What is the history of digital colour? How did Moore’s Law shape the future of computer animation? Mike Seymour sits down with Pixar co-founder and lifelong innovator Dr Alvy Ray Smith for a discussion on building businesses from dreams and overcoming roadblocks.
Dr Alvy Ray Smith was a guest of the University of Sydney Faculty of Engineering for the 2019 Dean’s Lecture.
‘Genesis effect’ for Star Trek II – the Wrath of Khan, the first use of 3D CGI used in a movie that was shown to the public
Short film, Sunstone, showcasing pioneering animation techniques
From Pencils to Pixels, BBC documentary featuring John Lasseter
Faculty of Engineering, Dean’s Lecture 2019
You can subscribe to this podcast on iTunes, Spotify, Soundcloud, Stitcher, Libsyn, YouTube, or wherever you get your podcasts. You can follow us online on Flipboard, Twitter, or on sbi.sydney.edu.au.
This transcript is the product of an artificial intelligence - human collaboration. Any mistakes are the human's fault. (Just saying. Accurately yours, AI)
Sandra Our special guest on this episode has worked with Steve Jobs, George Lucas, Bill Gates and Steven Spielberg. Alvy Ray Smith dreamed of using computers to revolutionise movie making, and animated movies in particular. Movies and their infinite possibility was the connective tissue between Alvy and the cast of famous people he has worked with. And just like in the movies, not all of the relationships had happy endings. But they do make a great story.
Intro From the University of Sydney Business School this is Sydney Business Insights, the podcast that explores the future of business.
Sandra Alvy Ray Smith has been described as a "charismatic Texan with a PhD in computer science and a sparkling resumé". Not my words, but Ed Catmull's who, along with Alvy, founded Pixar. At the time, a tiny computer animation studio that went on to bring the world the Toy Story movies, A Bugs Life, Monsters Inc., Finding Nemo, The Incredibles and more than 21 hugely popular movies. Alvy's passion to make animated movies using computers is embedded in Pixar's DNA. However, he left the company after a tremendous fight with Pixar's owner, none other than Steve Jobs. Across his brilliant career Alvy has worked at four of the world's most innovative organisations, not only Pixar, but Xerox, Microsoft and Lucasfilm. In his corporate life, Alvy experienced recognition for his technical innovation, but also frustrating rejection. He was fired from Xerox, where he was developing the paint colour program because Xerox management thought the future of personal computing was black and white. No need for Alvy's colour palette. His work on digital paint systems eventually earned Alvy an Academy Award. The statue sits alongside his other Oscar, also for technical innovations that changed movie animation. Alvy was a guest of the University of Sydney's Faculty of Engineering, where he delivered the 2019 Dean's Lecture. He generously sat down in the SBI studios with my Business School colleague, and long-time Alvy fan, Mike Seymour. Mike's prodigious knowledge of Alvy's work is a bonus for us. Like Alvy, Mike has deep experience in visual effects and the entertainment industry. His academic research into digital humans plans to take cutting-edge technology from the entertainment industry, and apply it in a wider business context for the good of all.
Mike So, thanks so much for agreeing to talk to us, we really appreciate it.
Alvy Pleasure being here, Mike.
Mike When we spoke before, you were thinking about coming out with a book, which I think was the title Digital Light: A Biography of the Pixel. Where is that at?
Alvy I have just about completed it.
Mike So conceptually, a pixel for you is more than just a dot, more than just a rectangle. It's a moment of things coming together, almost a unification of technology, isn't it?
Alvy A pixel is a non-trivial concept, and that's one of the things I want to get across. You said rectangle, it's not a rectangle. It's a point, it's a sample of a continuum. You can't see a pixel, that kind of surprises most people. What you see on a digital screen are little glowing, we call 'em display elements, that put out little blobs of coloured light in a regular array, and the little blobs of light overlap, creating a continuous service of lightnesses. And that's what we see. That's what our visual system is designed to see. I call it spreading the pixels. You can only see spread pixels, and they're spread by the display elements.
Mike It's interesting, though, isn't it, because there's this world of maths, sampling theory. But by the same token, you could have most of that discussion and we'd be talking about nuclear physics. You could have most of that discussion and it would be relevant to modern data science, or...
Alvy It applies everywhere, right.
Mike It's an incredible thing to think how much these seemingly disparate technologies are all actually connected.
Alvy It's a simple idea with profound consequences. Yes, in all kinds of fields. But of course, it's what made digital light possible in my world.
Mike And your career has been nothing short of remarkable in realising that ability to basically simulate the world, which is, not putting too fine a point on it, a pretty ambitious and audacious undertaking.
Alvy It's interesting you put it that way, I never thought that that's what we were doing. Although many of my colleagues just assumed that's what we were doing, was trying to simulate the world. In my book, I've tried to elicit the notion called the 'central dogma of computer graphics'. The central dogma says we shall make models of the real world using Euclidean geometry. We shall project that model into two dimensional space, using Renaissance perspective, and the light fields and gravity and so forth shall honour Newtonian physics. Now that's all real world stuff, but there's nothing about a computer that requires any of those three things.
Mike Well, let me take you back now to Xerox PARC, because I want to see if we can map a couple of big, kind of, views that you must have had on the world, and how that sat with the view of the future at that point in time. Now I know Xerox PARC because of laser printers, the mouse, so many developments in graphical user interfaces. And I was astounded to know that when you were at Xerox PARC, you had trouble getting them to look at colour.
Alvy Yes. They fired me because I was doing colour.
Mike You've got to understand that doesn't gel with my world view, because this is a place rarefied in now almost mystical levels of inventiveness and story. And yet, you're saying that they didn't see colour as being important.
Alvy They didn't, it was one of the big surprises in my life, because it clearly was the hottest place going on the planet in computers at the time. The world as we now use it, was being invented there.
Mike Ethernet, everything.
Alvy Everything. I was doing colour pixels. So when they fired me, I went in, of course, and said, why? And they said, we've decided not to do colour. And I said, but the future is colour and you own it completely. And my boss actually said, you may be right, but it's a corporate decision to go black and white.
Mike Can we just unpack that a little? Because here we've got an organisation that is renowned for being innovative and being inventive. And yet, you are seeing a future and it seems in retrospect beyond obvious, so obvious that almost like you'd be insulted to bring it up at a meeting.
