artificial life kiran lakkaraju. outline i.what is a-life ii.the early beginnings of a-life....

27
Artificial Life Kiran Lakkaraju

Upload: pearl-cummings

Post on 19-Jan-2016

217 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

Artificial Life

Kiran Lakkaraju

Page 2: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

Outline

I. What is A-Life

II. The early beginnings of A-Life.

III. Cellular Automata

IV. Current Focus

GA’s, Evolutionary Programming, Emergent Programming.

V. Differences between AI & AL

Page 3: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

What is Artificial Life

• “collection of methods for building discrete event simulations with evolving multiple agents”

• study of the dynamics of living systems, regardless of substrate

– Biology – Study of CARBON-based life.

• Artificial life (A-Life) uses informational concepts and computer modeling to study life in general, and terrestrial life in particular. (Freeman quoting Langton)

• “form of mathematical biology”

Life is a tale told by an idiot, full of sound and fury, signifying nothing.

W. Shakespeare

Page 4: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

A Definition…

• Artificial Life, as defined by Chris Langton, one of its founders, is a field of study devoted to understanding life

by attempting to abstract the fundamental dynamical principles underlying biological phenomena, and

recreating these dynamics in other physical media, such as computers, making them accessible to new kinds of

experimental manipulation and testing

Life is a foreign language; all men mispronounce it.

Christopher Morley (1890 - 1957)

Page 5: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

A Philosophy

• Abstraction

"There's nothing implicit about the material of anything - if you can capture its logical organisation in some other medium you can have that same 'machine', because it's the organisation that constitutes the machine, not the stuff it's made of..." - Chris Langton

• Look at principles of life without reference to the implementation.

Page 6: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

Methodology

• Alife – synthetic. It tries to achieve complexity from the combination of simpler structures

• Biology – analytic. Breakdown of Complex phenomena to basic components

"We are no longer the knights who say 'Ni.' We are now the knights who say 'Ikki Ikki Ikki Ikki, P'tang Zoopdah blohlfdsjfdslkjn. (Ni!)" -Monty Python's The Quest for the Holy Grail"

Page 7: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

History

• As with most things in AI, the early results were with Turing and Von Neumann

• Theoretical Accounts of Self-Organization

Page 8: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

Turing

• Univeral Turing Machine

• Can a human be represented as a Turing Machine?

Page 9: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

John Von Neumann

• Abstract requirements for Self-Replication

• Universal Constructor – using CA’s– Interpreted code and Uninterpreted data

• Copy Errors Evolution!

Self-reliance is the only road to true freedom, and being one's own person is its ultimate reward.

Patricia Sampson

Page 10: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

Goals

• Turing’s Universal Automata• Von Neumann: Universal self-reproductive

automaton.• Logical – Not Physical.• Never became functional – too complex• Codd 1968 successful, but still very complex.• Langton succeeded 1986 – restarted the field.

Page 11: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

Cellular Automata

• Formal Definition: “A cellular automaton is a lattice of sites, each of which can take on k values. Each site of the automaton is updated at discrete time-steps by a finite state automaton residing at each site and assigned a value depending on the value of the sites around it.” (Adami 1998)

),,( )(1

)()(1

)1( ti

ti

ti

ti aaaa

If something is so complicated that you can't explain it in 10 seconds, then it's probably not worth knowing anyway. Calvin & Hobbes

A rule:

Page 12: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

Informal Definition

• We have a n-Dimensional array where at each time step, the value of a certain entry depends upon the values of its neighbors in the previous time step.

Page 13: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

Example

• Dimensionality: 1D

• States: _ and 1

• Neighborhood: r=1, look at two neighbors to the left and right of you N C N

Page 14: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

Time 0 : _ _ _ _ _ _ _ _ _ _ 1 _ _ _ _ _ _ _ _ _ _

Time 1 : _ _ _ _ _ _ _ _ _ 1 1 1 _ _ _ _ _ _ _ _ _

Time 2 : _ _ _ _ _ _ _ _ 1 _ _ _ 1 _ _ _ _ _ _ _ _

Time 3 : _ _ _ _ _ _ _ 1 1 1 _ 1 1 1 _ _ _ _ _ _ _

Time 4 : _ _ _ _ _ _ 1 _ _ _ 1 _ _ _ 1 _ _ _ _ _ _

Time 5 : _ _ _ _ _ 1 1 1 _ 1 1 1 _ 1 1 1 _ _ _ _ _

Time 6 : _ _ _ _ 1 _ _ _ 1 _ _ _ 1 _ _ _ 1 _ _ _ _

_ _ _ -> _ 1 _ _ -> 1

_ _ 1 -> 1 1 _ 1 -> 1

_ 1 _ -> 1 1 1 _ -> _

_ 1 1 -> _ 1 1 1 -> _

Page 15: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

1D CA Examples

Copyright David G. Green 1993.

