From Abstract Machines to Agents: An engineers perspective

1
From Abstract Machines to Agents An engineers
perspective

• Hürevren Kiliç
• Department of Computer Engineering
• Atilim University

2
Content

• The Question
• Indicators for its answer
• Intermediate Conclusions
• Today Agents, Environments, Objects, and
  Human-being
• Conclusions
• Problems to be solved

3
The Question

• Mathematics is a game played according to certain
  simple rules with meaningless marks on paper. (D.
  Hilbert) 1.
• Engineering is
• 1 the application of science and mathematics by
  which the properties of matter and the source of
  energy in nature are made useful to people,
• 2 the design and manufacture of complex products
  (Merriam-Webster dictionary) 2.
• Related phrases mathematical model,
  mathematical proof, mathematical
  representation, mathematical formulation,
  mathematical equation etc.
• A classical response Anything can be a
  mathematical model so long as it is a
  well-defined set of rules for playing with some
  abstract constructs. But the question is A
  mathematical model of what ? (D.Cohen) 3.

4
The Question

• Mathematics, only a tool for engineering ?
• Not the question of time/space representation in
  mathematics but the existence of mathematics in
  time/space (possibly virtual !).
• Symbols in virtual space !!! Engineerability of
  intangibles !!!
• Any evidence/indicator for this ?

5
Indicator 1

• An analogy between Constructive Proofs and
  Engineering Effort
• NB -- No computer algorithm can perform the task
  of determining whether a mathematical statement
  is true or false 4 (K.Gödel, A.Turing,
  A.Church)

6
Indicator 2

• Mathematician Lets write this mathematical
  conclusion (e.g. Lemma, equation etc.) down here
  on this page, later we will turn this back !!!.
• Idea Storable mathematical entities ? Stored
  program concept ? (by J.Von Neumann)
• Mobility for them ? (between memories, server,
  client, peer environments ...)
• Codable/Storable/Interpretable/Movable/Executable
  mathematical entities ?

7
Indicator 3

• Operators vs operands --- Data vs methods
  ---Capabilities/Resources
• Input / Output y F(x)

F(...)
x
y
Transductional Behavior
ENGINEERING SYSTEM
output
input
ENVIRONMENT
8
Indicator 4

• Three necessary and sufficient things to do/make
  mathematics
• Paper
• Pencil
• Mathematician (Anon ?)

the mathematician
pencil
paper
9
Indicator 4 cntd.

• An abstract machine with universal computation
  capability called Standard Turing Machine.
• Definition (Standard Turing Machine) 4
• Q finite set of machine states,
• S finite set of input alphabet, not containing
  the special blank ? symbol,
• G the tape alphabet where, ? G and S
  G
• d Q x G ? Q x G x L, R is the state
  transition function,
• q0 Q is the start state of the machine,
• qaccept Q is the accept state,
• qreject Q is the reject state, where qaccept
  ? qreject

10
Indicator 4 cntd.

• Turing machine vs
  Turing

the mathematician
Control Unit
Read/Write Head
.......
.......
pencil
Tape
?
Today
Today
paper
Today
Today
11
Intermediate Conclusions

• Mathematical statements are mathematical
  entities.
• Proving them (at least for constructive ones)
  is analogous effort to engineering system design
  (or specifically program writing). How about
  inductive and by contradiction type of proofs
  ?
• They are codable/interpretable/storable/movable/ex
  ecutable.
• They can be realized on electron-based hardware
  environments, on programming language/operating
  system like software environments or
  alternatively biological, chemical, nano-scale
  quantum environments.
• Mathematical entities living 7/24 on the internet
  at run-time !!!

12
Evolution of Abstractions on Mathematical
Entities

• abstract machine descriptions (Universal Computer
  !!!)
• storable machine description
• machine code
• assembly language
• machine-independent programming languages
• sub-routines
• procedures functions
• abstract data types
• objects
• to agents.

13
What is agent ?

• Definition by M.Wooldridge An agent is a
  computer system that is capable of independent
  action on behalf of its user or owner (figuring
  out what needs to be done to satisfy design
  objectives, rather than constantly being told)
  5
• S.Russell and P.Norvig definition for rational
  agent For each possible percept sequence, a
  rational agent should select an action that is
  expected to maximize its performance measure,
  given the evidence provided by the percept
  sequence and whatever built-in knowledge the
  agent has. 6

14
What is environment ?

• Basic characteristics of the environment 5
• ubiquity
• interconnection
• intelligence
• delegation and
• human-orientation

15
Objects vs Agents 5

• Are agents just objects by another name? (dynamic
  objects ?)
• Object
• encapsulates some state
• communicates via message passing
• has methods, corresponding to operations that may
  be performed on this state

16
Objects vs Agents 5

• Main differences
• agents are autonomousagents embody stronger
  notion of autonomy than objects, and in
  particular, they decide for themselves whether or
  not to perform an action on request from another
  agent
• agents are smartcapable of flexible (reactive,
  pro-active, social) behavior, and the standard
  object model has nothing to say about such types
  of behavior
• agents are activea multi-agent system is
  inherently multi-threaded, in that each agent is
  assumed to have at least one thread of active
  control

17
Human-being vs Agents

• Agents do it because they want to ?
• Objects do it for free, agents do it for money
  ???
• A natural consequence of pumped rationality !!!
• Rationality vs bounded-rationality !!!
• Human vs agent
• (Heart Mind centric) vs (Mind centric)
• Delegation of authority,
• Agents should behave as their cooperating owners,
• Interactions between human and agent (Turing
  test).

18
Turing Test (by Alan Turing in 1950s)
?
19
Human-being vs Agents

• Danger If you are not able to construct
  human-like agents then try to manipulate
  human-being to behave like rational agents.
• The trust problem between the agents (Solvable ?)
• Family metaphor against organization metaphor
  ?

20
Related Issues

• Single agent vs multi-agent.
• Environmental setup through protocol design
• Cooperative vs competitive environments
  described by interaction protocols.
• Equation-based modeling vs Agent Based Modeling
  (ABM)

21
Conclusions

• Triggered by mathematics, the fields of computer
  science and engineering provided interaction
  among mathematical entities and their mobility
  over interpretation based virtual environment
  setups.
• Mathematics not only to understand the matters
  properties and sources of energy but also to
  understand the living mathematics that is
  realized and become recognizable - over the
  internet-like environments as a consequence of a
  complex historical engineering effort by means of
  game theory, logic, complex systems, dynamical
  systems theory and more from other disciplines.

22
Agent Problems to be Solved

• Rationality of agents.
• Existence of an alternative metaphor to
  organization (Legacy computing systems)
• Establishment of trust among entities (for
  human-being, lack of common reference !!!).
• Social problems due to the delegation of rights
  to agents for making decisions/negotiation on
  behalf of their human owners.

23
References

• 1 Quoted in N Rose Mathematical Maxims and
  Minims (Raleigh N C 1988).
• 2 http//www.merriam-webster.com
• 3 D.Cohen, Introduction to Computer Theory,
  Wiley, 2nd Ed., 1997.
• 4 M.Sipser, Introduction to the Theory of
  Computation, PWS Publishing Company, 1997.
• 5 M.Wooldridge, Introduction to Multiagent
  Systems, John Wiley Sons, 2002.
• 6 S.Russell and P.Norvig, Artificial
  Intelligence A Modern Approach, Pearson
  Education, 2003.

24

• ! Thanks !
• ? Questions ?