Alvy That's what I thought, it was so obvious! So for years, I was rather angry with Bob Taylor, who was the supposed visionary in charge of the group I was working with for missing this. I kept saying, well, he's not much of a visionary, he can't see colour. And it was only when I was writing this book that I finally realised, yes he did have a very clear vision. It's the desktop computing world that we now know, minus colour pixels. And he pulled that off. I even thought about calling him up and apologising to him that I had missed the fact that he was a great visionary. That I had a different vision that was beyond his, but he got there and he implemented his, he made it happen. And I called him up to say all that, and his son answered the phone and said Alvy, you're too late, he's on his deathbed. In fact, he died two days later so I never got to tell him that. Actually, my admiration of him had shot up once I put everything into perspective.
Mike There are so many aspects of everyday life where we sort of think that people would bump up against people that didn't have a vision, and thus would have trouble articulating a view of the future. Yet you're here at Xerox PARC, the centre of defining the future. How does that work and what can we learn from that? Because if you can't communicate colour at Xerox PARC, what are the rest of us stand any chance?
Alvy The only thing I can figure, Mike, is that Bob Taylor kind of understood what it would take to get his vision implemented, black and white. And he saw me as delaying that project because colour was too expensive still for the Alto computer, which they wanted to be a personal computer.
Mike I think I should set the context. At the time that we're talking about, in fact, a lot of the things we now take for granted in terms of, and I'm not even talking now about advanced computer graphics in films, I'm talking about the simple things like paint programs. You guys were inventing these things, right down to the fundamentals that now are the visual language that we use to communicate with computers. We'll discuss a bunch of companies that you'll go to and obviously help and co-found like Pixar, but that sort of university academia ethos permutated your entire career, and those of the companies that you established. The big international graphics conference SIGGRAPH is where people publish, and Pixar from day one was always a publishing company. You haven't had an ethos in your life of holding back, and sort of squirrelling away and not communicating. There's a very kind of academic sense, to the companies you...
Alvy Well we were very, we were always an academic outfit because both they and I are PhDs, we've come through the system. We weren't in industry, in fact our first two patrons were the businessmen. We weren't. We were just, they'd pay for us to do research. So that continued the notion that we were just an academic outfit. Ed and I understood that the pay was not money, the pay was fame.
Mike Well, academic recognition,.
Alvy That's fame. It's the same thing. People know you because of your work, and that is worth more than money, and we knew that. So there were certain things we would not let people publish because they were key, but mostly everything was let people know.
Mike So how is it that you got attracted to Long Island, the other side of the continent and what became the most influential academic hub for the next 40 years?
Alvy So I got fired at Xerox, and Xerox at the time had the only colour frame buffer in the world. My friend Dick Shoup had built it. A colour frame buffer being enough memory...
Mike Like a graphics card.
Alvy It'd be a graphics card now. It was enough memory to hold a picture made out of pixels.
Alvy Okay. Well, my buddy David DiFrancesco (he's an artistic friend of mine) and I had cast our lots together at Xerox by submitting to the National Endowment for the Arts a grant proposal to exploit this new art medium. Well, we had to have a frame buffer, we had to have a graphics card, as you might put it now, even though it was the size of a refrigerator. We had to find the next one. So we heard a rumour they were building one at Utah. So we drove over to Evans and Sutherland. And indeed, they were building the next frame buffer there. But they wouldn't take on too hairy looking guys like David and me.
Mike I think that we should not skip this point. We're talking about things that are so rare that there's like one or two of them in the world. Like today, there are graphics cards in every computer. You had to drive to another state to just find...
Alvy The next one. It literally was the next one. And they said, guys, we're Department of Defence funded. We can't have artist here. We didn't use the word art, but they figured us out pretty fast. Right. So one of them, Martin Newell, finally said. But, you know, there was a rich man who came through here the other day and he bought one of every machine in sight. And I said, including the frame buffer? He said, yes. So there was a guy who was going to have the next frame buffer, after Utah.
Mike The guy you nicknamed 'Uncle Alex'.
Alvy Yes. His name was Alexander Schure, and by one of the strangest coincidences in the world, he was the uncle of the fellow I'd been living with in Palo Alto, working in a Xerox PARC. It's one of those unexplainable coincidences.
Mike So there are gonna be four huge figures in the narrative as we move forward. He's the first obviously at NY Tech. You're going to have George Lucas in a moment at Lucasfilm. And then, of course, you're going to set up Pixar, which was funded by Steve Jobs, and then Microsoft with Bill Gates.
Mike So I'm wondering if we can look at those four periods and starting with this first one, we're at NY Tech, was there any characteristic of his role as the leader of that group that you now look back on as just being a lesson you could learn?
Alvy He was an entrepreneur. New York Tech was a private university and you know, he just made money.
Mike But he had a vision for the future. Because why would he go and get these incredibly rare as hen's teeth...?
Alvy Well, he wanted to be the next Walt Disney. Alex Schure, the owner, New York Tech, indirectly knew Walt Disney. And he was already underway making an animated film on the campus when we showed up the old-fashioned way, or the Disney way, where you draw characters in black ink on celluloid, and then you paint in between the curves with opaque coloured paints. It's inking and opaquing, that's what he was doing, he had a hundred person team on the campus. And so a travelling salesman from Evans and Sutherland named Pete Ferentinos just cold called him one day. His territory was the eastern half of the United States. He found this guy and talked him into believing that computers would make the production of this film called Tubby the Tuba more cost effective. And Alex Schure, being who he is, thought that meant that he could buy computers, fire all the people, and just have the computers make the animated movie. When he said that to us he says Alex, you can't say that it's not true. We don't have any idea how to replace the creativity. We can do the groundwork, and do it well.
Mike And when you say we, this is like you and Ed Catmull.
Alvy Ed Catmull, that's where I met Ed. He had just left University of Utah with his PhD and had joined this guy on Long Island.
Mike And of course, you and Ed will go on to form that computer group at Lucas, but also, of course, Pixar as the founders.
Mike What was that like? Just from a management point of view, like how did that team run? Were you guys independent?
Alvy Yeah, we were. Basically, you've already said it was collegiate. We were just independent entities. Everybody was intelligent, and everybody kind of knew what the problem was. We were going to make movies with computers. And there were a host of problems to be solved. Nobody had to order anything to be done, just somebody would volunteer to do that problem. The way we'd hire people is that they show up with that look in their eyes, and they had the knowledge, and they were just smart enough. If we'd actually had to manage, neither Ed or are very good manager so it wouldn't have worked.