Page 16: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

Conway’s Game of Life

• 2D CA.• 2 state, 9-neighbor • Rules:

– A living cell (state “1”) with 2 or 3 neighbors remains alive

– A dead cell (state “0”) with exactly 3 neighbors is born

– All other cells die (from loneliness or overcrowding), or else remain dead.

Page 17: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

Conway’s Game of Life cont.

Why do I get the feeling that someday I'll be describing this to a psychiatrist? The Simpsons

Page 18: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

Von Neumann, Langton

• Von Neumann able to construct a highly complex CA entity

– 29 States

– > 150,000 cells

– Self-Replication

• Langton (1986) restarted the field by constructing a simple one.

• Why important?

– Self Replication important

– Evolution

– Life logically implemented

Page 19: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

CA Animation

Page 20: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

Main Tenet

• Self-Organization– Spontaneous emergence and maintenance, of

order– Autonomous, intrinsic development.– Higher-level properties resulting from

interactions between simpler properties.

Page 21: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

Turing Again.

• “The chemical basis of morphogenesis” (1952)

• Showed how interacting chemical diffusion gradients could produce higher-level structures from initially homogeneous tissue (MITECS, Artificial Life)

Page 22: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

Application

• Simple rules complex behavior model used in many simulations:– Flocking– Electric Power Grid– Model Traders in the Commodity market.– CA implementation of partial differential

equations– CA model for freeway traffic.

Aah! Hey, get off my sugar. Bad bees! Bad! Ow. Oww! Oh, they're defending themselves somehow!Homer Simpson

Page 23: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

Evolution

• Follows the general principles of Evolution.

• Have populations of partial solutions that are selectively recombined and mutated in the hopes of creating a better solution.

• Genetic Algorithms, Evolutionary Programming

“…Let no one suppose that evolution will ever exempt us from struggles. 'You forget,' said the Devil, with a chuckle, 'that I have been evolving too.‘”

William Ralph Inge

Page 24: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

Applications

• Genetic Programming – Koza– Population of Programs– Apply the evolutionary operators (crossover,

mutation) to the program– Create better programs through this process.

Page 25: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

Simulations

• CoreWorld

• Tierra – Programs that can self-replicate existing in a simulated world with limited resources– Parasites, hyper-parasites

• avida, amoeba

Page 26: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

AI vs. AL

1. Complex human functions

2. Top-down3. Global Control4. Complex Rules5. Preprogrammed

behavior6. Knowledge-

Representation

1. Basic natural behavior

2. Bottom-up3. Local Control4. Simple Rules5. Emergent behavior6. Avoid Knowledge-

Representation & Planning

I used to be with 'it', but then they changed what 'it' was. Now what I'm with isn't 'it' anymore and what's 'it' seems weird and scary.

Abraham Simpson

Page 27: Artificial Life Kiran Lakkaraju. Outline I.What is A-Life II.The early beginnings of A-Life. III.Cellular Automata IV.Current Focus GA’s, Evolutionary

References

• Explorations in Artificial Life (special issue of AI Expert), pgs. 4-8, September, 1995. Millef Freeman. An Introduction to Artificial Life

• MIT Encyclopedia of Cognitive Sciences: Artificial Life, M.A. Boden.

• Introduction to Artificial Life, Christopher Adami, pgs. 27-28. Springer-Verlag, New York.

• “Introduction & overview of ‘Artificial Life’- Evolving intelligent agents for modeling & simulation” A. Martin Wildberger, Proceedings of the 1996 Winter Simulation Conference (ACM)

• Artificial Life: The Quest for a New Creation, Steven Levy,1992 Pantheon Books.