Dawn of the new everythi.., p.11
Dawn of the New Everything, page 11
Loads of people in the technical world are obsessed with the kinds of games that I find boring, and humiliating, in a way, because of the lab rat role you must embrace. I see these games as math’s way of portraying moral and societal failure.2 Life should be about rejecting claustrophobic games like that, not becoming adept at them. The most important math concerns avoiding games with fixed rules and sharply defined winners and losers.
And yet, games were the only interactive art form that was making money. How could I not go there?
My first job interview was actually all the way up in pretty Marin County, over the Golden Gate Bridge. George Lucas was starting an organization to create digital effects for movies but also for video and audio editing, with ambitions to get into video games. You might think I was interested because of Star Wars, but no, it was because a student of my hero Ivan Sutherland, a fellow named Ed Catmull, had started these digital efforts.
I entered a big, unmarked industrial building and was greeted by a giant painting of the Organ Mountains, the pinnacles I stared at so often as a child in New Mexico. How could this be? Turned out that one of the other prime digital gurus at the place was Alvy Ray Smith, another migrant from our corner of the desert.
It was cool but a little disorienting to see Alvy, as if universes were colliding. He had grown up just a stone’s throw from the dome. I knew him mostly for his wonderful work extending something called Conway’s Game of Life.
The Game of Life was a program—created by mathematician John Horton Conway—that showed a grid of dots that blinked on or off according to simple rules about whether neighboring dots were on or off. By tweaking the rules and the initial pattern of dots, you could get amazing, unpredictable things to happen, as if the Game was a miniature living universe.
Alvy proved that you could even cause a fully powered computer to come into existence within the confines of the game, so that there could be worlds within worlds—an insight that would be popularized by Stephen Wolfram years later. It was natural to speculate that we might be living within something like a Game of Life.
Here was a “game” that expanded. It didn’t stick the player into a tiny abstract prison.
Alvy’s work comforted me. Once I understood that a deterministic game like the Game of Life could produce unforeseeable results, then a dark anxiety I had felt melted away. There was no longer tension between determinism and free will. If the only way to know the future was to actually run the universe, then it no longer mattered to my philosophy if there was a deterministic floor to the thing. There might or might not be. We could never know, from inside the universe. Moot point.
Of course the most useful physics might include randomness, or not, but that no longer mattered to philosophy. Math doesn’t kill freedom! Faith in the reality of free will makes as much sense as its rejection.
Hackers used to argue about these ideas all the time. “The ability to reject the idea of free will is an example of free will.” “Are you saying that thing you just said couldn’t be stated in a universe where there is no free will? Wrong! I can write a program that says it right now.”
Alvy as a person is as comforting as his math. He has a cheerful approach to computers and life that I still enjoy. Abstractions are emotional! A physicist who works on theories in which the universe is emerging and unpredictable will tend to be warmhearted and funny, like Lee Smolin, for instance.
But let’s get back to my story.
Loop Skywalker
The person who interviewed me wasn’t Alvy, alas, but yet another polished young Suit. He obviously wished he were working in the glamorous movies instead of in what were still considered the second-rate digital boondocks.
“We want to eventually bring Star Wars to life, so that it’s you directing Luke Skywalker. You’ll be able to make him swing the light saber with the joystick. Do you think you could make a digital light saber look like it’s glowing on one of these 8-bit machines?”
“Oh, you know, I don’t think this is the job for me.”
“Wha … How can you say that? This is the biggest thing ever.”
“I don’t mean any offense. It should be an amazing job for the right person. I just don’t like Star Wars that much.”
“What the fuck! Why are you here?”
“Well, I didn’t know what the job would be.”
“How can you not like Star Wars? Everyone loves Star Wars!”
“Oh gosh, I don’t hate it … I can explain if you want.”
“Oh yeah, I have to hear this.”
“Well, look. A few years ago, when I was a kid, I used to play music to accompany Robert Bly when he gave readings back in New Mexico.”
“Who that?”
“The poet? You know … He was reading his translations of Rumi—the ancient Sufi.” I was clearly not communicating. “Uh, that’s the sort of hippie mystic part of Islam—goes right back to the origins—anyway … We got booked along with Joseph Campbell, who would give talks.”
“Oh yeah, we all know about him. George used his book Hero with a Thousand Faces as a template for Star Wars.” As if this guy was best buddies with Lucas. “Wait, you know Campbell?”
“Not really, just performed on the same billing at this retreat place by some hot springs.”
“I don’t believe you.”
“Well, yeah. Anyway, Campbell—really great guy—has a theory I don’t really like much, that all human stories are variations on the same shared story. Kind of like how Noam Chomsky says there’s a core of language.”
“Don’t know this Chomsky, but hell yes, if you make a pure version of that universal story, you strike gold. We did and keep on doing it. What’s your problem? You hate money?”
“It’s so confining. Not money; I mean this idea about stories. What if we don’t really understand the stories of other cultures? Who are we to say they’re telling the same story we are? And if there’s only one story, how can we have hope that stories will get better in the future? If we believe there’s only one story, maybe we’re trapping ourselves in a small loop, like we’re in a primitive, crappy computer program. Alvy, who works here, proved there can be these expansive kinds of programs…”
“What the fuck are you talking about? Star Wars is set in the distant past, not the future, and it’s cool! Robots, faster than light spaceships! That would be a great future!”