Mike Well, you say that, but history will prove you wrong. OK, so you don't have to really articulate the future here, because you just allowed to go do it.
Alvy Yeah, that was one of our blessings. We had a very easy, clear vision: Make the first movie.
Mike Which is a kind of a moonshot definition of how to have a successful, impactful project, isn't it?
Alvy We had no idea it was going to take 20 years.
Mike Yeah. So you have this super-creative environment, self-motivating team. You've got the latest toys. It's a beautiful place as well, Long Island.
Alvy It's gorgeous.
Mike So now you go over to the other side of the country to join one of the most successful filmmakers ever at that time, George Lucas and set up perhaps slightly unimaginatively named 'Lucasfilm Computer Division'.
Alvy You know the story there, we wanted a really sexy name like Industrial Light and Magic,.
Mike But you just couldn't agree on one.
Alvy We couldn't agree. Again, we're collegiate. At least two people had to agree on the name and no two people, and even George Lucas threw in name suggestions. Couldn't get anybody to support his name.
Mike So you're setting up this computer division over in California now. And the thing that I find astounding here is, again, this is now the most progressively, I mean, it's making Sci-Fi. It's imagining the future in the films. And yet it wasn't immediately easy to sell your vision of the future at Lucasfilm to one of the visionary filmmakers of his time.
Alvy And that was one of our big misperceptions. We assumed that he wanted us to come and make content for his movies. He being George Lucas. But after I was there for a while, I was the head of computer graphics and I just started hiring the best guys in the world, like Tom Porter and Loren Carpenter and these geniuses. My group of genius.
Mike I mean, it's the dream team of computer graphics.
Alvy Everybody wanted to be in a movie, and it's Northern California, it's a fantastic place to live, it's like living in Sydney, it's really a wonderful place. And I start waiting for George. And he never came. And all of a sudden it dawned on me. I said, oh my gosh, he doesn't have a clue what he's got. We had told him we could build hardware and software. See, this is another guy who had a clear vision, and it just didn't happen to be ours. His vision was he was going to digitise the moviemaking machines of Hollywood. And he had a very clear idea about that. He wanted to make editing film digital, and editing audio digital, and he wanted to use computers to do all the logistics tracking. You know, it is how Hollywood is done now.
Mike Yeah. In fact, in that period, that was the EditDroid, which is the precursor to offline editing that we now think of as just Premier or any of the editing programs. But here's the thing I find really fascinating, and this is not without its irony: for a guy who made Star Wars, the first sequence that you directed in a major motion picture was in Star Trek.
Alvy That's right.
Mike It's just, you couldn't write that in a script.
Alvy No, it's like, OK, so he's not going to show up, what do we do? Then in another one of the many lucky breaks along the way was Star Trek producers showed up to hire Industrial Light & Magic, the special effects division.
Mike But not directed by George Lucas now.
Alvy No, no, no. It was a special effects house for hire when they weren't busy doing George's movies. So ILM, Industrial Light & Magic says yeah, we'll do special effects for Star Trek too. Well, the producers of Star Trek said we want some of this newfangled computer graphics. The guys at ILM said, we don't do that. We think the new guys next door to that. Well, the new guys next door were me and my team.
Mike So you get a chance to make a sequence, you personally directing a sequence, in a major motion picture, ironically for Star Trek. And I got to just say, at this point in time, this was such a quantum leap in the imagining of what computer graphics could do, because everybody else up until this point thought we were doing good stuff if we did logos, and shiny metal, and obvious things, maybe a robot if you were lucky. And you guys went, oh no, we'll make a planet. And not only we'll make a planet, we'll grow mountains and trees and stuff. And it was as if somebody opened the roof and sunlight came in and we went, Oh my God, this is what computer graphics could do.
Alvy It's pretty crude compared to your description there.
Mike Well, it was fractals and particles, and it was marvellous, but it was organic. That's a word over-used now. But in that day, I literally, like, that was the only thing out of that movie, I was like, oh, my God.
Alvy I can't remember the rest of the movie
Mike Yeah, I'm sure Shatner was great.
Alvy When I landed that job. Two things, one, I told the movie producer, I said, you know, we can't do movie resolution yet. Computers aren't fast enough yet. We can only do television resolution. And the guy said, that's perfect because this is gonna be a video demo to Admiral Kirk. I said perfect, and I walked out of the room saying, OK, we've got this one. And I called the team together and said, all right, we've got our first big break. We've got a major motion picture that's probably going to be a successful motion picture. There had been computer graphics in the movies before, but they were failures, so nobody had seen them. This is going to be a success and we're going to do a great job. We've got a 60 second shot. Paramount's going to be happy. They were the production house. And it's going to make perfect narrative sense, and it's going to heighten the story, it's going to make sense to the audience. But what this really is, is a 60 second commercial to George Lucas. So he'll know what the hell he's got.
Mike So you are imagining the future by literally making it and then saying here, this is the future. Look, I made one to show you what it's going to look like.
Alvy I still don't quite know why I knew this, but I knew that George had a way of watching movies that nobody else does, so far as I can ascertain. He never, ever loses track of the camera.
Mike He's aware of the cinematography, of how it's shot.
Alvy Always, non-stop. And if you think about it, if you're watching the cameramen, the director's failed.
Mike You're out of the story.
Alvy You've not been sucked into the emotion. I think George could do both. He was always aware of the cameraman. And I gotta assume that he was also tracking the story, too. So I knew that, and I told my group that I said, we're gonna put a camera shot in this Star Trek piece that will blow his socks off. He'll know that no camera could do it, no real camera could do it. But it won't be gratuitous. It won't be computer graphics 101 where you just flip people around for the fun of it. It'll make narrative sense.
Mike Yeah, did. It was an evolution of a planet as it grew.
Alvy And the camera's tracking all the flying, it's coming over the edge and all of that. And it's turning upside down as the spacecraft pulls away. It's a really complex move. And sure enough, the day after the premiere, George Lucas, a pretty shy man, stepped one foot into my office and said, great camera move. And he was gone. He got it. And then he put us in his next movie.
Mike Because you've literally visualised the future in another space thing. He's got a franchise that was space. You took an example of what it could look like in a space context, deliberately tailored to something that he would appreciate, and then presented it. Did that mean that from then on out, the future was an easy thing for you at Lucasfilm?
