British Telecom Industry News

A New Generation of Intelligent Computers

The Alvey Program is spearheading Britain’s drive to develop a new generation of ‘intelligent’ computers. Jane Bird talks to its director and describes an example of the research it encourages.

timeline_computers_1959_stretch

Brian Oakley has been picking teams ever since his days as a physics undergraduate at Oxford when he directed engineers in student theatricals. It is the skill at casting that he reckons is his greatest asset as director of the Alvey Program, which is intended to keep Britain in the race to build the so-called ‘fifth generation’ of intelligent computers.

‘They could have chosen a director with more technical, expertise. But it’s picking the right people that counts – there are some really unbalanced individuals in artificial intelligence’, he says.

Despite his modesty, Oakley has impeccable credentials in the key fields that Alvey combines – scientific research, industry and government. It was set up over two years ago after a committee headed by John Alvey, BT’s development supreme, recommended that Britain should join the race.

Oakley acquired a taste for pushing research into commercial products while working on micro-waves and computer applications at the Ministry of Defense’s heavily-funded Royal Signals and Radar Establishment. Not surprisingly, civil applications received no encouragement there.

But Oakley pursued his mission to keep scientists and industrialists collaborating, when he became secretary of the Scientific and Engineering Research Council in 1978.

Now, of all his achievements with the Program – less than half way through its five-year span, and 85% through its pounds 200m budget – it is the newly forged links between industry and academe that he is most proud of.

‘That is our most obvious success, and to a large extent it is irreversible’, he says.

The links are controversial. ‘Some people think I’m stifling the seed corn of new ideas in academe by forcing researchers to concentrate on commercial products’, he admits.

But he argues that teaching gets very arid if not constantly refreshed with a breeze from the outside world. ‘Innovation involves technology, marketing, finance and psychology. It’s very exciting, but often teachers are too remote to put that across.’

computer-internet-100128-02

Oakley is frank, even stubborn, and not shy of telling fellow committee members what he thinks.

At times he even seems to be deriving some mischievous fun from watching the opposition wriggle. Recently he responded to Japan’s repeated invitations to take a British high-tech delegation to Tokyo.

‘They wanted us to bring our academics, so I took the businessmen instead,’ he says, winking.

But Oakley is a great fan of Japan’s fifth generation computer project with its consensus on technological objectives. It was a compliment to him to be accused of copying the Japanese consensus approach.

If he could have his time again, Oakley would spend it cracking the complexities of how humans speak and understand language. He is fascinated by the challenges that elude us, despite our ability to build voracious number-crunching computers.

‘Speech recognition and language translation are tasks which we have always underestimated. In the early 1950s people thought all we had to do was load the rules of grammar and some vocabulary into a computer and out would come automatic translation.

‘It’s been a terrible disappointment – but the discovery of the extraordinary complexity of language has spurred on research into lots of technologies.’

Oakley is a hard taskmaster – short notice meetings have to be squeezed in over breakfast. Says a rueful colleague, ‘the trouble is that, in the nicest possible way, he expects all of us to be as dedicated.’

He has had some flak. The most widespread criticism is that he has starved small firms and individual British innovators by allocating 75% of the funding to the industry giants – GEC, STC, Plessey and Ferranti.

But Oakley argues that these are the companies with the staff and laboratories to accommodate research. So far he has approved 102 joint projects involving 60 firms, 40 universities and 15 polytechnics.

The Alvey directorate is a slimeline affair which Oakley keeps deliberately at minimum staffing on the basis that fewer people can keep in touch better with what each is doing.

Already 58, his leadership of a ‘son-of-Alvey’ program is unlikely.

But a successor who so skilfully combines the qualities of diplomat and shrewd high-tech missionary will be hard to find.