What is CyberLife?

By Steve Grand, former Technology Director at CyberLife Technology.

CyberLife is Creature Labs' proprietary A-Life technology based on the application of biological metaphors to software-complexity problems. As software becomes increasingly complex we start to face problems of how to manage and understand the systems we build. However, the levels of complexity of these systems are trivial in comparison to those of even the most modest biological systems. Why then with all our genius, logic, and organisational abilities do we find it so difficult to build complex systems? After years of research it seems the reason and the problem all lie within the way we think of and approach complex systems.

From mechanistic to systemic

Traditionally, science was all about breaking down systems into their constituent parts. These parts would then be analysed to reveal their structure and the functions they perform. This was the prominent endeavour of the 19th century, and was very useful as a method of gaining understanding about many things including simple biology, medicine and physics. During the 20th century, our endeavours focused on building systems, from the industrial revolution through to the digital revolution. However, somewhere along the way we had a paradigm shift and decided that the way to build or model complex systems was to consider the behavior required and try to capture this in high level constructs. Massive rule bases were developed in order to capture the intelligence and subtlety of human and animal behavior. Needless to say, these systems failed.

Adaptive life

The route of the problem seems to be that the abstracted knowledge has no grounding - there is no actual physical meaning to any of the concepts. Therefore, if the programmer of the system had not considered a possible situation, then the response of the system may turn out to be erratic, wrong or non-existent. Natural systems are rarely this brittle. All animals learn from experience and generalise. An animal will never be in the exact same situation twice, however it has the innate ability to reason about the similarities between its current situation and those it has experienced in the past. The animal will then usually perform some action that was profitable to it in the similar situations of its past. If this is a bad thing for the animal to do, it will learn from its mistakes and try out some other behavior if faced with a similar situation in the future. Why then don't we base our artificial systems on biological systems?

Modeled systems

Well, that is exactly what we are doing with CyberLife. If we want a system that behaves like a small creature, then we build a small creature. We model large numbers of cells in the brain (neurones), and connect them up and send signals between them, in a way similar to natural cells. We model blood-streams and chemical reactions. We model a world for the creature to inhabit, and objects for the creature to interact with. Finally we model diseases, hunger, emotions, needs and the ability for the creature to grow, breed and evolve. Only then do you get a system that behaves like a creature.

The first results form this philosophy can be seen in Creatures. Take a look, interact with them. Decide for yourself.

A short history of A-life

We believe that true intelligence is an emergent property of lifelike systems. Conventional approaches to artificial intelligence do not lead to true intelligence, just "smartness." This is because they attempt to create intelligent behaviour without regard to the structures that give rise to such behaviour in the real world (i.e., organisms). In the space of all possible machines, there may be many regions that show intelligence, but we only know where one of those regions lies - the region occupied by living creatures. Approaching AI without regard to Biology is just thrashing around in the dark.

CyberLife approaches the problem through simulation. We argue that certain types of simulation can become instances of the thing being simulated. By simulating suitable brain-like structures, we create brains, and (given suitable inputs and outputs) those brains will be intelligent and have minds of their own. By simulating biological organisms in the correct way, we create biological organisms. Artificial intelligence is not achieved by trying to simulate intelligent behaviour, but by simulating populations of dumb objects, whose aggregate behaviour emerges as intelligent.

Origins of Artificial Life

Though A-Life and artificial intelligence approach a common problem from radically divergent perspectives, historically they are closely related, both evolving from the work and research of Alan Turing and John Von Neumann. Turing's wartime effort cracking code in 1930's Britain initiated computer science in general and set off a wealth of scientific and philosophical discussions into the viability of a thinking machine. The Turing machine is a theoretically defined computing system with an infinite tape, capable of performing any possible computation.

Von Neumann's design for a digital computer in the 1940's was inspired in part by his research into computational neuroscience and theoretical research into cellular automata and self-reproducing systems. Studying the "logic" of reproduction, Von Neumann defined a universal replicator, a computational system capable of reproducing any system and realised that the system must function as both instruction and as data. He also remarked that errors in copying self-description could lead to evolution, which could be studied computationally.

