Transcript The Choice Methodology: A New Foundation of

An Introduction to Artificial Life
The Choice Methodology: A New Foundation of Structured Machine Life
Christopher A. Tucker
Department of Adaptive Systems
Institute of Information Theory and Automation
Academy of Sciences, Czech Republic
[email protected]
http://www.paradoxtechnologies.org
1
Topics covered in this lecture:

A Hierarchy of Life Forms
• Cognition Hierarchy – Mimicked Forms
Under Consideration
• The Choice Complex

The Research Question
• Experience & Emergence
2
Topics covered in this lecture:

Three Theoretical Assumptions
• Research Paradigms

Experimental Methods
• Choice Hardware & Software
• The Queen-Drone Autonomous System
3
The Cognition Hierarchy
4
The Choice Complex

If a machine can make a choice and if it
did, would it be a meaningful one?

Empirical testing by embedding behavior
in architecture
5
The Research Question
If a machine can directly experience
entropy, is the consequence an
emergence of consciousness?
6
Theoretical Assumption #1:
If a machine can make a choice, it
portrays some level of consciousness
7
Theoretical Assumption #2:
Only if a machine mimics biological life
can it have access to those kinds of
experiences.
8
Theoretical Assumption #3:
If a machine has knowledge of its death,
it fosters emergent behavior.
9
In order to satisfy the research question, the machine…

Choice
• Determines Experience

Life “Conditioning” & Survival
Adaptations
• Entropy

Mimic
• Form & Function
10
Choice

Determines Experience

Fosters Behavior

Forecasts Evolution
11
Life “Conditioning” & Survival
Knows that it has “life” functions within a
set of states—its operational life—and
also knows the inverse.
 The ability to die fostering emergent
behavior in the form of survival
adaptations

12
Entropy & Death

Catalyzed by a unique circuit in the
machine mainframe

A special circuit attached to a generic
robotics platform
13
Mimic

The Mimic or Animae (synthetic animal)

Satisfies the condition of form

Satisfies the condition of function
14
An Artificial Life Prototype
The Experiment
Based on the theoretical principles
15
Experimentation




How are all these ideas put together in a reasonable
methodology whereby we can test some of the notions
illustrated?
How can we generate reproducible results and publish our
ideas?
How can we create a methodological framework to extend
to future versions and generations of machines?
How can we advance the knowledge of science?
16
Experimental Conditions
Clause:
The machine should be made to exist on its own with
as little interference from humans to be as a close
an approximation to life as possible.
Two Experimental Conditions
17
Experimental Conditions
There can be no qualitative comparison with human intelligence.
18
Experimental Conditions
No superposition of one intelligence system with another, all
objects in the system remain equal.
How?
19
Motif: Environment, Intelligence, and Behavior
“The properties commonly ascribed to any object are, in last analysis, names for its behavior.”
20
The Hardware Paradigm
21
Choice Hardware
22
The Wireless Power System
23
A Prototype Wireless Power Transmitter
24
Hardware Layout

Cooperating with the transmitting array,
there is a third card that is onboard in the
physical device
25
Note: Card less conversion circuitry
Method of Power Delivery

Two methods under consideration:
1)
CW at radio frequency
2)
Pulsed power at near-audio
frequency
26
CW – Radio Frequency

Tests already conducted and proof of
prototype accomplished (2004):
27
At a transmission carrier frequency (f0) of 78MHz
Pulsed Power

Currently under development for a
upcoming research grant sponsored
by the University of Reading…
28
Choice Hardware & Software
29
The Entropic Circuit
The Core of the Domain-Specific Robotics Platform
30
The Software Paradigm




Power Choice—A, B, C, high, mid, low;
signals exchanged between Tx and R′x
Coupled to 802.11x device
Power delivered relative to feedback
between power and load sensors
Relationship modeled in software
diagram…
31
Queen-Drone XML Layout (Software Controller)
32
What are the first experiments?





Test the existing wireless power system
Apply software controller (choice)
Refit a physical robot with the Entropic
Circuit
Deploy and test
Watch the magic happen!
33
Results of Physical Prototype

Fully autonomous animae

Adapted to its environment

Dynamic model of living systems
34
Project goals over the next 24 months:





Test the tenets of distribution of energy &
information in a cybernetic paradigm
Allow a continuous operation of the machine
A long-term study of behavior
The Entropic Circuit to test survival in artificial
forms
Understanding the feature of choice in living
systems
35
Goals already accomplished since inception of project
Software architecture deployed in
www.o2active.cz
–
–
–
–
Commercial application of the choice complex
Testing of transformation-based logic (Ashby 1957)
Includes properties illustrated in this lecture
Decentralized parallel processing with drone tasking
36
Questions…Comments…
A paper containing the information in this lecture is slated
for publication in early in 2007.
In the meantime…
http://www.paradoxtechnologies.org
Paradox Technologies
You Dream It, We Build It
http://www.paradoxtechnologies.org
37
Thank you…
38
A special thank you to UTIA-AS and Dr. Kárný