Factors Affecting the Future Evolution of the Internet

1
Factors Affecting the Future Evolution of the
Internet

• George Sadowsky
• New York University and the Internet Society
• george.sadowsky_at_nyu.edu
• CEENET Network Training Workshop
• Budapest, 20 August, 2000

2
I learned in track 3 today

• Speakers have to know their audience
• You represent liquid intelligence
• My intelligence is more crystallized, and I
  suffer from memory constraints
• You are motivated to learn
• English is your 2nd or 3rd language

3
Introduction

• Internet is important to us all
• Wed like to know where its going for a number
  of reasons
• Planning
• Making money
• Intellectual curiosity, speculation
• Career progression

4
Introduction

• Progress and direction may vary according to what
  country youre in and where you are in the
  country
• Ill discuss many trends from developed counries
• And some applicable to developing countries
• For CEENet region, lag is 3-10 years, maybe
  narrowing
• FORMAT Consider this in part a discussion, a
  debate

5
Futurists

• Claim to know the future
• Do not have a good track record
• Early Book of Revelations predicts end of world
• Yardeni and Yourdon predicted a major depression
  and the end of civilization, respectively, in
  2000

6
Futurists

• Either do individual people who should know
  better
• U.S. Patent Commissioner in 1899 Everything
  that can be invented has already been invented
• Ken Olson Why would anybody want a computer in
  their house?
• Bill Gates, 1981 640KB should be enough for
  anyone

7
Futurists

• Track 2 showed a list of such bad predictions
  today
• Insult ones intelligence truth is not anyones
  monopoly
• I am NOT a futurist
• More modest, more sensible, and more useful to
  examine forces that will determine where the
  Internet goes, and then assess for self, rather
  than having to accept someone elses digestion in
  faith

8
Futurists

• Approach as economist and informatician
• Look at demand for and supply of network related
  stuff
• Examine a number of issues not so easily
  classified or separated
• CHALLENGE Identify significant factors that I
  have overlooked

9
Success factors for the Internet until now

• Interoperability rules (protocols) were
  straightforward
• There was distributed control over parts of the
  net
• Research and education drove initial years of
  Internet
• Guiding institutions (IETF, IAB) were
  non-commercial and non-governmental

10
The current Internet

• N nodes, in almost every country
• Massively commercial, following privatization in
  1994-5
• Complex structure, private peering, no backbone
  per se
• NSI claims 400,000 new DNS registrations per week
• From 1 to 80 registrars in 2 years (work of ICANN)

11
Categorizing the effects

• Supply push - positive (strength, encouragement)
• Supply weaknesses -negative
• Indeterminate supply issues
• Demand pull - positive (strength)
• Demand weaknesses - negative
• Indeterminate demand issues

12
Supply push - positive

• Cost of electronics - Moores law
• Twice the number of transistors on a chip every
  18 months at same cost
• Affects memory, processing speeds
• Important for router capacity
• In 1995 no. of core routes in Internet
  threatening to overwhelm largest router memory
  possible

13
Supply push - positive

• Increasing returns to scale - Metcalfes law
• Technological progress in wavelength multiplexing
• Cost per wavelength predicted to go down
  dramatically in next year
• Massive movement of content to web
• Commercial firms HAVE to be there
• Electronic publishing a strong force
• 5 of content on web today?

14
Supply push - positive

• Training trickledown - CEENET workshops a factor
• Rapid growth in wireless devices with screens
  (Palm Pilots, Handspring, cell phones, etc.)
• Strengthening of open source (software) movement
• Linux, Gnome, StarOffice

15
Supply push - positive

• Approaching useful digital signatures and their
  legal and business acceptance
• Tools for easy encryption how fast are they
  being delivered
• RSA public key encryption patent expiring this
  year (2000)

16
Supply push - positive

• Geographic positioning inexpensive, small,
  accurate GPS receivers - an entirely new
  dimension of information
• BACNET - control and automation network
• Moving to TCP/IP protocol
• 18 sensors per room at NYU

17
Supply weaknesses

• Longer run pessimism regarding persistence of
  Moores law
• Engineers can see 3 product cycles out in 10
  years
• 10 year horizon for progress narrowing
• Problem of intellectual property protection - no
  protection implies suply constraint

18
Supply weaknesses

• No successful micropayment schemes have yet
  emerged
• Electricity meters and French telephone
  impulsation counters implement micropayment
  schemes
• In North America, much telephone traffic is
  billing data for small amounts

19
Indeterminate supply issues

• Speed and difficulty adopting IPv6
• Speed of progress of UNICODE
• Movement of high tech labor between countries
• Visa approvals from developed countries
• Brain drain from developing countries
• Effects of outsourcing, e.g. Bangalore
• If people cant get to work, work will be sent to
  people

20
Indeterminate supply issues

• When will effective QoS be realized, allowing
  differential services and pricing?
• Even the Post Office (PTT) has implemented QoS

21
Demand pull - positive

• Apparently insatiable demand for content
• Number of people on Internet increases
• Number of applications learned increases
• Time per application may increase as learning
  occurs
• New applications make increased use of Internet,
  sometimes through transparent applciations, e.g.
  net agents
• Is demand a 4th power function?

