by Steven Levy
Minsky would say, “Brave Fredkin,” acknowledging the clandestine nature of Fredkin’s activity, which would have to be done without the knowledge of the hacker community; they would not tolerate an organizational structure that actually dictated where people should go.
While the destination could be industry—besides Systems Concepts, Fredkin’s Information International company hired many of the MIT hackers—it was often another computer center. The most desirable of these was the Stanford AI Lab (SAIL), which Uncle John McCarthy had founded when he left MIT in 1962.
In many respects SAIL was a mirror image of MIT’s operation, distorted only by the California haze that would sometimes drift from the Pacific Ocean to the peninsula. But the California distortion was a significant one, demonstrating how even the closest thing to the MIT hacker community was only an approximation of the ideal; the hothouse MIT style of hackerism was destined to travel, but when exposed to things like California sunlight it faded a bit in intensity.
The difference began with the setting, a semicircular concrete-glass-and-redwood former conference center in the hills overlooking the Stanford campus. Inside the building, hackers would work at any of sixty-four terminals scattered around the various offices. None of the claustrophobia of Tech Square. No elevators, no deafening air conditioning hiss. The laid-back style meant that much of MIT’s sometimes constructive acrimony—the shouting sessions at the TMRC classroom, the religious wars between grad students and hackers—did not carry over. Instead of the battle-strewn imagery of shoot-’em-up space science fiction that pervaded Tech Square, the Stanford imagery was the gentle lore of elves, hobbits, and wizards described in J.R.R. Tolkien’s Middle Earth trilogy. Rooms in the AI lab were named after Middle Earth locations, and the SAIL printer was rigged so it could handle three different Elven type fonts.
The California difference was reflected in the famous genre of computer games that the Stanford lab eventually developed after the heyday of MIT’s Spacewar. A Stanford hacker named Donald Woods discovered a kind of game on a Xerox research computer one day that involved a spelunker-explorer seeking treasure in a dungeon. Woods contacted the programmer, Will Crowther, talked to him about it, and decided to expand Crowther’s game into a full-scale Adventure, where a person could use the computer to assume the role of a traveler in a Tolkienesque setting, fight off enemies, overcome obstacles through clever tricks, and eventually recover treasure. The player would give two-word, verb-noun commands to the program, which would respond depending on how the command changed the universe that had been created inside the computer by Don Woods’ imagination. For instance, the game began with the computer describing your opening location:
YOU ARE STANDING AT THE END OF A ROAD BEFORE A SMALL BRICK BUILDING. AROUND YOU IS A FOREST. A SMALL STREAM FLOWS OUT OF THE BUILDING AND DOWN A GULLY.
If you wrote GO SOUTH, the computer would say:
YOU ARE IN A VALLEY IN THE FOREST BESIDE A STREAM TUMBLING ALONG A ROCKY BED.
Later on, you would have to figure all sorts of tricks to survive. The snake you encountered, for instance, could only be dealt with by releasing a bird you’d picked up along the way. The bird would attack the snake, and you’d be free to pass. Each “room” of the adventure was like a computer subroutine, presenting a logical problem you’d have to solve.
In a sense, Adventure was a metaphor for computer programming itself—the deep recesses you explored in the Adventure world were akin to the basic, most obscure levels of the machine that you’d be traveling in when you hacked in assembly code. You could get dizzy trying to remember where you were in both activities. Indeed, Adventure proved as addicting as programming—Woods put the program on the SAIL PDP-10 on a Friday, and some hackers (and real-world “tourists”) spent the entire weekend trying to solve it. Like any good system or program, of course, Adventure was never finished—Woods and his friends were always improving it, debugging it, adding more puzzles and features. And like every significant program, Adventure was expressive of the personality and environment of the authors. For instance, Woods’ vision of a mist-covered toll bridge protected by a stubborn troll came during a break in hacking one night, when Woods and some other hackers decided to watch the sun rise at a mist-shrouded Mount Diablo, a substantial drive away. They didn’t make it in time, and Woods remembered what that misty dawn looked like and wrote it into the description of that scene in the game, which he conceived of over breakfast that morning.
