The Past, Present and Future of the ICT Revolution

1
The Past, Present and Future of the ICT
Revolution

• Kenneth I. Carlaw
• Associate Professor, Economics
• University of British Columbia Canada
• And
• University of Waikato New Zealand
• kenneth.carlaw_at_ubc.ca

2
Broad Research Agenda

• Lipsey, Richard G., Carlaw, Kenneth I. and Cliff
  Bekar (2005) Economic Transformations General
  Purpose Technologies and Long Term Economic
  Growth (Oxford University Press Oxford, UK)
• Agenda Understand (very) long run growth driven
  by technological change
• Fact set History, 8000 BC to present
• Proposition 1 General Purpose technologies
  (GPTs) are necessary to rejuvenate and sustain
  growth in material wealth
• Proposition 2 Humans are technological animals

3
Preliminaries

• Transforming Power of Technology and Economic
  Growth
• People in the 21st century experience measured
  real consumption more than ten times larger than
  that of people at the beginning of the 20th
  century.
• But the whole story is not solely in the
  measurements
• New Techniques
• Electronic, CNC, robotized production versus
  steam powered, belt driven factories
• New Commodities
• ipods versus cheap cotton textiles
• New forms of Organization
• Just in time delivery systems versus steam
  factories populated with children

4
GPTs

• Definition
• Initial scope for improvement
• Range and variety of use
• Complementarities
• GPTs in history (24)
• E.G., Writing (3000 BC), Electricity (19th 21st
  century)
• Evolution of a typical GPT
• Efficiency and Applications
• Spillovers
• Path dependence
• Facilitating Structure

5
The GPT of the Modern ICT Revolution

• Programmable Computing Networks (PCN)
• Devices that exploit flexible machine logic to
  generate, transmit and process electronic
  information signals
• Includes computers and the internet

6
The modern ICT revolution

• Structural Evolutionary transformation versus
  Technological change
• S-E Transformations Modern ICT Revolution
  compared to the First and Second Industrial
  Revolutions
• General Purpose Technologies (GPTs) are the prime
  movers that drive economic growth and
  productivity growth
• In this case PCN

7
The two major questions

• What does the productivity slowdown tell us?
• Where is the PCN in its development trajectory?
• (i.e., how much more transformation of economies
  and societies is to come from PCN can we foresee?)

8
Two Views of Economic Processes and Productivity
9
Productivity slowdown and ICT diffusion

• Traditional view
• Productivity paradox Solow (1987)
• The computer is everywhere except in the
  productivity statistics
• No paradox because ICT is not a revolutionary
  technology Triplett (1999), Gordon (2000)
• The error here is that TFP is interpreted to be a
  measure of technological change

10

• Structural-Evolutionary View
• TFP is not a measure of technological change
• See Lipsey and Carlaw (2004) CJE
• GPTs
• arrive in crude forms
• require costly, complementary innovation
• require investment to become embodied in
  pre-existing economic Structure
• often generate a productivity slowdown if they
  require significant structural transformation
• See Paul David (1990), Cliff Bekar (Ph.D. Thesis)
  and LCB (Chapter 15)
• Productivity bonus if there are any follow
• Carlaw (2004) DCITA, and Carlaw and Lipsey (2005)
  DCITA, Carlaw and Oxley (2005)

11
Answer to question 2

• Hypothesis the GPT of Electronic ICT is at a
  point in its efficiency and applications
  trajectories that is roughly comparable to the
  GPT of Electricity in about 1920-30

• Implication much of the economic and social
  impact of PCN is yet to come via the continued
  (accelerating) proliferation of its applications

12
Placing PCN in Its Efficiency Curve

• We simplify the many dimension of efficiency of
  PCN into two broad categories
• Engineering Process Advances
• E.g., Shrinking etching size of transistors
• Logic Advances
• E.g., the stored program computer, multi-core
  processors, the software compiler and software
  optimization

13
Engineering Process Advances

• Moores Law
• the exponential revenue growth of the
  semiconductor industry (averaging 16 per year)
  has continued unabated for 40 years
• there are more transistors produced per year (1
  quintillion) than grains of rice, and each rice
  grain costs the same as 100s of transistors