Alvy No. No. It did not mean that. He did put us into a very short piece in the next movie, Return of the Jedi. A very brief piece. The most important thing he did was tell his buddy Steven Spielberg about us. And Spielberg had us in his next movie, and then it just started.
Mike Which was what?
Alvy The young Sherlock Holmes.
Mike And the famous stained glass sequence...
Alvy The stained glass man jumps out of the stained glass and fights with a real-world priest.
Mike Another huge stepping stone.
Alvy Every time we did one of these, we upped the quality of the rendering, and in terms of the central dogma we were honouring more and more of Newtonian physics. We did another piece after the Genesis demo, André and Wally B.
Mike By the way, for those who don't know, that's a short film and was the start of the huge history of shorts that have been almost the experimental playground for up and coming directors in computer graphics. But that started with that one André and Wally B, right? Which I think is also where John Lasseter enters the picture.
Alvy John Lasseter showed up, just saved me, by the way, he saved me. I thought I was an animator.
Mike Well, you had done art pieces.
Alvy But not character animation. We wanted to be character animators. And I thought I could do it. Okay, so we're coming back, Ed and I are flying back from the 1983 SIGGRAPH, the big annual conference.
Alvy And on the plane we're chatting and we decided that at the next SIGGRAPH, the 1984, a year hence SIGGRAPH, we would announce to the world that we were character animators. Distinguish ourselves from all those flying logo people and advertising station shot people.
Mike I mean, technology moved on a little, but the Apple Macintosh hadn't even been released at this stage.
Alvy Oh no, those were all toys. We never paid any attention to those little machines.
Mike This is all. Yeah, this...
Alvy These are the so-called minicomputers.
Mike Yeah. This is all...
Mike Industrial stuff. Yeah.
Alvy VAX, we had PDP-11s and VAXs and things like that. You know, half a million dollar machines. Everything was very expensive and very slow.
Mike So is this another moment of you articulating a vision in a very concrete way that you had no idea how you were gonna pull off?
Mike I mean you did pull it off eventually, right? Having a clear vision seems to be an important trait to get stuff done.
Alvy I keep coming back to having a clear vision is, now I see how valuable that is, and it's one of the hardest things to come up with. And we had this clear, we wanted to be the group to make the first movie with the computer. It was really simply stated, we didn't know what it meant, it could've been...
Mike So from New York.
Alvy ...two-dimensional. It could've been three-dimensional. We didn't care. We just wanted every frame to be done on a computer.
Mike Now Lasseter was, of course, at Disney as an animator. So he had the right credentials.
Alvy Oh, yes, he did.
Mike But he was not the rock star that he became. He was a junior.
Alvy No, he just, but he was good. He introduced himself to Ed and me. So Ed and I would make secret trips to Disney every year. We always thought Disney should be the company supporting us. Why this crazy man on Long Island or why George Lucas, it should be Disney. You know, we'd grown up with Walt himself, teaching us animation on the weekly television show.
Mike On the Sunday night, yeah Disney's weekly show, whatever that was.
Alvy So we would take trips out every year to visit Disney and kind of say, are you guys interested? And they always had some reason why they weren't interested. I remember the first year it was, can you boys do bubbles? Well, what they didn't get was a computer can do anything, it wasn't a bubbles, and smoke and...
Mike The bubble button.
Alvy All their other technical people got it, who we were. And they told us very frankly, you should be here. If Walt were still alive, you would be here. Or if Fireworks were still alive, you would be here. But the guys in charge were, you know, it was a football player who was in charge. Ron Miller had married Disney's daughter. And that's how he earned the presidency of the company.
Mike So you're still at Lucasfilm and you've gone down on one of these secret trips to Disney to presumably just check in and see whether they are interested in commissioning a whole lot of amazing animated feature film work.
Alvy As usual, they're interested, but nothing happens. But this time, this young kid who was working there, John Lasseter, invited Ed and me down into the archives, the Disney archives. Here's this bright kid who had already won the, I think it was called the Student Animation Academy Award, or the Student Academy Award at California Institute of the Arts, which was the college that Walt Disney himself had founded to train the animators for his own studio. John was the champion animator from a really high-powered class of animators including Brad Bird and.
Mike Tim Burton.
Alvy Yes, Tim Burton. So he's just, are you guys interested in the archives? Well, we grew up, we were aficionados, you bet we're interested in the archives. I remember going down to the archives with John. And he said, what do you want to see? And I said, anything? He says, Yeah, anything. I said, well, I want to see the dancing hippo from Fantasia.
Mike You don't mean the footage of it, you mean the actual drawings.
Alvy I mean the original drawings by Preston Blair, the actual great animator, John looks up chart and goes over and finds a manila envelope and does that thumb thing that animators do, pages flip by and there's Hyacinth the Hippo dancing, the original Preston Blair drawing.
Mike Preston Blair would literally write the book that every animator bought to learn.
Alvy I learned from his book. That's why I thought I was an animator. He didn't teach you that crucial thing that you have to have. Animators can't tell you what it is. They just have something special.
Mike And so you and Ed ended up obviously hiring John in, but you're still at Lucasfilm at this point.
Alvy George Lucas wouldn't let us do animation He says only Disney can do animation. And so we realized John can be visible. So we hired him as a User Interface Designer.
Mike So for the second time in your career, you're in what is deemed by the rest of the world to be one of the most imaginative places on earth, and your vision of the future doesn't sit with the vision of the future of the visionary person that's kind of running it.
Alvy See I never quite saw it in perspective, as clearly as you just said it. But yes, that's true.
Mike Let's say you see the writing on the wall. You've got a clear view of the future, and obviously, Ed. So how do you get from being there, producing terrific work, but obviously not your film yet, to having Pixar, which does go on to deliver on your promise of 20 years of making the first full CG animated film?
Alvy Well, let's see. George got divorced and that's when I went to Ed and we just said, okay, let's start a company. Gotta make sure you understand, we're two computer nerds. We're not business guys at all, right. We're just two nerds.
Mike Rumour has it you actually crossed the street and bought a couple of 'how to write a business plan' books..
Alvy We went across the street to a bookstore in Marin County called A Clean, Well-Lighted Place for Books. I remember the name of the store because it's a good name.
Mike And you literally bought two textbooks each...