“But the people are the same. They’re stuck in stupid petty power games. They’re cruel and selfish. Even the good guys are clannish and macho. Who needs more royal families? America was all about getting rid of them.”
“Oh god, you hippie idealists are so full of it.”
“Oh, don’t say I’m like that—am not! Um, but, science fiction can be about people getting better, not just gadgets getting better. I mean, in 2001: A Space Odyssey there’s this sense of transcendence, like we might outgrow our petty little conflicts. Well, maybe that’s not a great example—it’s pretty abstract and amoral. What about Star Trek? Gene Roddenberry had this idea that people would get kinder as the machines got better. That’s so much more exciting. I think it’s actually happened in human history already.”
“Star fucking Trek?”
“I guess I should go. Would you say ’bye to Alvy for me?”
“Not a chance.”
You Have to Get Awfully Weird to Avoid Becoming a Behaviorist
So, the Lucas world was not my cup of tea. But what an amazing feeling to be able to say no to an opportunity like that. I could take my pick of hundreds of jobs.
It was the 8-bit era. I coded a few games for various companies and was paid insanely well. College flunk-out debt gone like a dandelion in the wind.
It was especially satisfying to design sound effects and music. In those days the programmer might do everything from art and music to the instruction manual.
I wasn’t the only immigrant to Silicon Valley thinking this way. I started to meet other game hackers who thought of themselves as artists and scientists, and some of them would eventually help found the first VR company, VPL Research.
I met Steve Bryson, hippie physicist musician, dressed a little like Robin Hood, when we were both coding 8-bit games in a generic low-slung Sunnyvale office building. It had the usual grooved, prefab cement exterior; perfunctory hedges around the parking lot; show-off cars parked by the front door, my Dart parked around back.
A sterile habitat hosted marvelous, exotic people. What amazes me most when I think back on those days is how many incredible coders were also accomplished musicians. I remember storming into piano stores with five or six buddies, and everyone could not only play classical pieces expertly, but each was literate in jazz and had a personal, refined style. Steve Bryson, David Levitt, Bill Alessi, Gordy Kotik.
By 1981 I finally codesigned my first commercial video game. I worked with a toy and game expert named Bernie DeCoven on a title we called Alien Garden. It did fairly well. Then I designed my own game for the first time.
This was Moondust, which was a Top Ten home computer game when it was finally released in 1983. (The world was slow back then, before it was optimized. It could take years to release a program.)
Steve Bryson.
Moondust was sold in boxes! You used to go to a big store that mostly sold vinyl records and there would be a section with video game cartridges. I was infinitely proud to see Moondust on sale in its custom floor display, with promotional posters high up on the wall.
The best version ran on a Commodore 64, if anyone wants to dig it up. The music was algorithmic and pretty, with an echo and wetness, which was a trick in those days. The music was driven by the action, a first in gaming. The graphics had a glimmering soft quality instead of being blocky. Also quite a trick back when silicon was slow.
But the gameplay was bizarre. You influenced a whole swarm of spaceships at once to try to get them to smear a flowing ribbon of color into a ghostly, shimmering target, which would undulate orgasmically if you succeeded. The gameplay was too complex to approach analytically. One had to become intuitive, and then there was the oddly sexual quality.
It was amazing that throngs of customers bought copies of Moondust. I suspect they were drawn in by the graphics and sound but then quickly gave up on playing it. It was too strange, too open.
Grounded
Shortly after I arrived, I had rented an uninsulated hut in Palo Alto; old railroad worker’s shack on a dirt road, tilted on borrowed time in an un-devoured scrap of orchard by a creek.
You could tell who really understood Silicon Valley by their attitude toward real estate. I remember the real estate agent saying something that revealed her to be one of those horrid creatures who didn’t get it.
“You’re crazy not to buy a Victorian bungalow; those will be worth ten times as much in a few years.”
One of my hacker acquaintances was nearby and set her straight. “Code will run the world directly. Money is only an approximation of the code of the future, while we wait for computers to get cheap and connected. We’re creating a kind of power here that is much more important than money. Money’s obsolete, or will be any minute.” Yes, hackers talked like that. Everyone felt the call of oration.
The real estate agent looked to us like an uncomprehending dinosaur, staring right at the asteroid of doom.
My old home, even the ancient creek, has been untraceable for three decades now; instead, a mapping satellite detects only a smear of self-similar condos. I remember the rural smells of the gravel in the road and the moldy stained wood, the same inside and outside. California used to smell like grass, and it sounded like bugs and frogs.
Palo Alto was the spiritual center of Silicon Valley; less dismal than the grim empires of blandness down the road, like Sunnyvale, but still too dismal for me.
I stared up at the tall trees every evening in the endlessly perfect weather. The sky was always empty. No distant desert vista, no infinite sea, not even the wretched-but-fascinating urban New York mulch stretching to the horizon. All we could know was garden paradise, as imagined by immigrants from the snowy reaches of earlier American wealth. As if demons had tricked us with a simulation of heaven. Such a limited place, so discordant with my interior.
