The Supermen by Charles Murray, Wiley,ÂŁ19.99, ISBN 0 471 04885 2

SEYMOUR CRAY was a man who burnt his boats behind him. So writes Charles Murray in The Supermen. A key anecdote relates how Cray spent a summer building a boat, then sent it up in flames after he purchased a new one: a perfect illustration of Cray’s craving for a fresh start.

Murray, a senior editor of Design News, has written a lively history of Cray and his contemporaries. These men founded a string of start-ups, and pioneered the fields of computing and supercomputing, long before the days when venture capitalists would bankroll anyone who could say “dot com”. The story deftly interweaves social backdrop with individual drama, corporate manoeuvring with technical innovation, beginning with the birth of computers during the Second World War, and ends with the death of Cray last year. Next to the biography are set the stories of the managers who marketed his inventions, and the engineers with whom he collaborated and competed. It is a tale of market ups and downs, of successful and failed designs, and of a series of machines that generated over a billion dollars in sales.

Cray’s policy of starting afresh applied equally to computer architecture and corporate culture. As soon as he was sure one design would work, Cray left the finishing touches to others, and went to work on a radically different design for the next generation. And as soon as one corporation grew too large and conservative for his innovations, he would discard it and found another. When faced with a stagnating design, Cray liked nothing better than to found a new laboratory in a new town, leaving him with no distractions and no excuses for failure.

The success of the early computer firms was fostered not by venture capitalists, but by the Pentagon. One early Cray computer was designed to guide intercontinental ballistic missiles, while others found their first application in designing nuclear bombs. Computing, more than physics, became the crucial edge in the arms race. Cray chose to live and work in a remote location, in part to get away from corporate management, and in part to be far away from the bombs if they ever fell. One question not answered is whether Cray and the others believed that Mutual Assured Destruction was vital to national security, or whether it simply provided an exciting opportunity to build the world’s fastest computer.

Regardless, Cray’s life was devoted to design. He turned down a nomination for the National Medal for Science (and an almost certain visit to the White House) just so that he could devote more time to his next design. He worked long hours at home, and installed a full set of design tools (including a mainframe computer and backup power supply) in his holiday cottage. But he was not a workaholic: he played golf weekly and was devoted to skiing and windsurfing, and frequented dark restaurants with leather upholstery. He was quick to fire managers, but he cut his own salary in preference to laying off engineers. His bosses pandered to the press with stories of Cray’s eccentric genius—but this ploy backfired when Cray chose to go his own way.

Murray’s book is a paean to the creative process. You read of engineers so devoted that they beg to be allowed to finish their computer designs when laid off, even at no pay, and of fiendishly compact three-dimensional stacks of circuits, which generate so much heat that refrigerators must be built to cool them. The book’s main shortcoming is technical: Murray bandies about terms like “vector pipeline” and “multiprocessor” without ever explaining them. He worships the engineer’s creativity, but he never explains their insights into the core question of how to organise computation.

Despite leaving me hankering after more technical details and a leaning towards hagiography, Murray spins a story I found difficult to put down. This twin history of business acumen and design genius may dwell more on stock prices than on computer architecture, but it is clear its heart belongs more to the engineers than the managers. This story of the search for the perfect creative environment is a Shangri-la tale for today.

Rawlins, associate professor of computer science at Indiana University,
Bloomington, foresees a stark choice—either computers will protect our
privacy with codes not even a government can break, or they will destroy it by
leaving our banking and legal histories online for all to see. He then envisions
how electronic publishing will enhance our lives, and weighs the new modes of
communication enabled by the Internet against the likelihood that, much like
television, it will be corrupted. The book goes on to detail how computers will
replace us in war and the workplace, and to recall a few famous software
failures. It concludes that computers will ultimately succeed humans in the
struggle of evolution.

Rawlins is ambiguous as to whether this outcome is to be dreaded or desired,
but he is adamant that our fate is as inevitable as that of the eponymous moths.
His theme is that economic incentives—or greed—will lure us to hand
over more and more power to computers, until humankind itself becomes
obsolete.

Moths to the Flame is salted with speculative scenarios and peppered
with pithy phrases. But it is surprisingly short on science. Rawlins spins tales
about the future of artificial intelligence, but says nothing of its past
successes and failures. Nor does he touch upon the use of mathematics to vastly
improve the way we build software.

If you want to fire your imagination with charming stories of where computers
may go, then look here. But if you want to temper your imagination with
knowledge of where they’ve been and where they’re at, you’ll have to look
elsewhere.