Despite this pioneering research, computer science neglected A-Life for many years, focussing instead on AI and cybernetics. Computational evolution eventually developed once genetic algorithms were formally defined by John Holland in the 1960s. Still, the field of A-Life had to wait until the late 1980's to achieve unity and visibility.

In September of 1987, the first workshop on Artificial Life was held at the Los Alamos National Laboratory in the United States. One hundred and sixty computer scientists, biologists, anthropologists and other researchers were brought together and the term A-Life was officially coined. The organiser of the conference, Christopher Langton, there presented a paper which is now largely regarded as the manifesto defining A-Life's agenda.

CyberLife philosophy

Around four billion years ago, chemistry learned the art of co-operation, and as a consequence life began on this planet. Since then, the combination of random mutation and non-random selection known as evolution has pressed us onwards and upwards: ever more complex; ever more adaptable. Over the aeons, evolution gave many of us more and more powerful brains, with which we gained ever increasing control over our destiny.

A mere few thousand years ago, some of us (those who call ourselves Humans) began to understand ourselves and our world well enough to start to interfere with that process of evolution. First came agriculture, where deliberate selective breeding led to life-forms that would otherwise never have existed, such as wheat or the dairy cow. After this, the development of scientific reasoning led to a greater understanding of ourselves as machines. This in turn accelerated the technology of medicine, whose power to overturn the random accidents of evolution has now all but stopped our own natural selection in its tracks.

Very recently, through our understanding of the theory of machines, we have begun to comprehend and be able to manipulate life at a very profound level indeed. We are now ready to return to the Garden of Eden, whence we came. However, this time we will not be mere produce of the garden, but gardeners ourselves. Human knowledge has brought us to the verge of being able to create life-forms of our own.

Humans have been able to generate more humans for a very long time (and a good deal of fun has been had in the process). Never before, however, have we been in a position to create life to our own design. Scientists are already able to alter the genetic structure of existing simple organisms such as bacteria, in order to produce 'designer' life forms. We can make apples stay fresh longer and breed giant strawberries. In the future it is probable that they will be able to synthesise whole organisms from basic chemicals, creating life where there was none before. However, this is not only a long way off, it is also, in a sense, the least profound way in which we wil be able to create life. Synthetic life of this kind is merely 'life in our own image', yet carbon chemistry is only one of the ways that life can exist.

'Thinking different'

We are beginning to realise, the more we study the attributes of life, that life isn't so much a property of matter itself (so that you can only generate living things from carbon chain chemistry), but that it is a property of the organisation of matter. Much of science is currently undergoing something of a revolution in its thinking, and it seems that one consequence of this shift will be the genesis of other classes of living things, whose minds, if not whose entire bodies, lie within the memory of a digital computer and eventually, perhaps, collections of networked computers.

With CyberLife, we are making our first tentative steps towards a new form of life on this planet. Sitting in a tank, on the very PC with which these words were written, are two small and rather stupid creatures. One is called Eve, and the other, for reasons best left to another paper, is called Ron. They are not highly intelligent, and they hardly ever do as they're told (just like children). Yet they are quite easy to become attached to, and hopefully they will have many generations of descendants, throughout the world. Some of these offspring, or their cousins, may learn to do useful jobs for people or simply to keep people entertained until the day comes when we know how to create truly intelligent, conscious artificial beings. It is also hoped that those conscious beings will find a place in their hearts for the memory of Ron and Eve.

"By the middle of this century, mankind had acquired the power to extinguish life on Earth. By the middle of the next century, he will be able to create it. Of the two, it is hard to say which places the larger burden of responsibility on our shoulders."

- Dr. Christopher G. Langton

Embedded life

Once upon a time, all machines were integrated with living things. Every plough was pulled by oxen and guided by a man; every lathe turned by hand and controlled by eye. The Industrial Revolution removed the need for muscle power, and the progress of automation has reduced our reliance on human supervision for the control of machines. However, many jobs cannot be done, or are done badly, without a living organism at the helm: a tractor can pull a larger plough than a team of oxen can, but unlike the oxen it cannot refuel itself or navigate rough terrain without a human brain to guide it. CyberLife is concerned with the re-vivification of technology. Through CyberLife we are putting the soul back into lifeless machines - not the souls of slaves, but willing spirits, who actually enjoy the tasks they are set and reward themselves for being successful.

