Henry Holt & Company
Reinventing Man is a journey into the state of the art in robotics as well as an exploration of the possibilities of ‘reinventing’ human intelligence in machines. Although the modern robot owes its existence to a progression of new technologies – among them, computers, cybernetics, silicon memory chips – more than this is required if we are to create a true thinking robot: human biology, psychology, and even philosophy must come into play. Roboticists now at work on the problem of artificial intelligence are trying to bridge the gap between mind and body. If they succeed, they are likely to shed light on one of mankind’s greatest mysteries: the nature of the mind itself.
What are the prospects of creating machine intelligence comparable to our own? While it is a sophisticated piece of machinery, the present-day robot in large part has lacked two very important ‘human’ qualities: intelligence and perception. Until recently, the robot has been unable to see and react to its environment – to learn – the same way a human being does.
Technological breakthroughs made in the past few years are changing all of this. Reinventing Man offers a look into a future that is already here, by introducing WISARD, probably the closest anyone has come to date to creating a thinking, seeing, sensing, self-programming learning machine. WISARD is a sharp break from the traditional idea of programmed computers functioning according to sequential circuits. It is, instead, made up of unprogrammed silicon memory chips, randomly arranged like neural networks in the brain. Here is a machine that, unlike humans and unlike program computers, can actually learn from experience and by induction.
WISARD consistently outperforms the programmed computer in tests such as the recognition of human faces, where it accurately discriminates between different faces, even with changes in light or facial expression or disguises like false moustaches and eyeglasses. The future is indeed here, and WISARD opens up extraordinary possibilities not only for the study and development of robotics, but for what is always likely to be our central concern: the sources in the human system of memory, intelligence, and learning.
A Professor of Electronics at Brunel University and head of the British Cybernetics Society since 1980, Igor Aleksander led the team that developed WISARD. In 1962 he was the recipient of the Charles Babbage Award for ‘outstanding contributions to the understanding of computers’, and in 1982 he was the Rank Prize Lecturer at the Royal Society on the topic of ‘Practical Vision Systems ‘ for robots. Piers Burnett is a writer and editor with a long-standing interest in making complex scientific material accessible to the general public.