It was at Stanford that gurus were as likely to be faculty people as systems hackers (among Stanford professors was the noted computer scientist Donald Knuth, author of the multivolume classic The Art of Computer Programming). It was at Stanford that, before the Adventure craze, the casual pleasures of Spacewar were honed to a high art (Slug Russell had come out with McCarthy, but it was younger hackers who developed five-player versions and options for reincarnation, and ran extensive all-night tournaments). It was at Stanford that hackers would actually leave their terminals for a daily game of volleyball. It was at Stanford that a fund-raising drive was successfully undertaken for an addition to the lab, which would have been inconceivable at MIT: a sauna. It was at Stanford that the computer could support video images, allowing users to switch from a computer program to a television program. The most famous use of this, according to some SAIL regulars, came when SAIL hackers placed an ad in the campus newspaper for a couple of willing young coeds. The women answering the ad became stars of a sex orgy at the AI lab, captured by a video camera and watched over the terminals by appreciative hackers. Something else that never would have occurred at MIT.
It was not as if the SAIL hackers were any less devoted to their hacking than the MIT people. In a paper summarizing the history of the Stanford lab, Professor Bruce Buchanan refers to the “strange social environment created by intense young people whose first love was hacking,” and it was true that the lengths that hackers went to in California were no less extreme than those at Tech Square. For instance, it did not take long for SAIL hackers to notice that the crawl space between the low-hanging artificial ceiling and the roof could be a comfortable sleeping hutch, and several of them actually lived there for years. One systems hacker spent the early 1970s living in his dysfunctional car parked in the lot outside the building—once a week he’d bicycle down to Palo Alto for provisions. The other alternative for food was the Prancing Pony; named after a tavern in Middle Earth, this was the SAIL food-vending machine, loaded with health-food goodies and potstickers from a local Chinese restaurant. Each hacker kept an account on the Prancing Pony, maintained by the computer. After you made your food purchase, you were given the option to double-or-nothing the cost of your food, the outcome depending on whether it was an odd-or even-numbered millisecond when you made the gamble. With those kinds of provisions, SAIL was even more amenable than MIT for round-the-clock hacking. It had its applications people and its systems people. It was open to outsiders, who would sit down and begin hacking; and if they showed promise, Uncle John McCarthy might hire them.
SAIL hackers also lived by the Hacker Ethic. The time-sharing system on the SAIL machine, like ITS, did not require passwords, but, at John McCarthy’s insistence, a user had the option to keep his files private. The SAIL hackers wrote a program to identify these people, and proceeded to unlock the files, which they would read with special interest. “Anybody that’s asking for privacy must be doing something interesting,” SAIL hacker Don Woods would later explain.
Likewise, SAIL was in no way inferior to MIT in doing important computer work. Just like their counterparts at MIT’s AI lab, SAIL hackers were robotics fans, as implied by the sign outside SAIL: CAUTION, ROBOT VEHICLE. It was John McCarthy’s dream to have a robot leave the funky AI lab and travel the three miles to campus under its own physical and mental power. At one point, presumably by mistake, a robot got loose and was careening down the hill when, fortunately, a worker driving to the lab spotted it and rescued it. Various hackers and academics worked
at SAIL in important planner fields like speech understanding and natural language studies. Some of the hackers got heavily involved in a computer music project that would break ground in that field.
Stanford and other labs, whether in universities like Carnegie-Mellon or research centers like Stanford Research Institute, became closer to each other when ARPA linked their computer systems through a communications network. This "ARPAnet" was very much influenced by the Hacker Ethic, in that among its values was the belief that systems should be decentralized, encourage exploration, and urge a free flow of information. From a computer at any “node” on the ARPAnet, you could work as if you were sitting at a terminal of a distant computer system. Hackers from all over the country could work on the ITS system at Tech Square, and the hacker values implicit in that were spreading. People sent a tremendous volume of electronic mail to each other, swapped technical esoterica, collaborated on projects, played Adventure, formed close hacker friendships with people they hadn’t met in person, and kept in contact with friends at places they’d previously hacked. The contact helped to normalize hackerism, so you could find hackers in Utah speaking in the peculiar jargon developed in the Tool Room next to the Tech Model Railroad Club.