CyberLife is thus the art of creating and embedding living things into machines, either in software or hardware. However, underlying CyberLife is a set of more fundamental principles, themselves quite far-reaching, that need to be understood by all participants if the CyberLife promise is to be fulfilled.

Underlying ideology

"Cyberspace" could be defined as "the location at which two people meet when they are engaged in a telephone conversation." This definition expands to encompass all networked communication quite easily, and the concept of the internet as a 'container' of cyberspace, or a channel 'into' cyberspace is now commonly understood throughout the world. Since the emergence of the notion of "virtual reality," cyberspace has become a broader and more powerful concept, representing a world 'inside' a computer, and people are now beginning to be able to walk around cyberspace, see it, manipulate it, meet people in it, and shoot them.

The idea that a computer is a container and life-support system for cyberspace has begun to have profound implications for programming methodology. Originally, computers were designed to be fast calculating machines (hence the name), and were even used to compute such abstract operations as Calculus, despite that fact that Newton (and Leibniz) only invented the Calculus because they didn't have computers to iterate their approximations for them! Once computing got into its stride, however, the notion of 'Procedural Computation' took hold, and computers became machines for expressing algorithms, rather than merely solving equations.

Now, expressions describe relationships and algorithms describe processes, and both of these tools very quickly became used to make computer models of real-life systems, for example to forecast the weather or compute the behaviour of bridges in a wind. The expressions and the algorithms in such models directly describe the behaviour of those things (air masses or bridges); they don't necessarily, however, describe the things themselves. The recent adoption of Object-Oriented Programming turns that view around, and considers a computer to be a device for modelling things, each of which has certain properties, and it is the relationship between those properties which then describe the behaviour of the system. The modern programming paradigm is therefore one in which the programmer constructs objects, which reside in the cyberspace within a computer: A computer is a machine which contains cyberspace, which contains machines.

Object-orientation is rapidly becoming popular, at least partly because certain advantages such as portability, reusability and robustness come along with it. However, old habits die hard, and many programmers who lap up these undoubtedly valuable side-effects still fail to appreciate the profundity of the concept of a computer as a machine which contains cyberspace which contains machines. Yet profound it certainly is.

Emergent behaviour

The tools which have made science so powerful over the last three thousand years are largely analytical and reductionist: if you want to study something very complex, take it to pieces first, then study the pieces; if the behaviour is too messy, develop a simplified model, without the messy bits, then study that; if the proper equations cannot be solved, invent some simpler ones that can. Nobody would doubt that this reductionist approach has been very successful. However, many people are equally aware that there are fundamental flaws in this kind of reasoning: everyone knows that sometimes "the whole is greater than the sum of its parts". If you study only the parts, then you are missing something crucial about the whole. This 'something' is known as 'emergent behaviour.'

As a trivial example, 'soccer' is an emergent property, which is vested in no single soccer player. Only when a whole team is together can soccer exist. Similarly, the mind exists because of the interaction of the billions of neurones in its brain, yet no single neurone could be said to contain the mind, nor even any part of it. Only when the assembly is acting as a co-ordinated whole does the mind exist. Moreover, the existence of the mind would come as a complete surprise to anyone who was only given one of its neurones to study in isolation.

Not only is the whole more than the sum of its parts, but in general the whole cannot even be predicted by a study of the parts alone. Therefore, science has been missing a great deal by its reductionist approach. What is more, it has missed out on much of what we (as more than the sum of our parts) actually find interesting in the world. One of those things is the secret of life itself.

Many hands make light work

Once you accept the idea of a computer as a container of virtual objects, you have available a plurality that isn't there when you think of a computer as a processing machine: a computer is one machine, yet it can contain many virtual machines. There are some very important things that many machines can do together that one machine can't do alone.

Think of an ant, for example. A marvel of complexity and sophistication it may be, but no ant is smart enough to design, memorise or communicate the plan for an ant nest. There is no master architect ant, who stands there in a hard hat and red braces instructing the other ants on where to build, yet an ant nest is a very complex and organised structure.

Think of a raindrop. Examine it. Construct differential equations to describe its behaviour. Do you see Niagara in that raindrop? Would you predict such a thing from it? Yet Niagara is only a bunch of raindrops acting in concert.