I felt dead lonely inside for years, for decades after my mother’s death.
A Club That Would Have Me as a Member
Hackers were always showing off their latest projects. Since the computers weren’t connected, you’d have to drive to see a demo, or bring it along with you. Instead of goats, the back of the Dodge Dart was now piled with computers, so I could haul around demos of my work. I remember having to occasionally pick bits of old hay out of hard drive slots.
I’d show everyone Moondust. I showed it to Alan Kay and his team up at Xerox PARC, to the people at Apple who would end up creating the Mac, to Doug Engelbart’s group at the Stanford Research Institute, to the people at NASA working with flight simulators.
One time I hefted one of those giant old CRT monitors up onto a table at this dimly lit dim sum place in an alley near Stanford—to show people Moondust, of course. (Don’t recall the name of the joint, but if anyone wants to figure it out, this was the one that put almond oil in their shrimp dumplings, and everyone talked about that.)
The diners that day would go on to found companies like Pixar and Sun. Moondust was a hit with this crew, and they started pestering me.
“How’d you do it? There are pixels changing all over the screen at once.”
“Oh, I’m using a compressed lookup table through these shifting masks…”
“Wait! Don’t tell them how you’re doing it!”
“I thought the hacker ethic was all about sharing code.”
“Well, yeah, if it helps bring down the big, bad old power. But this is your personal stuff.”
“I don’t know what to do.”
“Well, anyway, you’re one of us now.” One of us said with that emphatic, grunting rhythm from the movie Freaks.
Code Culture
Our world wasn’t made for us, yet. We were still profoundly strange.
The Valley already had elite pockets, but it was mostly not so rich, and much of it was raunchy and depressing. All of America, including the Valley, retained a slimy coating from the 1970s. Rusty signs with missing blinking lights offered live sex shows just north of Menlo Park, and beleaguered streetwalkers crowded the corners.
And yet this was our gathering place. We needed to stay close together, as there was not yet an Internet, but we needed network effects.
I remember playing pool at a rough dive bar on El Camino Real, the main drag, and thinking that a hacker in Palo Alto was like a cue ball that spins in a fixed spot after knocking another ball into faraway action. We spun in place in our new home while our momentum was transferred outward, reformatting the whole rest of the world.
Coding all night long, all the next day, coding until your brain absorbed a big abstract structure and perfected it. The experience was different than it is for coders today, because at that time you worked directly with the chip to get decent enough performance. That meant you weren’t dealing with languages, tools, or libraries from other programmers.
Everything important was fresh, entirely made of your own mind. You were an abstract explorer, facing only wilderness. If you wanted to get a circle to appear on a computer screen, you had to figure out a way to code a circle that would be fast enough to matter. I remember going with Bill Atkinson, who coded the graphical aspects of the original Macintosh, to see the legendary guru of algorithms at Stanford, Don Knuth, to present new ways of drawing circles. It was like visiting the code pope.
Push anything far enough and it transforms. This principle applies even to computers. At the core of the coding experience, when you are functioning at the very highest level of excellence, you reencounter a mysterious sense of the world that is not code-like.
There is—or at least there used to be—an amazing feeling in the gut when code was correct. An incredible, almost messianic feeling. We used to talk about it with a bit of embarrassment, our hidden store of mysticism buried under a fortress of rationality.
Whenever I had that feeling, the code in question would thereafter prove to be bug free. It was a strange and almost holy moment, and you only got to feel it once in a blue moon.
That experience of the apex of programming has become ever more elusive, because programs are no longer ever written by a single person; new programs of any importance are usually made by teams, and when they run, they’re spread out like moss grown upon myriads of preexisting software structures, which in turn aren’t even running on an identifiable computer, but instead roam secretly between the world’s uncharted, interconnected computers. One can no longer really know a piece of software; one can only test it, as if it were a newly discovered piece of nature. Yet another link to the old world of intuition severed.
Anyway, after days of concentration would come sleep like a velvet sea, often in your clothes, and then you might venture out and see other humans, but all of them had been doing the same thing. You looked like code to each other. You’d speak of the world as if it were an incomplete puzzle you were inventing.
I wish I remembered the names of all my earliest friends in the Valley. At least I remember the conversations. “I’ve kept data on all the sushi bars so that we can choose the optimal one.” “So have I.” “Did you timestamp your data? We could use a Bayesian method to correlate.”
This means of encountering the world was still done on paper! We carried around mini-notebooks and pencils. Hackers would mount their notebooks in little pretend metal cases to simulate what it would be like to have portable digital devices someday. There were a lot of fancy belt mounts, wrist mounts, and vest mounts. After all that calculation we’d eat our sushi and then go back to our coding.
After you spend all day coding, you dream in code; you think of the world as code. Scott Rosenberg wrote a book that in part recounted my experience of dreaming in code, called, naturally, Dreaming in Code. You’d wake up and realize you were coding in your sleep, coding the events that were taking place around you in the dream. A loop for one’s heartbeat.
