The Soul of a New Machine

by Tracy Kidder

  "Artificial intelligence" had always made for the liveliest of debates. Maybe the name itself was preposterous and its pursuit, in any case, something that people shouldn't undertake. Maybe in promoting the metaphorical relationship between people and machines, cybernetics tended to cheapen and corrupt human perceptions of human intelligence. Or perhaps this science promised to advance the intelligence of people as well as of machines and to imbue the species with a new, exciting power.

  "Silicon-based life would have a lot of advantages over carbon- based life," a young engineer told me once. He said he believed in a time when the machines would "take over." He snapped his fingers and said, "Just like that." He seemed immensely pleased with that thought. To me, though, the prospects for truly intelligent computers looked comfortably dim.

  To some the crucial issue was privacy. In theory, computers should be able to manage, more efficiently than people, huge amounts of a society's information. In the sixties there was proposed a "National Data Bank," which would, theoretically, improve the government's efficiency by allowing agencies to share information. The fact that such a system could be abused did not mean it would be, proponents said; it could be constructed in such a way as to guarantee benign use. Nonsense, said opponents, who managed to block the proposal; no matter what the intent or the safeguards, the existence of such a system would inevitably lead toward the creation of a police state.

  Claims and counterclaims about the likely effects of computers on work in America had also abounded since Weiner. Would the machines put enormous numbers of people out of work? Or would they actually increase levels of employment? By the late seventies, it appeared, they had done neither. Well, then, maybe computers would eventually take over hateful and dangerous jobs and in general free people from drudgery, as boosters like to say. Some anecdotal evidence suggested, though, that they might be used extensively to increase the reach of top managers crazed for efficiency and thus would serve as tools to destroy the last vestiges of pleasant, interesting work.

  Dozens of other points of argument existed. Were computers making nuclear war more or less likely? Had the society's vulnerability to accident and sabotage increased or decreased, now that computers had been woven inextricably into the management of virtually every enterprise in America?

  Wallach and I retreated from the fair, to a cafe some distance from the Coliseum. Sitting there, observing the more familiar chaos of a New York City street, I was struck by how unnotice able the computer revolution was. You leave a bazaar like the NCC expecting to find that your perceptions of the world outside will have been altered, but there was nothing commensurate in sight — no cyborgs, half machine, half protoplasm, tripping down the street; no armies of unemployed, carrying placards denouncing the computer; no TV cameras watching us — as a rule, you still had to seek out that experience by going to such places as Data General's parking lot. Computers were everywhere, of course — in the cafe's beeping cash registers and the microwave oven and the jukebox, in the traffic lights, under the hoods of the honking cars snarled out there on the street (despite those traffic lights), in the airplanes overhead — but the visible differences somehow seemed insignificant. Computers had become less noticeable as they had become nailer, more reliable, more efficient, and more numerous. Surely this happened by design. Obviously, to sell the devices far and wide, manufacturers had to strive to make them easy to use and, wherever possible, invisible. Were computers a profound, unseen hand?

  In The Coming of Post-Industrial Society, Daniel Bell asserted that new machines introduced in the nineteenth century, such as the railroad train, made larger changes in "the lives of individuals than computers have. Tom West liked to say: "Let's talk bout bulldozers. Bulldozers have had a hell of a lot bigger effect on people's lives." The latter half of the twentieth century, some say, has witnessed an increase in social scale — in the size of organizations, for instance. Computers probably did not create the growth of conglomerates and multinational corporations, but they certainly have abetted it. They make fine tools for the centralization of power, if that's what those who buy them want to do with them. They are handy greed-extenders. Computers performing tasks as prosaic as the calculating of payrolls greatly extend the reach of managers in high positions; managers on top can be in command of such aspects of their businesses to a degree they simply could not be before computers.

  Obviously, computers have made differences. They have fostered the development of spaceships — as well as a great increase in junk mail. The computer boom has brought the marvelous but expensive diagnostic device known as the CAT scanner, as well as a host of other medical equipment; it has given rise to machines that play good but rather boring chess, and also, on a larger game board, to a proliferation of remote-controlled weapons in the arsenals of nations. Computers have changed ideas about waging war and about pursuing science, too. It is hard to see how contemporary geophysics or meteorology or plasma physics can advance very far without them now. Computers have changed the nature of research in mathematics, though not every mathematician would say it is for the better. And computers have become a part of the ordinary conduct of businesses of all sorts. They really help, in some cases.

  Not always, though. One student of the field has estimated that about forty percent of commercial applications of computers have proved uneconomical, in the sense that the job the computer was bought to perform winds up costing more to do after the computer's arrival than it did before. Most computer companies have boasted that they aren't just selling machines, they're selling productivity. ("We're not in competition with each other," said a PR man. "We're in competition with labor.") But that clearly isn't always true. Sometimes they're selling paper-producers that require new legions of workers to push that paper around.

  Coming from the fair, it seemed to me that computers have been used in ways that are salutary, in ways that are dangerous, banal and cruel, and in ways that seem harmless if a little silly. But what fun making them can be!