Alvy We bought, he bought two 'how to start a company' books, and I bought two other others. So we bought four 'how to start a company' books.
Mike This is your business school education.
Alvy Yes. And I did call my buddy Jim Clark, who had been with us on Long Island. I was born in Texas. He's a Texan. We grew up back 60 miles apart. So that's like next door in Texas. So we were homeboys. And I said, well, here's a guy who's actually started a successful company. I'll call him up and ask him how to do it. And he said, he's very Texas, he said, well, Alvy, it's really easy, he says, all you have to do is you gotta learn the terminology, it'll take you about a year and then you'll have it. It's not that hard. And I should have noticed that he didn't volunteer to invest!
Mike So you go and produce a bunch of business plans in that kind of classical sense, because I think this is an important story point. History sometimes, maybe at the Wikipedia level, implies that Steve Jobs bought this from George Lucas, 'cause they were best mates.
Alvy Oh Steve had very little to do with this. He saved us investment-wise.
Mike But it was your company.
Alvy Yes. So Ed and I had this idea, and we wrote up the business plan. And by the way, it wasn't to make movies because we knew Moore's Law really well. And we knew that our computers, we needed five more years of Moore's Law development, just standard Silicon Valley development, was going up by an order of magnitude, a factor 10, every five years. And we were still a factor 10 shy of having cost-effective computers. What are we gonna do for five years while we wait for the computers to achieve computation?
Mike You're articulating the vision very clearly: the future is coming, I just know exactly when it's going to hit, and it's not for like, X number of years.
Alvy The thing about Moore's Law is it's been right on schedule all along.
Mike Did you know who Steve Jobs already? Had he visited?
Alvy Yeah. Alan Kay who is an old friend who hired me in at Xerox PARC. Alan Kay had gone to work for Steve Jobs as the chief scientist of Apple. And Steve showed interest in computer graphics. And so Alan said, well, you've got to go up to Lucasfilm and see the best guys in the world. And he did come up, Steve did, and Alan. And Steve gave a presentation, but he almost immediately got fired from Apple.
Mike Right. So it's before the firing. He's legendary, but obviously still to go through the trip that takes him off to do NeXT, and before he comes back to Apple.
Alvy That's true. Right, so this is the first time at Apple.
Mike And he was a internationally recognized, kind of, industry rock star. I mean, he was a personality.
Alvy Oh, definitely. And I'd met him once before that, too at a design conference on the Stanford campus, we had set up the same table and chat. So this is my second in gallery. So, you know, we were there in the valley. You get to know people pretty quickly. All right. But so Ed and I wrote the business plan for a hardware company because we thought, well, we had built a computer called the Pixar Image Computer for George Lucas.
Mike With transputers, the first chips I learned to program on.
Alvy Really? I think that's forgettable.
Mike I do know though, that they actually went back and found one of your original Pixar image computers. And it sits now at the front of the render farm at Pixar.
Alvy We were terrible at this hardware business, but we did it. It kept us alive for five years, and got us an investor.
Mike And if I'm not mistaken, when you got Steve to invest, he invested 10 million, being 5 million for getting the licencing rights to the tech, and five for working capital.
Alvy Yeah, it's was the working capital, right. Ed and I turned the first cheque for five right over to Lucasfilm.
Alvy To purchase our rights to the technology for everything we had developed there.
Mike So it's an amicable parting with the guys at Lucas.
Alvy Yeah, it was. I mean, George was unhappy because he said, Steve's gonna get all the credit, and I did all the work. He said that to us as we were departing.
Mike So now you're with another visionary.
Alvy Oh we went through 45 funding opportunities before we got to Steve.
Alvy 45 people turned us down.
Mike And I think one of those almost had you at General Motors, right?
Alvy General Motors was one of them. H. Ross Perot, ran for president.
Mike Had Ross Perot not run his mouth off and upset GM's Board, we would now be talking about the famous car company.
Alvy Boy, I had troubles with Steve Jobs, I can't imagine how much trouble I would have had with H. Ross Perot.
Mike But these are legendary figures again, but you're living this from the inside on a one-to-one basis, because obviously you are vice president, I think. And Ed was president.
Alvy Ed Was president and I was executive vice president.
Mike Right exactly. And so, of course, at board level, he's your major investor.
Alvy Yeah. The three of us are the board. Yeah. So I don't have anything positive to say about Steve except one, and it's a very important exception. He came through with the money when 45 other outfits turned us down.
Mike And he would continue to come through with money because that initial 10 million actually blossomed to about, to 50, right?.
Alvy Oh we burned through that, nothing left. Yeah, when these things happen is you look back and say, I don't know why they should have happened that way, but they did. So we kept running out of money. We were not a good hardware company. We kept running out of money, his money, you know, we'd just burn right through it. By the way, he's running NeXT at this time, hour and a half away. So it's Ed and I, you know, we're running the company and we're running it into the ground. We're just not any good at being a hardware company. Well, the product was not the right product as usual problem with a startup company.
Mike You were a bit ahead of the curve.
Alvy I think it was just a mistake, to be honest. Moore's Law was going to solve this problem, so.
Mike I can argue with you about that. But the parallelism that you are adopting in having the fast pipeline in the pixel...
Alvy Yeah, by going RGBA alpha in parallel, we've got a four time speed up. It's not an order of magnitude, but is four times, it's pretty good yeah.
Mike That parallel approach is echoed today in computer designs. I think you're being a little harsh on yourselves, I think you guys did a better job than you're willing to admit, but okay, lets...
Alvy You know, basically Moore's Law just left our Pixar Image Computer in the dust.
Mike Sure, Okay.
Alvy So we would run out of money and we'd go to Steve, and well Steve wasn't about to admit failure. He had been kicked out of Apple, remember? And this is his first investment after Apple. He could not sustain the embarrassment of failure. So he would tear Ed and me apart. But he'd write a cheque and take away our equity. So after about three of those refinancings, he owns everything, but we still have our company.
Mike It's funny 'cause he's stuck with you through thick and thin. But by the same token, I think you once said that if somebody had offered him 60 million for his 50 million investment, he'd have also taken it.
Alvy He would've gone for 50 million. I know because I wrote the business plans to offer these guys.
Mike But you said he did one really good thing, which is coming up with the money. But also the other thing that he did, and I think it was after you left, is he took the company public in an astonishingly early stage.
