Mark Ward

VIRTUAL ORGANISMS
Pan Books 1999


WardVirtOrg8


The continunm of living organisms stretches from viruses, which are little more than a stretch of DNA or RNA surrounded by a protein sheath, up to planetary ecosystems like Earth at the other. All of them use the same principles to - briefly - defy the first and second laws of thermodynamics and the attempts of the larger Universe to return them to the dust from whence they came.

Artificial Life research encompasses software simulations, robotics, protein electronics and even attempts to recreate the Earth's first living organisms. It is less concerned with what something is built of than with how it lives. It is concerned with dynamics and just how life keeps going. All you, I and everything else need is information.

Life Rules

The second law of thermodynamics says that the Universe is winding down. It is gradually becoming more and more disordered. Energy in useful forms is slowly being dissipated from the places where it is used to do work out into the air or space in the form of heat where it can do no one any good. This disorder is measured by a concept called entropy and it is slowly increasing.

Systems that possess order - you, me, whirlwinds and whippoorwills - have a low entropy. We are suffered to remain in our living state because we are so good at creating disorder. We consume energy in a variety of forms but we excrete it largely as low temperature, high entropy heat.

We pay our debts to the second law, which strictly says we should not exist because there is a net decrease in entropy in a living system, by generating far more entropy and disorder than the small amount of order needed to keep our bodies functioning. This is the con trick played on the Universe by every living thing and so far we have not been caught out. With a bit of luck it will be a long time before we are. As scientist James Lovelock says: 'If your excretion of entropy is as large or larger than your internal generation of entropy, you will continue to live and remain a miraculous, improbable, but still legal avoidance of the second law of the Universe.'

Information is the key to maintaining our inner peace and avoiding a descent into entropy. Built into all of our genes is a wonderfnl mechanism for preserving information -deoxyribonucleic acid (DNA). Genes have been characterized as selfish, but what they hold tightest of all is information about the body they are travelling aronnd in. DNA and its little helper RNA are remarkable molecules. Both are strands made up of subunits called nucleotides. Attached to each nucleotide is one of four different bases. DNA has guanine, cytosine, adenine and thymine, RNA has the first three but uracil in place of thymine. A triplet of bases is known as a codon and, via DNA's faithful servant RNA, specifies the instructions for making one of 20 different amino acids.

Life keeps going using this information. The relationship between information and entropy was first discovered by Claude Shannon, an engineer who worked at Bell Telephone Laboratories in New Jersey during the 1940s and 50s. Shannon was investigating what prevents information being transmitted across a channel or telephone line. He found that fault lay with a hard-to-define quality that always seemed to be increasing whenever information was lost. Shannon never witnessed a decrease in this quality in all the experiments he performed. Acting on the advice of mathematician John von Neumann, Shannon called this slippery quality entropy.

Life is all about ensuring information is passed on, or transmitted, while all the time preventing entropy from corrupting the message. Life has found a way to ensure that entropy keeps increasing but not at the expense of its own survival or the integrity of the information it wants to transmit. Evidence of this can be found if we take another look at amino acids. The more abundant an amino acid is the more care is taken to ensure the information is read and preserved correctly. Six different codons prodoce leucine perhaps because it is so abundant. This is the biggest and best trick that life has learned.

Now ALife is helping us to understand just how it does it. It is starting to show that living things do not rely on the properties of chemicals to foster another generation, they depend on information encoded in chemical form. ALife is showing just how the dynamics of information can come to dominate over the properties of the materials living things are made of. ALife research has revealed that this shift occurs when a system is acting chaotically. Not chaotically in terms of 'badly organized' but chaos in the mathematical sense.

In the real world stability seems to reign. The seasons come and go, our hearts keep on beating and the taps keep dripping but in truth nothing is as safe or secure as it first appears. Living things cannot be captured by arid equations, they are too messy for that. Intuitively we feel that we have more in common with clouds than clocks and chaos helps explain why.

N-BODY PROBLEM

Predicting the movement and interaction of two bodies as they orbit the Sun is relatively straightforward. Newton, comprehensively solved the mathematics of this problem. You could be forgiven for thinking the maths would not get much more complicated if a third body were added. In fact the addition of another object creates an intractable computation known as the n-body problem.

The mutual interaction of the three bodies around a common gravitational centre, the Sun in the case of our Solar System, makes their exact movements impossible to predict. Only in the last decade have researchers been able to use chaos mathematics to confirm that this is the case. The implication is that the same is true of any system where multiple forces and influences are at work.