  A reporter who had covered the computer industry for years tried to sum up for me the bad feelings he had acquired on his beat. "Everything is quantified," he said. "Whether it's the technology or the way people use it, it has an insidious ability to reduce things to less than human dimensions." Which is it, though: the technology or the way people use it? Who controls this technology? Can it be controlled?

  Jacques Ellul, throwing up his hands, wrote that technology operates by its own terrible laws, alterable by no human action except complete abandonment of technique. More sensible, I think, Norbert Wiener, prophesied that the computer would offer "unbounded possibilities for good and for evil," and he advanced, faintly, the hope that the contributors to this new science would nudge it in a humane direction. But he also invoked the fear that its development would fall "into the hands of the most irresponsible and venal of our engineers." One of the best surveys of the studies of the effects of computers ends with an appeal to the "computer professionals" that they exercise virtue and restraint. When the team came trooping back onto the bus — laughing, teasing, their faces glowing in the evening — it became obvious that many of them had not stayed long at the fair. They had gone off instead to see Greenwich Village and Times Square, and they looked refreshed. The conference didn't really fit their interests. "I don't care how computers get sold. I just build 'em," said one of the kids. Alsing remarked, "I don't even know how much an Eclipse M/600 costs."

  Some in the community of Eagle cheerfully professed ignorance about, and little interest in, the ultimate uses to which the machines they built were put. But they did not hold consistently to that attitude. Some had seen in use machines that they had made, and they confessed that was a thrill. Some didn't hold to the attitude at all. "No, I think about it a lot," Chuck Holland told me. "Initially, when I was starting out, I could have worked for a com-

  any that makes machines directly for the military, and for more money. But I'm not gonna design anything that directly bombs


  someone."

  Young computer engineers who professed anxieties about the fruits of their labors — those to whom I talked, anyway — usually

  named as the source of their worry the military applications of computers. One young man told me that if his company

  would ever start building devices of destruction, he would have to talk the executives into stopping, and if worse came to worst, he would make sure that those devices never worked. No

  doubt he meant it, but I think he overestimated his power. He worked for another outfit; no member of the Eclipse Group ever talked of sabotage. It would, in any case, have been an idle

  chatter.

  When the Eagle was done, according to the plan, Data General would ship the blueprints to an OEM, which would build a "ruggedized" version of the machine. Thus equipped for battle, Eagle would be sold to other OEMs, and they in turn would put it together in packages of military design for sale to the Department of Defense. Not everyone in the group believed that such applications would constitute an unworthy end, while others preferred not to think about that side of the process. What, in any case, could a dissident have done in the situation? Make sure that the plans for Eagle got fouled up and that the machine never came to life? Eagle, as it was planned, would live up to its designation as a general-purpose computer, competent at scientific and commercial chores, as well as military ones. If it did any, it could do all. A dissenter could indeed refuse to do work that might end up in the hands of soldiers: that would mean, in effect, not being a computer engineer.

  Shortly after he arrived at the company, at a party for newcomers known familiarly as "cocktails with the Captain," Jon Blau had asked de Castro if Data General did business with the Republic of South Africa. De Castro replied, Blau remembered, that the company did not at that time, but that Data General wasn't necessarily guided by politics — which was perhaps an understatement. Blau worried more openly than anyone else in the group about how people used computers, but he felt confused about the subject. "A friend asked me what is my social role. Obviously, what I'm doing is in demand, but it's hard for me to pinpoint why But for every bad use of a computer, I can think of a good one."

  The computer's reputation for awesome obscurity and the real complexities of the engineers' trade made barriers that were hard to cross. Their wives, some of the team said — and some of the wives agreed — didn't know much about what they did all day. "No one understands what we do," said Alsing. These youngsters led, in this respect, monastic lives. What were computers doing, going to do to society? They weren't often asked.

  A number of the engineers read prodigious amounts of science fiction. Dave Keating, for instance, read three or four novels of this sort a week. "For one," he explained, "science fiction has a lot of optimism in it. I mean, we made it that far at least. Part of it, too, we can identify with the technology; and I like the imagination in those books." Several of them, at idle moments, liked to conjure up stories of their own. Chuck Holland told me: "I do a lot of thinkin' about what my mind must do. The way I think it's gonna be is that the computer will grow up with the child. When you're born, you'll be given a computer. There'll be this little thing" — he grabbed his shoulder to show where it would sit — "that goes around with you. You'll teach it how to talk after you learn how." He imagined a teenaged boy teaching his computer how to drive, telling it, "Okay, now you give it a try," and correcting it when it made a mistake. "It's gonna be an extension of me," Holland said. But maybe, he went on, a time would come when the home would be "practically a simulator," and the computer would run every aspect of a person's life. "Then we get tired of it. We start growing plants or something. Maybe slowly we will turn around and go away from it. If computers take something away from us, we'll take it back. Probably a lot of people will get screwed before that happens."

  Alsing said: "I have a great idea for a science fiction story. What happens when infinite computer memory becomes infinitely cheap?" But that was on the ride back, and no one at that moment seemed interested in such speculations. They were having fun. The Microteam gave Holberger an Honorary Microcoder's Award — with some reservations, on account of a change he had made in UINST after they had already taken their unanimous vote (all such awards required unanimous approval, Alsing pointed out). Holberger accepted with a graceful speech from the back of the bus.

