Books

CDs

Events/Exhibits

Film/Video

By Faith D'Aluisio & Peter Menzel.
MIT Press, Cambridge, MA, U.S.A., 2000.
239 pp., illus.
ISBN: 0-262-13382-2
Reviewed by Robert Pepperell. University College Wales, Newport Caerleon Campus, Newport, NP18 3YH, U.K. E-mail: pepperell@cwcom.net


Good news for those fearing an imminent take over of the world by robots - according to the experts it's still going to happen, but not for another 50 to 100 years. One learns from Robosapiens, that all those predictions made in the 'sixties and 'seventies about how intelligent machines will be dressing us and serving our meals were over-optimistic (or over-pessimistic depending on your point-of-view). This book presents what appears to be a comprehensive account of the current state of robotics research in a snappy journalistic style made up largely from interviews with leading figures in the field conducted by Faith D'Aluisio with their work photographed by her partner Peter Menzel. The result is an information-rich coffee-table book that reads and looks like an extended WIRED article. Despite the earlier high expectations of Robotics we are now starting to realise the huge degree of complexity needed to replicate anything approaching human-like behaviour (or 'humanoid' as the terminology has it). Instead, research has branched out into smaller, more achievable, areas such as 'search and rescue' and surgical assistance. This is not to say there is an absence of exciting projects or people exploring deeply interesting ideas.

Of particular interest from a 'Posthuman' point of view was the move away from the strong-AI goal of modelling the human brain in order to replicate consciousness towards an understanding of consciousness as a brain-body phenomena. Such work is being conducted using a $1 million 'Dynamic Brain' robot at the Japanese ATR Centre just outside Tokyo under the direction of Stephan Shaal and Mitsuo Kawato. Equally interesting from a 'Postdigital' point of view is the seemingly awesome power of Mark Tilden's 'Unibug' made from cast-off electrical parts assembled for a couple of hundred dollars. The Unibug, almost uniquely amongst current robots, dispenses with digital processing and uses analogue feedback circuits which allow this little 'creature' to move about and learn. These units are highly efficient, very cheap and more reliable than many more expensive systems. However, what seems to unite much of the research documented here is US Defense funding since a large proportion of the projects described are directly or indirectly supported by DARPA. One can see in certain cases, like the remotely operated rifle, how the research might be immediately relevant, but this is less so in many others and one inevitably suspects covert motives in otherwise innocuous projects. I was left with the impression that whatever the outcomes of current thinking and developments in robotics, it is the ubiquity of defense funding which means that the armed forces that will get the first and greatest benefit, maybe making Karel Capek's fantasy nightmare even more probable.

top