Artificial Life Experiments
(or is that digital death?)

What is A-Life? Simply put, Artificial Life is a broad field of work in which attempts are made to simulate or recreate one or more natural processes. Most A-Life work is done on computers, but in some cases it's done in the real world with chemicals or just on paper. Some examples of natural systems that are often modelled are: predator/prey balances, darwinistic evolution, neural networks, behaviour and learning, ecological interactions, flocking, self organization, emergent behaviour or complexity, reproduction, and genetics, to name a few. A-Life's practical purpose can be two-fold, to use technology to understand and improve biological sciences, and to use biological understanding to improve technological science. But beyond the practical, insights into life, nature, and the universe can be gleamed even from the most simple model, experiment, or simulation.

me Ed T. Toton III

  • System Administrator / Programmer
  • Long-time A-Life enthusiast
  • Self-Organized Emergent Philosophical Process :)

Since the summer before I entered college, artificial life (or "A-Life") has been one of my interests. In general, I find all of the "fringe" computer topics interesting, including fractals, AI, digital music, 3D graphics, etc. But A-Life holds a certain appeal. There's something particular compelling about the concept of having life exist within the computer, something that evolves and develops it's own behaviour or morphology, something that goes beyond the initial design created by the programmer.

Artificial life takes on many forms, and different people approach it from different viewpoints and with different goals. For some people, it is a scientific tool, to help understand the natural processes in the natural world. Darwinistic evolution and predator/prey balances are two examples of such processes that can be simulated and observed in a controlled environment. In some cases, it's used to study emergent behaviour, or emergent complexity from a set of simple rules (a trait that fractals also exhibit).

To others, it's more of an artistic endeavor. Some A-Life projects have been developed to be a means of genetically generating pleasing patterns and designs (Biomorphs have been used for this). L-Systems can be used to model the external morphology and appearance of plants. And in some cases (such as some of my programs), they're done for the sake of creating something that takes on "a life of its own", and they can be interesting to watch.

And for others, a-life is for entertainment. Virtual pets and 'games' in which you take an active role in the development of the organisms, or populations of organisms can fall into this category (such as SimAnt, Creatures, Tamagotchi, Dogz/Catz, etc).

Despite the wide-range of practical applications (or even impractical applications), it should not be overlooked that A-Life is indeed a legitimate field of study, and serious scientific research is being done in the field. One that is only recently receiving wider attention, perticularly with the continuing increase in available computer power.

Many A-Life experiments fall into one category, and typically only model one or two specific processes or traits of individual organisms or populations of organisms. Often they can also be described in terms of other properties of the simulation- For instance, some simulations simulate the interactions between multiple individuals that are created on a high-level (such as BugFest), wheras others attempt to create a world with its own rules on a lower level, in which the organisms are constructed from components of that world (such as Tom Ray's Tierra or Helix-4).

While my personal interest doesn't specifically exclude any particular form of A-Life, my efforts typically fall within the entertainment/artistic range, since most of my projects are more for my own enjoyment, and for sharing the wonder that such things can inspire in people, though I do attempt to make them valid experiments in emergent behaviour, genetic evolution, or other natural processes. I've made a variety of types of A-Life programs, only the better of which are available here (meaning they're aesthetically acceptible, relatively bug-free, fairly bullet-proof, and interesting on some level).

Although I tend to release programs of this nature very infrequently, A-Life is the one field of interest that I can never seem to stray very far from. Most of what I've made so far has been fairly simplistic in many ways (though some of the technical hurdles to keep program performance high were not insignificant), but I have ideas floating in my head all the time. There will undoubtedly be more life springing forth from this website sooner or later. :-)


Click here for the old A-Life comments

My A-Life Experiments:

  • AT Robots 2 - Not really an A-Life project, but related and very interesting. This is a game in which you write a program to control a robot, and pit it against other robots in a simulated arena. I've had an online robot server running for several years, where you can test your robots against the current champions.

  • Unnamed Project - This is an A-Life project written completely in perl, only calling out to one ImageMagick program to convert the graphical portion of the output to a browser-compatabile image format. It's an online simulation, and the CGI script that generates the page you view predicts the weather patterns independantly (capable of forecasting with 100% accuracy to an arbitrary distance into the future). It's the classic "Living vs Undead" struggle, competing for control of territory, running as a cellular automata model with a multi-sinusoidal weather generator.

  • BugFest-1 - A graphically aesthetic predator/prey balance simulation using insects.

  • Helix-4 - This was an attempt at a Tierra-like A-Life project. Self replicating programs compete for space in the memory pool. The program/interface is a little clunky, the instruction set is a little fragile, and I'd do things differently now, but the program worked well and even included a genebanker and creature viewer.

  • Helix-1 - My very first attempt at Artificial Life. It doesn't really demonstrate anything, except how darwinism can improve efficiency and nothing more when there's no possible open-ended evolution. The interface is nice though. :)

A-Life Links:

Back to NecroBones Headquarters

(Copyright 1998, All Rights Reserved)