Yet even as the Hacker Ethic grew in the actual number of its adherents, the MIT hackers noted that outside of Cambridge things were not the same. The hackerism of Greenblatt, Gosper, and Nelson had been directed too much toward creating one Utopia, and even the very similar offshoots were, by comparison, losing in various ways.
“How could you go to California, away from the action?” people would ask those who went to Stanford. Some left because they tired of the winner-loser dichotomy on the ninth floor, though they would admit that the MIT intensity was not in California. Tom Knight, who hacked at Stanford for a while, used to say that you couldn’t really do good work at Stanford.
David Silver went out there, too, and concluded that “the people at Stanford were kind of losers in their thinking. They weren’t as rigorous in certain ways and they sort of were more fun-loving. One guy was building a race car and another was building an airplane in the basement . . .” Silver himself got into hardware at Stanford when he built an audio switch to allow people working at their terminals to listen to any of sixteen channels, from radio stations to a SAIL public-address system. All the choices, of course, were stored within the SAIL PDP-6. And Silver thinks that exposure to the California style of hacking helped loosen him up, preparing him to make the break from the closed society of the ninth floor.
The defection of Silver and the other MIT hackers did not cripple the lab. New hackers came to replace them. Greenblatt and Gosper remained, as did Knight and some other canonical hackers. But the terrifically optimistic energy that came with the opening explosion of AI research, of setting up new software systems, seemed to have dissipated. Some scientists were complaining that the boasts of early AI planners were not fulfilled. Within the hacker community itself, the fervid habits and weird patterns established in the past decade seemed to have solidified. Were they ossified as well? Could you grow old as a hacker, keep wrapping around to those thirty-hour days? “I was really proud,” Gosper would say later, “of being able to hack around the clock and not really care what phase of the sun or moon it was. Wakeup and find it twilight, have no idea whether it was dawn or sunset.” He knew, though, that it could not go on forever. And when it could not, when there was no Gosper or Greenblatt wailing away for thirty hours, how far would the hacker dream go? Would the Golden Age, now drawing to its close, really have meant anything?
• • • • • • • •
It was in 1970 that Bill Gosper began hacking LIFE. It was yet another system that was a world in itself, a world where behavior was “exceedingly rich, but not so rich as to be incomprehensible.” It would obsess Bill Gosper for years.
LIFE was a game, a computer simulation developed by John Conway, a distinguished British mathematician. It was first described by Martin Gardner, in his "Mathematical Games" column in the October 1970 issue of Scientific American. The game consists of markers on a checkerboard-like field, each marker representing a “cell.” The pattern of cells changes with each move in the game (called a “generation”), depending on a few simple rules—cells die, are born, or survive to the next generation according to how many neighboring cells are in the vicinity. The principle is that isolated cells die of loneliness, and crowded cells die from overpopulation; favorable conditions will generate new cells and keep old ones alive. Gardner’s column talked of the complexities made possible by this simple game and postulated some odd results that had not yet been achieved by Conway or his collaborators.
Gosper first saw the game when he came into the lab one day and found two hackers fooling around with it on the PDP-6. He watched for a while. His first reaction was to dismiss the exercise as not interesting. Then he watched the patterns take shape a while longer. Gosper had always appreciated how the specific bandwidth of the human eyeball could interpret patterns; he would often use weird algorithms to generate a display based on mathematical computations. What would appear to be random numbers on paper could be brought to life on a computer screen. A certain order could be discerned, an order that would change in an interesting way if you took the algorithm a few iterations further, or alternated the x and y patterns. It was soon clear to Gosper that LIFE presented these possibilities and more. He began working with a few AI workers to hack LIFE in an extremely serious way. He was to do almost nothing else for the next eighteen months.
