010 10, never 18

81
ACM 97 (Mar. 1997) attendee poll whats your
opinion?

• By 2047 will the majority telecommute? (68 of
  ACM attendees said yes)
• Will representative democracy be replaced by
  electronic participatory, democracy by 2047?
  (ACM 38 yes)
• TV, the internet and telephone will converge by
  2002 (ACM) ! clearly optimistic and very wrong

82
Lets poll (forecast) when certain events will
(or will not) happen
83
Forecast of Internet Telephony Videotelephony

• IP will carry 2B of existing voice or fax
  traffic by 2001, 2005, 2010
• IP will carry 20 of existing voice or fax
  traffic by 2001, 2005, 2010
• IP will carry 50 of voice and fax traffic by
  2005, 2010, 2020, never
• Note email can be used to carry fax traffic, but
  not visa-versa.

84
Forecast of xDSL CABLE of Internet to US Homes
Businesses

• xDSL use will cross-over ISDN installations by
  2001 55, 2005 37, 2010 8
• xDSL will deliver service to 10M by 2001 10,
  2005 69, 2010 22
• Cable will provide WWW to 1Ml by 2001 61, 2005
  31, 2010 8
• Cable will deliver WWW to 10M by 2001, 2005, 2010

85
Forecast of convergence?

• The datacom network (IP) will carry Television
  programs by 1/2001, 2005, 2010, 2020, never
• There will be a single net to access the
  majority of telephony, data, and television
  sources by 2005, 2010, 2020, never

86
Forecast of the access device

• One-half the homes with PCs will be always on
  by 2001, 2005, 2010, 2020
• The main (in units) web access device in 2001
  will be the PC, Telecomputer aka videophone,
  Set-top, TV, other
• The main web access device in 2005 will be the
  PC, Telecomputer, Set-top, TV, other
• The main web access device in 2010 will be the
  PC, Telecomputer, Set-top, TV, other

87
Agenda predicting and forecasting

• Introducing the Internet and Web driver
• Cyberspace the quest and vision my opinion
• Laws of Prediction or how to understand the
  technology of our future
• The big telecommunication bets if we could only
  forecast the future
• Overthrowing the incumbents a plan

88
Ten Niches to the IECs economic support base.
Reducing voice revenue is the only avenue for
changing the LEC-PUC-FCC controlled access
infrastructure!
89
Overthrowing bandwidth limiting incumbent IECs to
get Internet access ten, critical niches

• Plain Old Cellular replaces POTS
• 64 Kbps quality audio vs telephony quality
• Cable for Internet AND voice over IP
• CLECs aka Competitive LEC suppliers
• Newcos (e.g. Covad) IP service with xDSL
• Startup IP based telephony companies
• Large users by-pass LECs and use IXCs and IP for
  inter-organization telephony
• Fax services over Internet undermines base
• Bellcore plan for ISPs to provide telefony
• Maintain FCC support FCC for voice over IP
  without FCC/PUC interference

90
Telephone Quality Voice is an Oxymoron! We can
do better.

• Analog to digital transition must occur!
• E.g. a 6 Mhz TV channel can deliver about six
  higher DVD quality digital channels
• IECs use 64 Kbps channels to supply approximately
  8 kbps, low-quality voice
• New channels can offer higher quality audio using
  gt33 kbps.
• Datacom must drive to offer scalable audio
  bandwidths from 2 kbps-500 kbps

91
Attack plan for ISPs to deliver voice

• According to the rules, ISPs cannot offer
  telephony if it looks like a telephone
• Thee ways that make IP telephony NOT be a
  telephone
• Increased quality use all the bandwidth that can
  be obtained e.g. 33, 56 or whatever to provide
  high, not telephone quality voice
• It is a videophone
• It is a computer and conference device

92
How a New Player or ISP can Attack Incumbents
e.g. LECs

• Dial Me Up, Faster and Faster!!!!

Internet Backbone
IXC
Telafony Gateway
Voice over IP at 7.3 Kbps Using Modems
ISP
LEC
courtesy of Bellcore
93
Plan for Pipes and Wires
ATM Switches
IP Routers
LEC Network
LECs are free pipe and wire access providers. IP
over ATM lets competitors buildout a network
94
FCCs 4/10/98 Report to Congress How long will
it remain?

• If you use IP to transmit and store information,
  including telephony (unless it looks like a plain
  old telephone call), its Information Services.
• If you use circuit switching (no storage) its
  Telecommunications (and FCC/PUC controlled)
• IP Telephony has to look different!
• BetThis will hold until 1/2001.

95
The end