22
Demand pull - positive

• Search engines empower one to take over functions
  that were previously in province of library
• But not without error, omission, and
  inconsistency
• Question What of worlds content is on the
  web?

23
Demand pull - positive

• The personal imaging revolution
• Explosion of consumer tools for digital image and
  video manipulation
• Miniature cameras such as Sony announcement
  7/2000
• More digital cameras than film cameras will be
  sold in 2000

24
Demand pull - positive

• Ease of use browser interface multiplied number
  of possible users enormously.
• Recognition of problems posed by digital divide
• What is the digital divide?
• How serious is western commitment to eliminating
  it?

25
Demand weaknesses

• End to long period of economic expansion in west
• In some developing countries, lack of resolve (or
  money) by G-8 and others to follow through on
  narrowing digital divide
• Minor issue of taxation of goods and services
  delivered on the net

26
Indeterminate demand issues

• Is the home market real?
• How big is it?
• How quickly will it be developed?
• Where are the limits of its use?
• Will video on demand over the net be popular,
  with what effects on congestion (in absence of
  effective QoS mechanisms)
• Newly announced movie on demand service
  advertises 100 Gb for 3.00 (movie)
• How much difference will compression improvements
  make?

27
Indeterminate demand issues

• The entertainment market
• How will the entertainment market shake out?
• What will be the effects of convergence of
  entertainment, computing and communications?
• Is this a fad or is it permanent/
• What will be the next big plug and play
  breakthrough after the web interface? Speech?
  Other?
• Son of WEB?
• Every time one simplifies, more people can
  participate

28
Indeterminate demand issues

• Speed of development of integrated productivity
  applications
• Voice over IP a beginning ISP
• IP trunking a temporary solution to wrong problem
• Balance between centralization and
  decentralization
• HW Mainframes, time-sharing, stand-alone minis
  and PCs, LANs, Internet
• Depends upon relative costs, economies of scale,
  need for low latency, throughput, etc.

29
Indeterminate demand issues

• How will network security evolve
• Involves both net security and (web, e-business)
  server security
• Presently, electronic warfare on large scale
• UNIX, multi-user machines the target
• NT gaining a reputation for weaknesses
• Some UNIX systems shipped with no passwords
• Linux and OS-X aggravate the problem with amateur
  systems managers
• Every 3 months, secret Service visits NYU - how
  does that fit with publicly available IP tone or
  Web tone?

30
Indeterminate demand issues

• Probes, break-in kits, active counterculture,
  hackers
• Network of networks NOT integrated at management
  level (unlike at the technical level)
• Address spoofing
• No uniform protcol for handling complaints,
  assigning responsibility
• Every ISP needs to check packet return addresses
• Fear of massive meltdown,over-dependence on the
  network

31
Indeterminate demand issues

• What do we mean by the Internet
• Internet characterized by peer-to-peer, TCP/IP
  end-to-end
• Now have millions and millions of peers
• Always had larger network context
• UUCP, FidoNet, Bitnet
• But ALSO AOL!
• Shortage of addresses has caused growth of
  private addresses with less than full
  functionality
• New services I-mode in Japan, 10 million users
• Will have to think of more complex networking
  universes, with interlinkages between regions

32
Indeterminate demand issues

• Administration and governance
• History is one of RD work, informal
  administration, IANA, Postel
• IANA has led to ICANN
• ICANN does NOT govern net, it administers crucial
  net functions that are technical AND
  administrative
• Issue of formation, board, individual members,
  director election, 158K members
• Organized membership campaigns in Far East
  (Japan, Korea, China)

33
Indeterminate demand issues

• Governance issues
• Internet has been governed by consensus and
  emergence of business practices
• Deregulation of US telecommunications sector
  allowed for rapid growth of Internet
• Lesson still not understood in all countries
• OSI/ISO diversion in Europe hurt development in
  Europe
• ITU has been late in understanding impact, too
  litle, too late
• But if instability, may see ITU/UN make a play
  for authority over the net

34
Indeterminate demand issues

• Anti-trust issues
• Effect of eventual final Microsoft judgement
• What if Cisco were suspected of monopolizing a
  market. IOS is proprietary code.
• NSI continues to defend its quasi-monopoly status

35
Some developing country issues

• Inadequate supply of connectivity constrains
  demand
• Africa One abandoned, Latin American ring planned
• Satellite technology (WAN) and wireless
  technology (LAN) are helping

36
Some developing country issues

• Some issues are just being faced for first time
• Role of government in telecommunications
  privatization
• Role of a regulatory authority
• Governments role with respect to content control
• Tension between privacy and authority

37
Conclusion

• The space of Internet evolution is
  multidimensional and complex
• Reflects the complexity of the real world
• Compounded by the speed of technological progress
• Consumer interests
• Assure growth of coverage, universal presence
• Unimpeded technical progress
• Competition among providers at all levels

38
Conclusion

• Provider/Producer interests
• Provide huge markets
• Establish reliable revenue streams
• Make money
• Government interest are more diverse
• Ensuring enforcement of law
• Whats next?
• Well just have to wait and see .