This interplay of obiects and forces plays two roles. It keeps a system unpredictable, constantly threatening to upset the apparent equilihrium and plunge it into chaos, but at the same time the push-me-pull-you play of forces keeps it on the straight and narrow.

In complicated systems, such as dripping taps, air flows over wings and weather patterns, so many factors coincide that predicting their outcome over anything but the short term is, to all intents and purposes, impossible.

The best that can be done is map out all possible outcomes and watch the subject move around the territory to get a feeling for what forces are at work and how they are weighed against each other. These 3-dimensional behavioural maps can be plotted on computer and are said to exist in phase space. This is an abstract concept more often used by physicists. You can think of it as a place where anything can happen. The constraints on what does happen are decided by what it is that you are modelling. If you want to model the movement of a pendolum, the dimensions of your phase space will be velocity, position and friction. The behaviour of the pendulum can be plotted at any moment as these forces act on it and it slowly loses momentum and comes to rest.

Plotting this behaviour in phase space can be revealing. It shows the relationships between the forces acting on a system and any patterns in that behaviour. Edward Lorenz, one of the pioneers of chaos theory, gave the name of 'the Lorenz attractor' to one of the patterns he discovered in phase space. The pattern resembled the face of an owl or a butterfly with its wings open. Other phase space patterns resemble closed fists with sharp drops abutting flat planes that, perhaps, represent a period of stability in the development of that system. Others are spiky like medieval weapons with sharp precipices between peaks.

What is becoming apparent is that life has to exist on a knife edge. Observation of fish populations and ant colonies, as well as experiments with artificial creatures, have shown that too much chaos produces a system that never gets a chance to settle down and develop. But if there is too little stimulation, everything stagnates. Between these extremes is a narrow, fertile region where just enough disturbance gets through to keep a heart beating healthily or a population thriving. Life needs a threat to keep it sharp. Entropy, or too much disorder, is one threat but the environment throws up others in the form of predators and climate changes.

Life as a dynamic process

This conception of life as a dynamic process - rather than an inherent property of the bits and pieces that living organisms are made of - has only recently begun to prevail. Prior to this change in perception life and living systems were thought by ecologists and biologists to be working towards a static state where environment and organism work in harmony. Now it is known that feedback and conflict is the norm, with species constantly wrestling for the upper hand and managing to survive and develop together as a result.

The approach of using biological metaphors and computer models of chaos to try and improve our understanding of natural and artificial systems is being used everywhere. Rescarchers at the Bionomics Institute in San Rafael, California, are starting to look at national economies as organisms or ecologies and have been surprised at how usefnl the analysis has been. Companies are starting to be compared to living, dynamical systems too and again the insights are piling up.

This method works because of the realization that the essential properties of life can be abstracted away from the bustling flora and fauna we find on Earth. In constructing computer models of living things we lose nothing because life is all about the preservation of an abstract property - information. Provided the behaviour that a computer model or a robot exhibits is lifelike then it is legitimate to use it to draw conclusions about the animals scuttling aronnd on the planet. ALife pioneer Chris Langton says:

The 'artificial' in Artificial Life refers to the component parts, not the emergent processes. If the component parts are implemented correctly, the processes they support are genuine - every bit as gennine as the natural processes they imitate. The big claim is that a properly organized set of artificial primitives carrying out the same functional roles as the biomolecules in natural living systems will support a process that will be 'alive' in the same way that natural organisms are alive. Artificial life will therefore be genuine life - it will simply be made of different stuff than the life that has evolved here on Earth.'

In many ways any artificially created life is much more useful than any of the real living creatures that are running around getting on with their lives. It is impossible to make a rabbit or beagle experience something for the first time again and again. They are too good at adapting to be fooled more than once. Their memory will help them cope. The same cannot be said about ALife creations; the memories of these creatures can be wiped time and again and their reactions recorded as if they were encountering an obstacle, a predator or a person for the first time.

ALife also helps us say more meaningful things about life in general. Biologists are doing a sterling job of capturing many of the properties of the creatures on Earth but it is hard for them to claim that what they are finding out has any universal relevance. We have no knowledge of life on other planets and how it goes about its daily life to compare with how ants, anteaters and antbears spend their days.

ALife can help by creating new forms of life and watching what they do. In this way we can start to draw more meaningful conclusions about life because our theories draw on novel examples. ALife is as much about life-as-it-could-be as life-as-we-know-it. But before we can create new beings that share the properties of the life on Earth we need to know a lot more about current conceptions of living things - what they do and how they learned to do it. We need to draw up a checklist of the life all around us so we can be sure that our own novelties are doing the right things. To do that we have to begin where life started - at the bottom.












HOME      BOE     SAL     TEXTE