From Computer Business Review, a sister publication.
Peter Cochrane and fellow researcher Chris Winter were probably hoping for big headlines. After all, it is not every week that scientists from a prestigious research center announce the possibility of new life forms, a revolution in at least two industries, and hint at the downfall of one of the world’s most powerful companies: Microsoft. But then the last week in February was no ordinary week. Just a few hundreds miles to the North, fellow British scientists had announced to the world that they had successfully managed to clone an adult sheep. While the media attempted to explore and explain the possibilities for humankind of those DNA experiments in Scotland, they largely missed out on the opportunity for some equally sensational reporting on the bewildering but beguiling research work into artificial life being carried out at the British Telecommunications Plc laboratories based in Martlesham, UK.
By Andrew Lawrence
In a low-key press conference to announce the progress the $65m research project has made so far, the center’s research head Peter Cochrane casually forecast that in a few years time, a computer would not only be able to build its own software on the fly but that it would be able to construct its own circuitry using field programmable gate arrays. You say [to your PC] ‘I would like you to write a new word processor and it does it… instead of buying it from Bill Gates or anybody else. Within ten years, forecast Chris Winters, today’s graphical interfaces would be dead and the first biological machines would appear. In 20, today’s dominant computer ‘Von Neumann’ computer architecture would only be used for accountancy. But this was the more mundane stuff. Cochrane and Winter, who have just patented an algorithm called ‘Virtual Children’, happily speculated on the possibility – even probability – of creating autonomous life-forms. This raises the possibility that such a life form might not be friendly. When we start to build these things, we may have to program something which makes them look upon human beings kindly said Cochrane, adding We worry a lot about being smart enough. We’re carbon-based. We might not be smart enough to spot silicon life and all the implications. How will all this come about? It is all about mimicking the natural evolution in software: after all, nature has built systems – often very simple ones – that are capable of handling some very, very complex tasks. That is how British Telecom got involved in the first place, attempting to manage very complex telecommunications networks with very simple algorithms. But surely nature took millions of years to build its systems? True, but British Telecom has been asking – and answering – a deep and fundamental question: why, in nature, are there just two sexes? Why not four, or six? These are questions that the DNA researchers in Scotland are also interested in: they want to know why not just have one sex?