  Someone asked: "Hey, where's the Michelob? There was a whole case of it somewhere."

  "We drank it all," someone else confidently and proudly asserted.

  A little later on, a case of Michelob was handed back from seat to seat, while Alsing stood in the aisle holding forth hilariously on the perils that New York City's massage parlors held for young men from the country.

  They were free for a little while on a summer's night — comfortable, of course, in the assurance that interesting work awaited them tomorrow, but for the moment unbound from their machine.

  THE LAST CRUNCH

  In August, Carl Carman asked Ed Rasala when he thought they would finish the debugging.

  Rasala looked squarely at his division's vice president and said, "I don't know."

  West was greatly amused.

  As the last two items on his many debugging schedules, Rasala listed the Eclipse and the Eagle Multiprogramming Reliability Tests, the hardest of all the diagnostic programs. When Eagle could play those for a full night without failing, then it would be ready to become a computer. The debuggers would, Rasala planned, run Adventure, and then they would send one of the prototypes down the hallway to the Software Department. They would "ship it to Software," and the thirty or so software engineers now working on the project would continue to endow it with the complex set of programs known as an operating system.

  Each time they had approached the deadline for this consummation, West had named a new deadline, which he would announce to superiors and various interested departments. Accordingly, Rasala would draw up a new debugging schedule. West said April, and Rasala worked up a plan that might get them there by then. It didn't, so West proclaimed May to be the date, and Rasala followed suit. Then it was June, and finally West settled on the end of September. By then, Rasala couldn't stand it anymore. He wasn't making any more promises. On the other hand, he felt that they had better do it by the end of September. "Our credibility, I think, is running out."

  As a group, they often got ahead of themselves. They began to celebrate that summer. As their first favorite pretext for a party, they used the presentation of the Honorary Microcoder Awards that Alsing and the Microteam had instituted. Not to be outdone, *he Hardy Boys cooked up the PAL Awards, the first of which they gave to the Eclipse Group's own "CPU," after work, at a local establishment called the Cain Ridge Saloon. The citation read as follows.

  HONORARY PAL A WARD Rosemarie Seale

  In recognition of unsolicited contributions to the advancement of Eclipse hardware above and beyond the normal call of duty, we hereby convey unto you our thanks and congratulations on achieving this "high" honor.

  It came in a frame, and with a plastic cover. Glued to the center of the citation was a socket, such as is used to hold a PAL chip on a printed-circuit board. This being a PAL Award, the socket was empty.

  At a party at his home, Chuck Holland handed out his own special awards to each member of the Microteam, the Under Extraordinary Pressure Awards. They looked like diplomas. There was one for Neal Firth, "who gave us a computer before the hardware guys did," and one to Betty Shanahan, "for putting up with a bunch of creepy guys."

  Having dispensed Honorary Microcoder Awards to almost every possible candidate, the Microteam instituted the All- Nighter Award. The first of these to Jim Guyer, the citation ingeniously inserted under the clear plastic coating of an insulated coffee cup.

  Somewhat later, at a time more appropriate — at the Appreciation Dinner that Carman, with coaxing from Alsing, threw for the team — the wives got their due, the Eagle Awards. Alsing wrote all the citations. Penny Rasala's rea
d:

  EAGLE

  (Eclipse Appreciation and Gratitude for Lonely Evenings Award)

  This is to certify that Penny endured many lonely hours in the absence of Edward Rasala while Gallifrey Eagle was being developed, and on many occasions wondered what was the advantage of Ed being boss if he had to work 2 shifts.

  It was signed, in type, by one "Gallifrey Eagle," and witnessed by Alsing.

  Thus — as Alsing, himself the main instigator and organizer of parties, put it — "We congratulated ourselves on finishing Eagle and then we went back and finished it."

  It was an evening in August. They still had a long way to go. But, as if in concert with the season, the Hardy Boys were stuck in the deepest lull of their campaign. They had gone back to working single shifts. Even the pace of Rasala's conversation had slackened. Walking toward the lab, he spoke in short sentences and paused in between them. 'The momentum has slowed down dramatically. ... Back in the early days, when nothing worked, it was easy to find things to do Now almost everything works

  The problems are harder to find They take days The people are more tired, I think The problems may be less interesting. ... Some are more complicated This is the grind part, I

  think."

  Only a few were working in the lab that evening. None was too busy to talk. Seated in front of Gollum, Jim Veres discussed the picture on the blue screen of his analyzer. Hundreds of white dots blinked on and off while many white lines danced among them. The screen looked like the ceiling of a planetarium, the stars moving, white lines between them sketching out the constellations. It was lovely. Veres called this picture a map. "After staring at the map a long time," he said, "you get used to what it should look like." By now he and the other debuggers could tell which of the dozens of diagnostic programs was running just by looking at the map. Did I notice how the patterns kept changing in rhythm? he asked. If on some part of the screen images froze, that meant trouble inside. "This picture looks healthy."