Amongst my ordinary batch of email messages on the 3rd January I had a slightly unusual one from someone called simply 'Fred'. It was a short note to let me know that he had left the 'CyberPort' and was now roaming the 'TechnoSphere' in search of food and, er, sexual partners. For me this message was welcome news - in fact, it was something I had been waiting for a while. I had created Fred a few weeks ago and had been getting a little worried about his silence. Why had been there been no communication in this time? Had the set of body-parts I'd chosen for him failed to function in some way? Or maybe there had been an environmental catastrophe in the TechnoSphere, leading to a mass extermination of its inhabitants, including my poor nascent Fred? However, I needn't have worried - it seemed that the truth was far less dramatic. By the looks of it the administrators of TechnoSphere had been dealing with a large backlog of requests and therefore Fred's birthday had been delayed a little. Still, as the email message stated, he was now in existence - apparently alive and well - and I could look forward to many future messages as he explored his particular part of the Internet...

Alight, you've guessed it, Fred was not your normal email user. In fact Fred was, and still is, not human at all but is instead an Internet 'beastie' who I had designed when I visited the TechnoSphere Web site at [no more]. I interactively chose his eyes, body shape, legs (actually, he's got wheels!) and mouth-parts via the Net and then set him loose to hopefully grow, feed and reproduce in the TechnoSphere environment. Perhaps unsurprisingly the 'world' Fred (17498 to his friends) lives in, exists only inside a computer. It's an ever-changing mathematically-modelled Universe that stretches a (notional) 16km in each direction and is made up half of grass-covered open plain and half of craggy fractal mountains. Within this artificial environment the computer ensures that the grass 'grows' daily, providing food for the herds of herbivorous beasties who graze on the plain, who in turn are eaten by various - more solitary - carnivores inhabitants. Life in the TechnoSphere, so the creators say, is simple - revolving around eating and the avoidance of being eaten, sleeping and, of course, procreating.

Life in the TechnoSphere is also rather fast-moving when compared to the million-year timescales adopted by living things as they evolve in the 'real-world' around us. The rules-of-life in TechnoSphere combine Darwinian ideas of 'survival of the fittest' with a creationist world-view to produce a unique virtual environment that allows anybody who has access to the Net to quickly create a new lifeform and then watch it evolve over weeks rather than years. In a God-like way you can design a new beastie (like my Fred), place it to the TechnoSphere environment and then sit back and wait as it emails you regular news of its activities - giving you an insight into TechnoSphere's archetypal world of life and death.

You Can Call me AL

The TechnoSphere project is a very interesting (and literal) example of what has come to be known as an 'artificial life' system. Artificial life (or simply 'alife' or 'AL') is a well-established area of computing which has recently started to leave the research-laboratories and enter the public arena. Its broad aim is to successfully apply the processes adopted by biological organisms to the computing domain. The belief is that by using biological techniques such as 'self replication', evolution, cross-fertilisation and mutation when designing software we will not only be able to create computer programs that better achieve the tasks required of them (be they sorting programs, robot control systems or implementations of mathematical algorithms) but we will also gain a better understanding of the processes of life itself.

As the field of alife has developed a wide range of biological terms and techniques have entered the computing field. For example, the idea of 'genetic' programming has been widely explored by Computer Scientists. This technique involves solving a computing problem not by writing a single program, but instead by creating a population of partial solutions to the problem which can then breed by exchanging programming instructions. As the programs reproduce, an ever-better solution to the problem is generated until eventually the optimum program appears - a program that was, in some sense, created by the computer itself.

As well as alife experiments and genetic programming, the field artificial life also encompasses another application area - that of the simulation of real living systems. A good example of this can be seen on the Alife Page at http://www.fusebox.com/cb/alife.html, where, as well as non-real 'life' demonstrations, there are simulations of natural behaviours such as swarming and even forest fires. Many of you may also be aware that systems such as these were used to control the movement of the hunted dinosaurs in one of the famous T-Rex scenes in the film Jurassic Park. A case of an alife environment being used to simulate an extinct 'real' lifeform in a work of fiction!

The 'environments' inhabited by genetic programs, 'beasties' such as those in the TechnoSphere and simulated real lifeforms are normally created inside a single computer. Each 'alifeform' has a description stored within the computer environment that is updated as each unit of time passes inside the virtual world - with each update to the description being based upon some set of pre-defined rules. However, with the growth of the Internet a much larger space has become available for colonisation by would-be alife creatures. And while it may be some time before beasties such as Fred leave their limited worlds and begin to search the Net for e-food and data-swapping partners, some life-like forms are already out there and occasionally cause us humans major problems...