The group’s first effort was to try to find a configuration in the LIFE universe, which was possible in theory but had not been discovered. Usually, no matter what pattern you began with, after a few generations it would peter out to nothing or revert to one of a number of standard patterns named after the shape that the collection of cells formed. The patterns included the beehive, honey farm (four beehives), spaceship, powder keg, beacon, Latin cross, toad, pinwheel, and swastika. Sometimes, after a number of generations, patterns would alternate, flashing between one and the other: these were called oscillators, traffic lights, or pulsars. What Gosper and the hackers were seeking was called a glider gun. A glider was a pattern which would move across the screen, periodically reverting to the same pointed shape. If you ever created a LIFE pattern, which actually spewed out gliders as it changed shape, you’d have a glider gun, and LIFE’s inventor, John Conway, offered fifty dollars to the first person who was able to create one.
The hackers would spend all night sitting at the PDP-6’s high-quality “340” display (a special, high-speed monitor made by DEC), trying different patterns to see what they’d yield. They would log each “discovery” they made in this artificial universe in a large black sketchbook, which Gosper dubbed the LIFE scrapbook. They would stare at the screen as, generation by generation, the pattern would shift. Sometimes it looked like a worm snapping its tail between sudden reverses, as if it were alternating between itself and a mirror reflection. Other times, the screen would eventually darken as the cells died from aggregate overpopulation, then isolation. A pattern might end with the screen going blank. Other times things would stop with a stable “still life” pattern of one of the standards. Or things would look like they were winding down, and one little cell thrown off by a dying “colony” could reach another pattern and this newcomer could make it explode with activity. “Things could run off and do something incredibly random,” Gosper would later recall of those fantastic first few weeks, “and we couldn’t stop watching it. We’d just sit there, wondering if it was going to go on forever.”
As they played, the world around them seemed connected in patterns of a LIFE simulation. They would often type in an arbitrary pattern such as the weaving in a piece of clothing, or a pattern one of them discerned in a picture or a book. Usually what it would do was not interesting. But sometimes they would detect unusual behavior in a small part of a large LIFE pattern. In that case they would try to isolate that p
art, as they did when they noticed a pattern that would be called “the shuttle,” which would move a distance on the screen, then reverse itself. The shuttle left behind some cells in its path, which the hackers called “dribbles.” The dribbles were “poison” because their presence would wreak havoc on otherwise stable LIFE populations.
Gosper wondered what might happen if two shuttles bounced off each other, and figured that there were between two and three hundred possibilities. He tried out each one, and eventually came across a pattern that actually threw off gliders. It would move across the screen like a jitterbugging whip, spewing off limp boomerangs of phosphor. It was a gorgeous sight. No wonder this was called LIFE—the program created life itself. To Gosper, Conway’s simulation was a form of genetic creation, without the vile secretions and emotional complications associated with the real world’s version of making new life. Congratulations—you’ve given birth to a glider gun!
Early the next morning Gosper made a point of printing out the coordinates of the pattern that resulted in the glider gun, and rushed down to the Western Union office to send a wire to Martin Gardner with the news. The hackers got the fifty dollars.
This by no means ended the LIFE craze on the ninth floor. Each night, Gosper and his friends would monopolize the 340 display running various LIFE patterns, a continual entertainment, exploration, and journey into alternate existence. Some did not share their fascination, notably Greenblatt. By the early seventies, Greenblatt had taken more of a leadership role in the lab. He seemed to care most about the things that had to be done, and after being the de facto caretaker of the ITS system he was actively trying to transform his vision of the hacker dream into a machine that would embody it. He had taken the first steps in his “chess machine,” which responded with a quickness unheard of in most computers. He was also trying to make sure that the lab itself ran smoothly so that hacking would progress and be continually interesting.