Think of your childhood. You remember it clearly, don't you? Yet you weren't there! Not a single atom that's in your body now was present when you were a child. You're not even the same shape as you were then. No thing has remained constant, yet you are still the same person. Whatever you are, you are not the stuff of which you are made; yet without that stuff, you would not be anything at all. Material flows from place to place, and momentarily comes together to be you. If that doesn't make the hair stand up on the back of your neck, read it again.

These examples all have one thing in common, and the implications that can be drawn from that are varied and deep. The common feature is that each example shows a unified structure or process that exists only because many small things, with relatively simple properties, come together in one place and interact with each other. The principle that Niagara cannot be inferred by looking at a raindrop is known as emergence, and the school of thought that attempts to capitalise on emergence by building large edifices from many small building-blocks is known as bottom-up.

Starting from scratch

One very practical consequence of bottom-up thinking to a programmer is how it helps with the management of complexity. Imagine trying to define an adventure game in terms of a decision tree - this is the top-down approach. Suppose you progress through the game by taking decisions, always from a choice of two. The structure to describe this would be a binary tree: the first decision could be taken in two ways; each of those choices leads to another decision, giving you four different routes through the tree. Add a few more levels of decision-making and you have a tree with 2n leaves (32 decision steps = 4,294,967,296 possible routes). This in itself is not an especially intractable problem, even if routes can double-back, or many choices are available. However, imagine that we add a computer player, and that every decision he takes affects the choices you have available. How many nodes do we need on our tree now? Well, it depends on how the interactions are implemented, but the decision tree is not twice as big as before, it is many billions of times as big. Now add ten more computer players and stir?

Biological models

The great thing about living creatures is that they are general solutions to problems. A squid may solve a dramatically different set of problems than a mole does, but the methods of solution are only variants, not fundamentally different from each other. Throw a mole into an ocean and it will not swim like a squid, but both types of creature share a common ancestor, and thus each is an example of how that ancestor solved a different problem. The adaptable, building-block nature of living things and the fact that they rarely need to reinvent the wheel (because they inherit their solutions from their parents) makes organisms capable of solving a huge range of environmental tasks.

Living creatures generally solve only problems that are of interest to themselves: evolution adapts them to new ecological niches where they can thrive unmolested; brains help them to solve the problems associated with getting food, finding a mate and so on. This is fine for them, but not a lot of use to us. However, if we were able to create life-forms to our own design, we would be able to select the problems that we wanted them to solve. This has been the failed goal of Artificial Intelligence research. Unfortunately, AI has devoted decades to looking at the task from the wrong end - from the top-down. The real answer lies in emulating Nature's way and creating life from the bottom-up.

Every human being starts out life as a single cell. That cell divides into two almost, but not quite, similar cells. Each cell switches on slightly different genes before dividing again. Eventually you have one hundred billion cells, performing many thousands of distinct tasks, each doing its own little job without being overseen by any 'master cell.' Each contributing blindly to the working of the whole, to the emergence of us. Therein lies a very powerful idea: general-purpose building blocks, whose behaviour is controlled by data (genes and the local environment), and which interact locally to produce behaviour from the whole system that could not be predicted from, is not resident in and is not controlled by any single one of those parts.

Summary

• Think of a computer as a container for cyberspace; think of programming as the creation of machines that populate that cyberspace, rather than a recipe for the overall behaviour of the system.

• Swarms of simple objects interacting with each other have more power and subtlety than a single top-down structure can provide, they also minimise combinatorial explosions.

• Complexity cannot be forced; it must be nudged and cajoled into existence. Getting the right dynamics is as much an art as a science.

• Life is a complex network of feedback loops that allows a system to hover on the brink of chaos, a regime which is capable of sustaining computation.

• The best, if not the only, way to create living systems is to create models of the building blocks from which existing life-forms are constructed.

(copyright 2000 CyberLife Technology LTD)

Note: Steve Grand is no longer a part of Creature Labs nor participates in the development of the Creature series though continues research and development in the field of A-Life along with his wife, Ann, at Cyberlife Research in Shipham, Somerset, England. A description of Steve's latest book, Creation: Life and how to Make it, can be found at Harvard University Press.