Infrastructure for a Knowledge Based Economy

1
Infrastructure for a Knowledge Based Economy

• Walter Stewart
• June 12, 2006

2

• What I hope to do is provide you with perhaps a
  different context for your work.
• Context is everything.
• When you are up to your ass in alligators.

3

• e-Infrastructure
• Intelligent Infrastructure
• e-Science Infrastructure
• Cyberinfrastructure
• Grids

4

• FIRST, ID LIKE TO START WITH A STORY

5

• The S.S. United States

6

• The First Cars Looked Like Buggies
• The First Aeroplanes Looked Like Steamers and
  Pullmans
• WE HAVE TO BE CAREFUL HOW WE CONSTRAIN THE FUTURE
  BY THE WAY THE PAST CONSTRAINS OUR IMAGINATION

7

• WE MUST BE ESPECIALLY CAREFUL OF IDEOLOGY
• One generations revolutionaries are frequently
  the next generations reactionaries.

8

• As we replace the infrastructure of 1999, do we
  uncritically clone it with the 2006/7 version
  faster, smaller, cheaper?
• Or do we recognize that there is cause to develop
  whole new orders and types of infrastructure to
  take us towards the end of the first decade of
  the new millennium and beyond?

9

• That cause is data and the fact that we have
  begun a knowledge economy.

10
 

• So what about a knowledge based economy?
• What are the raw materials?
• Where are the raw materials?
• What are the processing stages?
• Where does processing happen?
• What are the finished products?
• Who is needed to create the finished products?
• Where are the markets?

11
 

• What are the raw materials?
• DATA

12

• Given the relative growth of Bandwidth, Moores
  Law, and Data, we clearly need a new model.

Data
Performance
Moores Law
Bandwidth
06
96
Source Gartner Group, June 2003
13
 
http//www.sims.berkeley.edu/research/projects/how
-much-info/ http//www.sims.berkeley.edu/research
/projects/how-much-info-2003/
14
 

• Note 2002 before
• Serious deployment of RFID
• Serious deployment of sensor networks
• Serious deployment of Grid Computing
• The LHC
• A number of new Telescopes, Synchrotrons, Shake
  Tables
• Development of Wireless Networks

15
 

• Sure much of it is noise!
• Thats why it needs processing or refining.

16
 

• Where are the raw materials?
• Just about everywhere!

17
 

• What are the processing stages?
• Ingest
• Storage
• Mining
• Processing
• Visualization
• Dissemination
• Multiple Reuse

18
 

• Where does processing happen?
• Multiple locations
• On the network
• Who cares?

19
 

• What are the finished products?
• Knowledge

20
 
Who is needed to create the finished
products? Multi-locational, multi-disciplinary,
multi-organizational teams
21
 
Where are the markets? Everywhere!
22
 

• A Disruptive Technology
• Team rather than individual
• Lateral rather than vertical
• Iterative rather than finished
• Network based rather than locational
• Intangible rather than tangible
• Managed complexity rather than imposed simplicity
• Access rather than ownership
• Trust rather than control
• Operating costs rather than capital costs

23
 

• A Disruptive Technology
• Team rather than individual
• Iterative rather than finished
• Network based rather than locational
• Intangible rather than tangible
• Managed complexity rather than imposed simplicity
• Access rather than ownership
• Trust rather than control
• Operating costs rather than capital costs

Note None of these are technical issues!
24
 

• CONSIDER THE COSTS OF NOT
• BUILDING IT
• USING IT
• REORGANIZING AROUND IT
• AND CONSIDER THE COSTS OF NOT BEING IN THE LEAD

25

• INFRASTRUCTURE AND THE BOTTOM LINE
• ORGANIZATIONS WHICH CANNOT MAKE EFFICIENT,
  EFFECTIVE USE OF THEIR RAW MATERIALS FAIL

26

• This is not the future it is now
• It is being built, and not just as a research
  project
• Bank of America
• Credit Suisse/First Boston
• Pratt and Whitney
• Novartis
• Boeing

27

• LINKED, THE NEW SCIENCE OF NETWORKS -
• Albert-Laszlo Barabasi,
• Emile T. Hoffman Professor of Physics, University
  of Notre Dame

28
RIDING REDUCTIONISM WE RUN INTO THE HARD WALL
OF COMPLEXITY
29

• COMPLEXITY NEEDS A DIFFERENT APPROACH AND A
  DIFFERENT INFRASTRUCTURE

30

• Lessons From History
• the lever, the catapult, the wind,
• the horse, the steam engine, internal
• combustion, the jet engine, nuclear
• power
• All productivity spikes have been as a
• consequence of harnessing
• unprecedented power

31

• We are now engaged in bring the process
• from the realm of physical force to the
• realm of the mind.
• Separating the grunt work from the
• insight
• We dont even have commonly
• understood language to describe
• what we are about.

32
A Guide to Grid Computing

• Solving the Human IO problem
• its a question of the medium
• Building Communal Intelligence
• its also a question of the medium
• the oil and gas example

33

• Is the possibility of genuine communal
• intelligence the basis of the next big
• productivity spike?
• GRID OR CYBERINFRASTRUCTURE IS TECHNOLOGY THAT
  MAKES COMMUNAL INTELLIGENCE POSSIBLE IN
  HETEROGENEOUS, DISTRIBUTED ENVIRONMENTS

34

• Again, this is context.
• Id ask you to think about these questions as you
  go about your work and particularly as you think
  of your future work.

35

• Thank You