|
Articles The Computer Bulletin Interview The visionary head of BT Laboratories tells Mike Hewitt his thoughts on the future of computing and communications technology and its impact on society 'I don't think you can predict the future - none of us can - but what we can do is build it,' he says. 'That's our job here. My 660 staff are among the best in the world at doing this. I've effectively advanced them into various stages of that future. As head of research and development I'm probably the farthest ahead, at about 10 years. From this perspective, I see a world of instant communication, massive bandwidth and awesome computing power. And it's my aim to make BT the leader in this future, technologically, managerially and operationally.' His criterion for success is quite simple: 'Can I fall in love via the technology? I'd say that, currently, you can't over the Internet, phones, or videoconferencing. Not by themselves, anyway. You need to press the flesh, too. But when we advance to the stage where the communications technology becomes as efficient as tactile or emotional contact, we'll have succeeded. That's the way I want to go, and I can already see things moving that way.' Peter Cochrane has worked for BT since the early 1960s, when it was part of the General Post Office, and he has seen it all: System X, fibre optic cabling, satellite communications, ISDN, ASDL..... He is well placed to predict the shape of tomorrow's communications and global communications infrastructure. And to say what it definitely will not be, too: the Internet, which he describes as 'a parasite'. 'The telephone network was originally dimensioned for people who on average would make three or four four-minute phone calls a day. That gives a peak-to-mean ratio of around 4:1. But the Internet, which has simply been tagged on top of our existing 100-year-old phone network, makes the ratio nearer 1,000:1. Ergo, you need a different kind of network and a different approach to the technology to cope. If not, by 2003, when it's predicted that Internet traffic will overtake telephone traffic, the parasite might kill the host by swamping it.' That 'different kind of network' will have broadband optical fibre as its backbone, he says, although peripheral networks such as radio, optical wireless, and existing systems using old copper cables will hang off it. 'It's essential that we don't build an information footpath or cart track. We've got to start with a vision of continual network growth and expansion for at least 30 years ahead. This requires single-mode optical fibre everywhere. Optical fibre that amplifies and maintains its bandwidth over hundreds of miles is now an engineering reality. And we need the minimum of electronic switching and routing.' What sort of information will be travelling over this network? 'I think in terms of computers being everywhere and in everything. For instance, a heart pacemaker will be in continuous contact with a hospital. Our clothes will have chips inside them. There won't be any need for credit cards or cash: a reader will capture my bank account details from a chip embedded in my skin and automatically debit my account.' All this will mean fundamental changes to society. 'People talk about an information society,' Peter Cochrane says. 'What we'll actually create will be an information economy: a world dominated not by the exchange of commodities but by the transfer of bits and bytes. 'What we must not do, however, is repeat mistakes made in the past. For example, when society changed from hunter/gather to farmer, the gatherers got poorer. With the industrial revolution, the industrialists got richer while the farmers became poorer. To ensure everyone benefits in the coming information society and there are no losers, we have to ensure that everyone acquires the requisite IT skills.' This means man-machine interfaces will have to become a lot more straightforward: 'Until very recently, computer interfaces were appallingly difficult. Perhaps only 20% of the population could cope with them. We've got to reach a stage where everyone can.' Even so, Professor Cochrane is dismissive of front-ends that try to mimic the real world: 'I get really irritated by things like virtual libraries and supermarkets where you "wander around" shelves. It doesn't work. My brain knows that in a real supermarket I have to transport my body 20 feet to reach the cornflakes. In the real world my brain is tolerant of my body's limitations and will wait. But in a virtual world my brain knows that to actually get what it wants all it needs do is point and click. It wants instant gratification.' Hence BT's current research into speech recognition, improved man-machine interfaces, and 'intelligent' machines. It is Peter Cochrane's contention that in the coming information economy there will be so much information to process that human beings alone will not be able to cope. They will require computers to do at least some of the thinking and decision making on their behalf, if only to tell them not to buy a certain brand of washing powder. 'We desperately need a third intelligence: a future of man, woman and machine,' Professor Cochrane says. 'Machines can do things we cannot, and will therefore play a complementary role, giving us information about the information flow. But to get to this stage, they must be able to interpret the world as we do, which means "seeing" it as we do. Current computers are suffering sensory deprivation because they don't. Only when we give them a rich sensory system will they acquire true intelligence. I see this happening within the next 20 years.' To this end the BT laboratories have recently taught a computer to recognise human faces and tell the difference between a man's and a woman's. 'We fed thousands upon thousands of photographs of men and women into a neural network and told it which was which. Then we took a photograph at random and asked, "What's this?" It usually gets the right answer - but not the same way we do. The traits we look for are very subtle: the hardness of the male features compared to the female's, for instance. But the computer does it by recognising whiskers, lipstick and mascara.' So how would it cope with Eddie Izzard or Lillie Savage? 'It makes the same sort of errors we might make, except perhaps a bit more often.' The computers that will eventually weed out the cross-dressers will be incredibly powerful. 'If we look forward, computers are going to be 1,000 times more powerful than they are now in the next decade, and about a million times more after that,' Peter Cochrane says. 'But the decade after that is going to be particularly interesting. The size of devices will of course come down, and they will be somewhat less digital.' Is this future computer-enhanced information society going to be a unified one, or will it be subject to the same divisions as today's world? Peter Cochrane, at heart, is very much of the 'one world' persuasion. His favourite poem, If, probably sums up his philosophy: yes, there are always obstacles, but if you never give up, you will eventually succeed. 'Everywhere I've travelled, I've found people to be more or less the same: the same hopes, loves, fears, and desires,' he says. 'What separates us is simply ignorance. I believe the communications revolution will banish that ignorance and help the world community become more cohesive. Already, I think World War Three has been forestalled by our ability to communicate. 'So what lies ahead for us all has the prospect of being something really wild and wonderful. And I hope I can play a valuable part in bringing it about. 'I don't want to have on my tombstone, "He came and he consumed." I want, "He came, he worked hard, and he made a positive contribution. He made a difference."' |