Alvy Well, that shows you what a brilliant business man he was. I don't think he's the visionary that people have made him into, but he was a marketing genius. He would see a good idea, and he would just make it his. He invented the Mac if you would believe it, and he invented the iPad if you believe it, and he invented the iPhone if you believe it. He didn't invent any of those. But, boy, did he sell the stuffing out of them, because he's really, really good at that. And his brilliant idea on Pixar... Okay, so while we're building these machines unsuccessfully, and running through all the money, 50 million or so, you know, Moore's Law finally got there and Disney knocks on our door and says, 'let's make that movie you guys always want to make, we'll pay for it'. So, they save Steve. They paid for the movie, and made him whole. So we start working on the movie. And when it was done, took it to the critics in New York City and they went nuts. They said this is gonna be a huge success. And soon as Steve heard that, boom, he pushed Ed aside so he'd be running the company when the cameramen showed up. And he had the brilliant idea to take the company, which had no money at the time, public. It was a really bold move, and it made him a billionaire overnight.
Mike And to pick up on your point about him being able to identify technology, you obviously famously fell out with him. But no matter how much you fell out, he still actually invested in your next company. So even though he clearly had a problem that ended up with you leaving the company, he was still saw in you...
Alvy I don't think it was that clean. If I'd gone to him and asked him to invest in my second company, he wouldn't have.
Mike But he did.
Alvy Well, he did indirectly. He did...
Mike And quite frankly you made him whole again, as you like to use that expression, because he, he did rather well out of it when it was sold to Microsoft.
Alvy So when we sold it to Microsoft, Pixar owned 10 percent of the company. Well, he was Pixar.
Mike Yeah, sure.
Alvy By this time. Bought all the rest of us out. So it went to him because he was Pixar. If he had had to make the deal. I think he would not have allowed it to happen.
Mike So you get to see firsthand that management style that you found abrasive from Jobs, you're now at Microsoft though, because you've left and you've sold.
Alvy Before you're going further, though, Steve never managed us.
Mike He was your investor.
Alvy He said he created Pixar and ran it, but he did not ever run it. He ran NeXT, and Ed and I ran Pixar. I mean, he never had a single say. We never allowed him in the building, for one thing, because he created havoc. And John Lasseter never allowed him in the story room because he created havoc. He was the money, and he got rewarded handsomely for that, because he was brilliant to take the company public when he did. That was a brilliant move.
Mike I guess my point was though, that you interacted with him a lot is what I was trying to say.
Alvy Yeah, in board meetings all the time, yeah.
Mike And of course, and I'm just skipping forward a little quickly. But you now end up at Microsoft because you've gone out, you've set up your own company that makes this, what I like to think of as what could have been the number one competitor to After Effects, effectively.
Alvy It could have been, it could have been a Photoshop eater. Microsoft didn't want to be competitive.
Mike Yeah, so you go to Microsoft, which in the history books is the number one arch-enemy of Apple, and therefore you've got Gates, as history paints it, the number one chief antagonist to Jobs. I mean, they're very few people that had your experience of saying these up close. So can I now peel back and say, what's the future look like inside Microsoft, was the Microsoft vision...
Alvy Microsoft was a total surprise to me. On my board of advisors at Altamira (my second company) was a man who was an adviser of Bill Gates, Gordon Bell. And he said, Alvy why don't you go up and I'll introduce you to the people at Microsoft, and they can give you marketing assistance. So I went up to introduce myself to get the marketing itself, the part I don't understand about business as marketing. And I meet this guy, Nathan Myhrvold, who's like second only to God, God being Bill Gates. He and I really hit it off, just Vulcan mind-meld. We were trying to out-talk each other on quantum mechanics, high-class photography, and somewhere I said, hey, Nathan, this is great, but I'm here to do some business, you know, I'm looking for marketing assistance. Nathan says, why don't I just buy that company? And I sort of felt like, ahh ah ahh, it's not me, it's the company. He says, I got it, that's the company, not you. I said, well, I didn't come here for that, but I can run it by the board and see what they think. And they said, yeah, let's do it! And so, to me, Microsoft was famous for not having any marketing sense whatsoever. It was about the nerdiest company on the planet. They didn't care what you thought about them. They were good, they knew they were good, and if you didn't see it, tough.
Mike When I was looking from the outside in, I would have painted Microsoft as uncool, kind of evil empire, not particularly innovative, not particularly imaginative, and certainly not...
Alvy Do you know who painted that story? Steve Jobs. Since Microsoft wouldn't paint their own story, Steve just jumped into the lurch and said...
Mike But, this wasn't your experience?
Alvy No, I got inside, it was some of the sharpest programmers I had ever seen in my life. They tested every line of code multiple times. They thought out the product from scratch before ever laying a line of code. It was very innovative. It was so complex. I remember watching in an interview one time Bill Gates was about to announce a new version of Windows, I think, and a reporter was interviewing, and said tell us something innovative in Windows. Well, Bill Gates is just straight nerdy guy. He just told her exactly something that was very innovative, but of course it went right over everybody's head, including this reporter, who then reported that he was this arrogant bastard. And I sat there and said, no, he told you exactly something that was innovative and you didn't get it.
Mike And that they shouldn't employ you as 'a guy'. They made you a graphics fellow, which was a title that until you yourself chose to violate it was to be reserved only for you.
Mike It was quite a hallowed kind of, you didn't just land there, they welcomed you with open arms.
Alvy It was a good deal. They made me financially successful, let's put it that way.
Mike So we've got now Xerox PARC couldn't see the future, and you moved, and you obviously had a similar kind of thing with Lucas to a certain extent. Was Microsoft open, because again, looking from the outside, some of the great stuff you brought that didn't go anywhere.
Alvy No. So I thought Microsoft didn't use what I brought them. And so basically, I had golden handcuffs, my options vested at four and a half years. So I stayed 'till the four and half years is up, and then I left because it was clear to me they weren't utilizing what I was bringing.
Mike Why do you think they couldn't see the future? Because the stuff that you brought them was, in retrospect, exactly what is used worldwide now.
Alvy I still use the product that I did for them, every day.
Alvy I use Photoshop, because that's the one that's out there.
Alvy But a lot of things you can't do in Photoshop, or it's really hard to do, which are easy, and a snap to use in my product, which is called PhotoDraw, and I use that every day.
Mike But how could they not see that?
