Rationale for IDEAs

1
VSS/DSS PROBLEMS

• Rationale for IDEAs
• How to solicit and sustain maximum IQ resources?
• How to Integration Disparate Knowledge, make it
  available and usable?
• How to Make Legacy Knowledge Persistent?
• Hypothesis - make it so that there no longer
  exists legacy knowledge that is, we are aware of
  all knowledge (its a flat system).

2
SCOPE

• First-Principle systems.
• Amenable to Engineering and Research systems
  (starting point).
• Possibly useful in business systems too
  (follow-up point).
• Hypothesis All systems are first-principle
  systems and if you say they are not, then that
  means you MUST go capture tacit knowledge!,
  otherwise, youre cheating yourself, the company,
  and the stockholders.

3
Target Participants

• Plasma Physicists
• Superconductor Specialists
• Thermodynamicists
• Systems Integrators
• Program Managers
• Sponsors, including politicians
• Educators
• Students of all ages
• General public everywhere in the world

4
Problem Questions

• Nonstandard 3rd-generation programming languages.
• Nonstandard documentation methodologies.
• Non-standard fields - how to integrated disparate
  Understanding (U)?
• Training costs
• Platform and software costs.
• Complexity
• Usability
• Transparency
• Design Prejudice
• parameter awareness
• value bias
• parameter dynamics

• Information Overload (Lerch Harter, 2001)
• Unfair RA competition
• Re-invention, loss of memory
• Small IQ access (loss of participants due to
  award winning)
• Non-invented here, hoarding information
• Cognitive Science dropped the ball that is, no
  cognitive cybernetic model to follow.
• HCI antiquated
• Point solutions, wisdom misused, inefficient
  designs, teaching wrong solutions and
  methodologies and bad habits industry and
  university engineering and science departments
  need upgrades, to teach ALL solution and
  Context-Down methodologies.

5
Problem Questions

• Why are DSSs nearly always closed to the masses?
• Platforms are not standardized
• Platforms require long learning times and special
  skills
• Intellectual property concerns
• Cost and access problems
• Maintainability and security
• Persistence and reliability
• Opacity to the embedded model (black box)
• Non-adaptability and non-scaling of the platform
  and embedded-models
• Platforms and tools are not well-designed to
  support natural human cognitive processes (NHCP)

6
Problem Questions

• Communication issues with other people.
• Is there a standard model of natural human
  cognitive processes?
• What are the properties of an ideal decision
  support system?
• Why do humans need a decision support system?
• Q. Can EXCEL notes be auto translated into
  Chinese?
• Include Business issues that youve seen with
  FAST, etc.

7
Problem Questions

• Prevent loss of knowledge over decades
• Integrate knowledge from
• disparate sources
• disparate subject fields
• legacy systems
• Make unique-complex systems common-simple
• software,
• hardware,
• knowledge model.
• Find the optimum solution over the Domain.

8
How to solicit and sustain maximum IQ resources?
9
Situation Competition for Scarce Funds
LANL HTc SC
JPL Integrate
JSC Missions
UMich Mirrors
LBNL ICF
10
Deplorable Result Loss of IQ
11
Better Sustain Original IQ
12
Best! Multiple Advantages (n2 IQ)
13
How to integrate disparate knowledge?
14
DOMAIN THINK COMMUNICATION SEMANTIC WEB
WORDS, MEANINGS, VALUES
15
Words Common Understanding Common Values
Unicode Accuracy Xlation accuracy / Culture
context is different, even if the language is
common there will be some discrepancy. Common
human context is key, and will likely be a focus
of future technology development. Common Values
Scott A. Carpenter, 20-Jan-2003-Mon-2311 pm
16
Carpenter Communication Paradigm (unicode must
understand words, meanings, values)
Person has a new idea. This person extrapolates
their goal list, and the idea appears to solve
their mission (feasibility). Person then expands
their model (generalization), and the idea
appears to be a feasible solution. The rules
are specifically described structurally that
appear to make the model work. Then the index
is expanded (vocabulary, unicode) and the data is
obtained from the environment. How are each of
these layers described by the person, by the
system? As the concepts work their way down into
atomic understanding, and then back up again in
a different language? The uncertainty in
understanding must grow! This implies that none
of us will be standing on the same layers of
understanding, as each of us will understand each
layer slightly differently than anyone else.
Fascinating. Note In the figure below, the
grapheme, syntax, and language might also affect
the index, rules, model
Same overall system, but in a different language,
and thus leads to possibly different
understanding certainly the fuzziness has grown.
A concise idea originates here, having a finite
expanse (interpretation, scope)
VISION
MindLang 2
MindLang 1
VISION
Mission
Mission
WISDOM
Lang 2
WISDOM
Lang 1
Goal
Goal
Lang 1
Lang 2
KNOWLEDGE
KNOWLEDGE
Model
Model
INFORMATION
INFORMATION
Syntax 2
Syntax 1
Rules
Rules
DATA
DATA
Grapheme 1
Grapheme 2
Index
Index
Environment
ENVIRONMENT (Complexity Mystery)
ENVIRONMENT (Complexity Mystery)
This is a paper! Scott A. Carpenter,
26-Oct-02-Sat-1154
17
How precise is even the smallest bounded concept?
Meaning in the Senders mind.
Meaning in, say 20 of, the local audiences mind.
Meaning in, say 40 of, the local audiences
mind, and say, 10 in the remote audiences mind,
and after translation, say 5 of those speaking a
different language.
18