14
Logical Advances

• Stored programmable logic instructions
• Both software and hardware instructions
• Stored on the device
• Changeable
• New tasks accomplished by combining and
  recombining basic logical instructions
• Instruction sets of greater complexity
• Instruction sets themselves can be expanded
• Device can be used for more applications
• expansion in size and complexity

15
Two Logic Efficiency Effects

• The software hierarchy presents efficiency
  opportunities
• Shifting complex functions between levels of the
  hierarchy presents other opportunities
• Too many instructions at the lowest level implies
  too much complexity in the hardware logic,
  reducing efficiency
• Too many instructions at the higher levels
  increase the complexity of the software logic
  layers, reducing efficiency
• Optimization balances these

16
Future Efficiency Gains

• Limits of CMOS
• Multi-Core
• Quantum computers
• Nano-switches
• Ray Kurzweil and the singularity
• Extremely optimistic
• There are always winners and losers with new
  technology

17
Placing PCN in its Applications Curve

• Diffusion and New Applications
• The distinction is blurred
• Diffusion
• Practical measurement problems
• We must rely on incomplete data
• Usually collected by private commercial agencies
• Expensive to acquire in terms of time and money
• Often price data only
• Diffusion data for specific categories
• E.g., Figure 5.1 computer sale in Canada over time

18
Figure 5.1 Annual Sales of Personal Computers
in Canada (millions) Source International
Telecommunications Union (2005), Series I422
19
New Applications Beget Applications

• We can only directly measure those that are
  realised
• There are those that are not yet realised but are
  enabled by the given GPT, given its current
  efficiency

20
Measurement

• We use a qualitative approach that exploits the
  definition of the applications curve and its
  phases
• In phase 3 we will observe a proliferation of
  applications that lead to further applications
• In phase 4 we will observe a slowing in both the
  rate of applications generated and the follow on
  applications possible from those that are
  generated
• Applications become more specific an narrow

21
New Applications

• Computers that read your mind
• Computing Researchers have developed a promising
  new way to control computers by thought alone
• Cows go wireless
• The use of electronic tags to track cattle and
  monitor their health is likely to accelerate
• Robots, start your engines
• Could a robot race funded by a military-research
  organisation help to advance the development of
  autonomous fighting vehicles? We are already
  seeing this technology being used in Iraq and
  Afghanistan
• An open-source shot in the arm?
• The open-source model is a good way to produce
  software, as the example of Linux shows. Could
  the same collaborative approach now revitalise
  medical research too?
• In dust we trust
• They have generated a lot of hype. But might
  sensor networks, also known as smart dust,
  actually be useful?

22
Comparing PCN and Electricity

• Electricitys transition to Phase 4 of its
  efficiency and applications occurred about 80
  years after the invention of the dynamo
• If we take the beginning of PCN as the late
  1940s, that GPT has been evolving in efficiency
  for about 60 years
• Electricitys efficiency increased by something
  just less than one order of magnitude over its
  first 70 years
• PCNs efficiency has increased closer to an order
  or magnitude every decade

23
PCN and Electricity

• Electricity in Phase 4
• PCN in Phase 3

24
Extrapolation of Real Price Index of Electricity
for Canadian Domestic Use
25
Conclusions

• Efficiency
• PCN efficiency has increased rapidly over an
  extended period of time
• PCN may be approaching Phase 4 of its efficiency
  trajectory if one focuses on CMOS architecture
  but it is not there yet
• Multi-core processors, nanoswitches and quantum
  computers offer scope for much further efficiency
  gain

26

• Applications
• We have seen from the sample of applications for
  which we could get reliable data that their
  diffusion is far from complete
• Applications we are observing are enabling
  applications
• Even if efficiency stopped increasing today (PCN
  reached late Phase 4 or Phase 5 on its efficiency
  curve), many applications would remain to be
  exploited
• The union of PCN with biotechnology and
  nanotechnology will spawn a set of new
  opportunities for inventions and innovations over
  at least the next half century with no
  foreseeable limit in size and scope of impact

27

• Given the evidence, it seems clear that PCN will
  continue to have a profound and formative
  influence on economic and social change for at
  least the next several decades, offering
  countless opportunities for the development and
  exploitation of new applications