Alvy Microsoft is a very complex company. It wasn't one giant company that executed business plans, there were these little fiefdoms, hundreds of fiefdoms, each doing their own products. When I came in, I said I've got an idea of how to bind all the image products of Microsoft into one giant scheme. They could care less. They just were interested in getting the next product out, which they were awesome at. I watched them take my product and generate five commercial products from it in five years. It was awesome
Mike So that was good productization.
Alvy Oh, they are awesome. And I realized everything I'd heard about them was just wrong. And they also didn't want to be seen anymore as the big bad guy. I thought I had a Photoshop-eater, I did. They did not want to be seen as an Adobe-eater because you know everybody thought they were the big, bad, evil empire.
Mike So let me ask you some broad questions, if I could, and I'm interested to get your opinion. If you're looking back now at these huge moments of technology and shifts in the industry, what would you characterize as your, not financial, but sort of greatest success? Like, what was it that you sort of look back on as being the pivotal successes?
Alvy Well, I'm so proud of Pixar.
Mike The culture of it, or the output of it?
Alvy The culture of it.
Alvy Yeah, that's a good question. It's the culture of it, and the culture of it is: everybody's the same. That sounds like I've said nothing. But the artistically creative people are just as important as the technically creative people. I haven't found that anywhere else. Microsoft, it certainly wasn't true. The technically creative people were heroes at Microsoft. But if you did art or marketing, which is considered one of those lower forms, they could care less about you, they look down their noses at you. And I've been in graphic studios where it's just the reverse. The artists were the kings and the gods, and there were these ‘technoids’, they call them, who did the technical stuff. That is so wrong. So the culture that I'm really proud of at Pixar was, yes, there are different types of creativity, no question about it. But they have to work together in mutually-admiring relationships. And so everybody had equal everything, promotions, salaries, dignity, health plan. I mean, everything was exactly equal.
Mike So when you started out, you were in a world where we didn't have very good track records at making software. And today, a lot of companies adopt Agile and Scrum and these modern programming techniques that are non-hierarchical self-organising teams. Do you view that transition of how people approach software as something that somehow you were echoing it or pre-empting?
Alvy Now, that's too big an idea for what, we just did. We weren't cresting a vision like that.
Mike Because a lot of the things you talk about in terms of how you respected the individuals, and how you didn't have this hierarchical stuff, are modern management approaches.
Alvy Good. It's really good. It's a wonderful, comfortable environment if you feel like you're being celebrated for what you're good at.
Mike Yeah, and the person that's closest to the job knows the most about it, rather than being sort of told what to do.
Alvy That might just be a reflection of Ed and me. I'm no good when people are ordering me around. That's probably why Steve and I didn't get along. You had to be his slave and I refused to be his slave, so...
Mike If you were teaching business students today, what would you want them taught or what would you teach them?
Alvy You know, I get asked, given opportunities to generalise like that, and I really don't know how to answer. Figure out what it is you really are good at. Become extremely good at it, and then go where the action is. So I say generalities like that, but I don't know how to...
Mike Seems to me one of the things that you guys did very well, and you know that I mean no disrespect in saying this, but you guys failed a heck of a lot, you never were failures. Does that make sense? Like you hit these roadblocks a lot, and yet you never acted like you were failures and you kept the faith, as it were, which is, ah I don't know what that was.
Alvy Well, I think it's luck. If Steve Jobs hadn't come along, we would have, we thought it was over when General Motors turned us down, because we had almost closed that deal. We were in a 40 somethingth floor of a building in lower Manhattan, and 20 people around the table and everybody saying yes and shaking hands. That's a done deal in the business world. Only thing we were missing was, overnight, the lawyers had to turn the verbal agreements and the paper, and we had to sign the papers. Overnight, though, was when the Ross Perot news broke that he had insulted the board of directors of General Motors. And everybody knew that anything that had to do with General Motors and Perot was dead, and our deal was right in there. Ed and I were just, now, what do we do? We're driving back in a limo to the airport, just miserable because we'd run out of possibilities. And that's when we came up with the Hail Mary of calling Steve.
Mike All right, so you say that it's hard to do the generality, so in the specifics how would you advise somebody that finds themselves in this position that you've seen or found yourself in where you've got a clear vision of the future? But quite frankly, the people around you don't share that vision or just can't see this quite, to you, obvious next step. Is there any advice you have about how to, not to sell that idea, but just to work in an organization to get that future realized?
Alvy All I can think of is trivial comments like keep trying. But the thing, it's just luck. If Steve hadn't said yes when we came over our Hail Mary plan, that we've been it. We would've been dead. We didn't have a backup.
Mike I think the other thing that I noticed when I look at this, is it's actually been really important to your career, all of these people that have travelled up with you in their own respective spaces. Like it seems to me that John at Adobe, or Alan at Apple, or wherever they are, the people we've mentioned in this story. Of course, they all independently had their own careers. But you kept a generous nature with all of them that meant that at various times they re-entered your life and helped to go to that next step. It seems to me that that idea of having, it's not like it's your crew or your particular like close, tight, tight friends, but just taking this sort of professional, dare I say, almost academic respect for colleagues, as everybody rises through their careers.
Alvy I like the academic respect part, I keep coming back to: academia was our model. You got respect for your ideas, and the money just came because if you've done a good job, money comes. I can't imagine going to a business school and saying don't be interested in the money, but I think that's what I was actually saying.
Mike What do you think the future is of entertainment and stuff moving forward from where we are now?
Alvy OK. So two things. One is I'm always kind of disappointed how my technology is being used in the movies. I'm sort of sick of exploding robots, and you know, okay, that's a little boy stage, we've got to get that out of our system. But maybe it's not. Maybe it's permanent. That's just not my world, I should just shut up. I like to see my technology used to tell stories that can't be told otherwise. The first time I saw it happen was in Titanic. They sunk the Titanic. Well, you can't sink the Titanic on a stage. I saw an early movie where they did, and it was so obviously a toy being sunk in a tank that, okay. They did it. But that wasn't the story. All those teenage girls that went to that movie weren't there for the computer graphics. They were there for what's his name, the romance. They were there for the story, right. All right. So that's one. What's out on the edge? I'm tracking very carefully VRA, AR and MR. I'm advising a little startup company in the Valley. It's a Chinese-American woman CEO, which I like helping that out. Who's got a studio called Baobab Studios, it's winning prizes right and left for virtual reality. She says, Alvy I want to be the Pixar of VR. That's her charming, the way she seduced me into being her advisor.
