Dream Machine: Exploring the Computer Age

Gregory V. Wilson

It no longer seems strange to me to write book reviews on a computer small enough to balance on my knee. Every once in a while, though, when I mention to students that I did my first programming with punched cards, and see the disbelief in their eyes, I am reminded how fast and how far computing has come.

"The Dream Machine" a new documentary series jointly produced by the BBC and WGBH-TV in Boston, tells the story of that journey. From the mechanical calculators built during the Enlightenment by Pascal and Liebniz, to the Victorian Gothic of Babbage's Analytic Engine and the fragile complexity of Colossus, the code-breaking machine constructed at Bletchley Park during the Second World War, the development of computers rarely touched everyday life. But by the 1950s, electronic computers were sufficiently cheap and reliable to be adopted by business and industry. Twenty years later, the invention of integrated circuits brought them into the home; today, children are as comfortable with drawing packages and Nintendo games as children of my generation were with television.

This story has been told so often that one must wonder what is to be gained from another re-telling, either on television or in this "book of the series". One answer is that any story well told is worth telling, and the authors of this book tell their story well. Their tone is informal, but not inexact; the introduce the technical terms they need where and as they need them, and use them un-self-consciously thereafter.

A better reason for doing this book is that it is not primarily about computers, but about who developed computers and why. The US Census Bureau twice played a crucial role, more out of desperation than design. Later, President Kennedy's decision to put men on the Moon gave the young integrated circuit industry a much-needed boost (as did the Minuteman missile programme, something which is given much less prominence). Apple Computers could not have become what it was without the technical genius of Steves Jobs and Wozniak, but neither could it have succeeded without the business acumen and marketing skills of Mikes Markkula and Scott. Technical wizards like Seymour Cray receive little or no mention; primarily, the authors are concerned with the background to the developments which have had the greatest impact on our daily lives.

"The Dream Machine" does, however, have two serious flaws. The first is its design: the stark, double-spaced typesetting is bad enough, but the enlarged sans-serif quotations are dreadful. (Although this is in one way a sign of the times: before laser printers and bit-mapped screens, how many programmers had the opportunity to be typographical critics?)

The second flaw is more serious. While the authors occasionally comment on business practices --- mentioning, for example, that IBM became the dominant force in computing not by building better computers, but by doing better marketing, and that TWA routinely monitors the productivity of its ticketing agents to ensure that they do not spend too much time at lunch or in the toilet --- they shy away from mentioning many of the less savoury aspects of the industry. Of Alan Turing's death, they say only that "Turing died in 1954... his death officially ruled as suicide." There is no mention of his homosexuality, or of the hormones he was forced to take to "treat" his "condition". Similarly, while the unprecedented growth of Silicon Valley is described, the dark side of living there --- suicide, broken families, drug abuse, and frequent exposure to the toxic chemicals used to produce silicon chips --- is not.

They are less reticent when it comes to discussing artificial intelligence (AI), one of computing's most famous, but least successful, offspring. AI has its roots in a paper by Alan Turing, in which he asked whether a machine could think, and proposed the so-called "Turing Test" as a way of answering that question. In the mid-1950s, researchers in the United States began writing programs that could prove simple theorems in logic and geometry. By the end of the decade, proponents of AI were confidently predicting that within ten years, computers would translate from one language to another and play chess better than the human world champion.

Thirty years later, we are still waiting. Just why AI failed to meet its original aims (or at least its original timetable) is given a thoughtful, even-handed treatment. Both proponents and opponents of the thesis that computers can be programmed to do anything which a human mind does are given a chance to air their views, and the philosophical as well as practical obstacles to machine intelligence are explained clearly.

Overall, this book is carefully researched and well organised. Anyone from a young teenager keen on computers to an adult who wishes to know where all of today's wonderful toys have come from will find something worthwhile in it. Now, if only someone could explain to me how the robot arm pictured on the front cover can possibly hold that letter 'A' without dropping it...
— Dr. Dobb's Electronic Review of Computer Books