Viruses, Worms and Agents

I am, of course, referring to the most feared of e-creatures, the 'computer virus' and its cousin the 'worm'. You may not initially think of computer viruses as being in any sense 'alive' but they do, however, have many features in common with biological lifeforms. Firstly, they have the ability to self-replicate or reproduce - the basic defining characteristic of life. Secondly, they aim to ensure their survival by maximising their population in a given environment - the more copies of themselves that exist, the 'happier' they are. And finally, some of the more advanced viruses are able to adapt to their world by 'mutation' and the exchange of instructions with other viruses.

Normally, viruses such as this spread around the Net in one of two ways. Firstly, they may 'hitch a ride' on the back of some legitimate program. You may download the program, run it, and then discover that you have let the virus loose within your computer. The virus will then 'infect' other files and spread to any other computer that you share disks or programs with. Secondly, the virus may take a more active role in reproducing itself. An example of this is the famous 'Internet worm' that almost brought the Internet to its knees in 1988 by exploiting a UNIX email bug. The program ran on one computer, copied itself to other computers via the email system, ran again, copied itself and so on. This elifeform had self-replication down to a fine art, in fact it was so good that its creator was prosecuted by the US government and almost ended up in prison!

However, before you start panicking about the possibility of your computer being infected by such a virus, it should be noted that viruses are actually relatively rare on the Internet. Most FTP archives screen against them before uploading a piece of software and the combination of a virus checker and reasonable security precautions will normally protect your computer from this kind of 'close encounter'. What's more, we should not forget that some of the techniques used by viruses and worms may not be necessarily all bad. One area of computing research is looking at creating benevolent elifeforms that reproduce like computer viruses and explore networks such as the Internet, returning information to their creators. Such programs, or Agents, could reproduce themselves and monitor the Net on your behalf - letting you know when they have found something that they 'think' you will be interested in. Wouldn't this take the strain out of Net-surfing!

The Art-ifical Life of William Latham

Another positive way of using elife techniques is in the creation of images and 3D structures. For example, artist William Latham uses the processes of mutation and selection to 'grow' his highly elaborate 3D artworks from sets of simple forms. Latham's work - which can alternately look like impossible sea-creatures, strange alien lifeforms, or delicate porcelain-like sculptures - relies heavily on the ability of the computer to simulate real-life behaviours and to create non-real virtual environments. When he is producing artwork he acts somewhat like a 'master gardener', deciding which 'art-ificial lifeforms' should be given a chance to grow further and which should be weeded out. From a common starting point he is able to grow all manner of forms to almost any level of complexity.

While much of Latham's artwork has been based around single computers, he too is interested in the possibilities presented by the Internet. When we chatted recently he talked of friendly art viruses that might one day be set free on the Net, spreading images from computer-to-computer, cross-fertilising and mutating to create populations of Latham-esque imagery! Also, he could imagine a distributed 'forest' of organic artworks - spread across the Internet - growing and evolving, almost as an ironic comment on the state of 'real' forests in our modern world.

If you want to see more of William Latham's artwork then point you Web browser at his Web site at [no more] or http://www.nemeton.com/axis-mutatis/latham.html. Alternatively, you will soon be able to have a go growing your own 3D images with the release of his Organic Art screensaver package.

The Next Generation

So, we have moved from artificial animals in the TechnoSphere, through to genetic programming, viruses, worms to organic art. The common thread through all of these subjects is that they use what we might call the 'rules-of-life' in some way. These techniques - clearly borrowed from biology - are producing computer systems and data structures that are becoming more and more life-like in the way they operate. The question arises, therefore, when - if ever - will we actually be able to say that a genuinely new 'lifeform' inhabits our computer or even the Internet? Well, the answer to this question is more down to definitions than hard-and-fast rules. While a computer virus may satisfy many of the criteria a biologist would use when describing life, very few of us would think of such a piece of software as ever being alive. Still, as viruses become more advanced (through the programming skill of their often destructive developers) and more benevolent alife experiments such as TechnoSphere expand, we may one-day have to seriously reconsider this view.

Perhaps the forces of life are even more prevalent than we think. After all, it may just be that life on Earth, with its never-ending search for increased diversity, has found a way into CyberSpace...

[ Features index ]

©1995 Sean Clark.