Mike The elevator pitch.
Alvy So course that got me interested in the state of augmented reality. And then the really hard problems, mixed reality, I call it where you take an internal model following the central dogma, of course it's Euclidean and all that. And render a computer graphics world into your goggles, and you bring in the real world at the same time, and you deduce what the structure of the real world you're looking at must be. So you can combine the synthetic world of the computer graphics with the assumed structure of the real world, so that the synthetic things can actually sit on the table and go behind objects in the room. If it's transparent glass, it'll show through them. This is a really hard problem, and people are working like crazy on in places like Magic Leap and Microsoft HoloLens, and other places are working at it. That was the hottest thing at SIGGRAPH this year. And I said, god I can see that could keep a you know, decade of people excited in trying to solve those problems.
Mike Central to a lot of that has been machine learning, or we sometimes call it AI, but I want to avoid the idea that we're talking about singularities. So machine learning, you've seen that kind of AI, machine learning push, fail in the past and you've been a bit of a critic of that. Do you wanna discuss that?
Alvy Well, I started out in AI in 1965. There was this new field called Computer Science. And at Stanford they had a department called Computer Science, one of the first ones. And I went there for graduate school to learn this sexy idea, artificial intelligence. Oh, that's so cool, maybe we can figure out how this human brain works using computer models. It appealed to my romance at the time.
Alvy Well, after a couple of years, I said nah it's not gonna happen in my lifetime, and I'm going to change to something I can actually make happen like a first movie.
Mike Do you still think that?
Alvy No, I don't. This is sort of like the fifth revival of AI in my career. And this time things seem pretty different.
Mike What do you think that is? Do you just think it's Moore's Law, or is it tech catching up? By the way, do you still believe in Moore's Law, or do you think it's run its course?
Alvy Oh absolutely.
Mike You think it's going to keep going?.
Alvy It's still going. And the thing about Moore's Law is, it's going to hit a trillion in 2025, just a few years from now. After that? Not clear whether it's got extended life, but the very nature of Moore's Law is you can't know how to extend it until you get there. I've seen it's death predicted four times in my lifetime. And every time we get there, the engineers just blow past it, because once they're there, they can suddenly see how to take it to the next step, but they can't see it until they get there.
Mike So this change that you perceive in the current AI machine learning that's delivering significant results, that's a Moore's Law phenomenon? Or is it that we've got better approaches?
Alvy It's Moore's Law to a large extent, but also we're figuring something out. Alison Gopnik, my wife and I were visiting Cambridge, England...
Mike I just want to establish your wife is actually a professor, right, at Berkeley?
Alvy Yes, doing AI these days. And an old colleague of mine, John Bronskill, who made his fortune writing filters for Photoshop, in other words, he's a pixel packer like me, came up to me. He says, Hey Alvy, we don't have to program anymore. I said, what are you talking about? He said, read this and handed me a paper which happened to be from Berkeley, which is where I live, where my wife teaches. And it is what I call the horse/zebra paper. They had taken a neural net, they had shown it 10000 random pictures of horses, arbitrary counts, colours, arrangements, unlabelled, just pictures. They had also given it 10000 pictures of zebras, also unlabelled, just arbitrary arrangements. They had a certain training system. The result was you can hand this trained neural net an arbitrary picture of horses, and it would hand you back that same picture where every horse had been turned into a zebra, or vice versa.
Alvy I looked at that and I said, you know, that's not even a well-defined problem. How can you? How can it? He said, we have no idea, it just works. And he says it's too complex to reverse engineer and all of a sudden it hit me. I says, you know, I've always assumed I would understand how intelligence works soon as we explained it, but I'm starting to wonder if that's true. It could be that, I don't know, maybe we're into the really esoteric parts of computation theory that Turing told us were there, and that we've avoided all these years. When you program, the operating system keeps the program from writing on the program. So there's the data over here, and the program can work on the data, but they can't compute on itself. Even though computers from day zero have been able to compute on themselves, we don't allow it because all kinds of havoc happens if you let a computer compute on itself. But these neural nets are computing on themself, I think, and I have a feeling we may be getting in to a realm where it's just so complex that you cannot rationally figure it out. But I'm talking through my hat here.
Mike No, I just think it's so great to see someone with your experience and your perspective on life so enthusiastic about what the future holds.
Alvy Well, I think back, what is it I really want to know more than anything else in the world? I want to know how this works.
Mike You're pointing to your brain.
Alvy I want to know what consciousness is. Come on, it's the most important thing in my daily life. What in the hell is it?
Mike Consciousness is the thing that everyone assumes we know what it is and no one actually knows what it is.
Alvy Nobody has a clue what it is. It's like, there are several things I have no clue about, by the way. One of them is I don't know how animators do what they do. They can make you believe a stack of polygons is conscious, and feels pain, and it conniving or, you know, like trying out with Wile E. Coyote, and it's so good that you are convinced, I mean, that's their artform. Actors do the same thing, they convince you they're somebody completely else, and you believe it. Or they're not a good actor. I don't know how engineers have figured out how to make Moore's Law go to the next order of magnitude. I can't figure it out, but they do that. And I can't figure out great programmers. You know, a computer is just a bunch of dumb steps, one after the other. And a computer program is just one dumb thing after the other. And yet, each step is meaningless but the result of a programmer's creativity is that long list of meaningless steps does something completely meaningful, like create a Pixar movie.
Mike Well, I just can't thank you enough for taking time to sit down with us and talk. It's a great pleasure, as always. Thank you so much.
Alvy Oh you asked very good questions, it was very good. You've done your homework, obviously.
Sandra This podcast with Alvy Ray Smith was recorded at the University of Sydney Business School. The sound editor was Megan Wedge, and the producer was Jacquelyn Hole. If you would like to see some of Alvy's visual effects or if you want to know more about some of the people and work mentioned in the interview, we have included the links in the shownotes. A special thanks to Mike Seymour for conducting the interview, and I hope you all enjoyed it.
Outro You've been listening to Sydney Business Insights, the University of Sydney Business School podcast about the future of business. You can subscribe to our podcast on iTunes, Stitcher, Libsyn, Spotify, SoundCloud or wherever you get your podcasts. And you can visit us at sbi.sydney.edu.au and hear our entire podcast archive, read articles and watch video content that explore the future of business.