Ed Catmull on the Metaverse

Announcer:

Today on Building the Open Metaverse:

Ed Catmull:

Artists understand constraints. The constraints change over time. In the case of technology, the artists want certain things. Whatever the constraints are, frankly, they're pushing on them. They know they're there and that often those constraints require them to be more creative about it, but they're still going to ask for things to sort of change where the boundary is.

Announcer:

Welcome to Building the Open Metaverse, where technology experts discuss how the community is building the open metaverse together, hosted by Patrick Cozzi from Cesium and Marc Petit from Epic Games.

Marc Petit:

Hello, everyone. Welcome to our show, Building the Open Metaverse, the podcast where technologists share their insight on how the community is building the metaverse together. Hello, my name is Marc Petit from Epic Games and my co-host is Patrick Cozzi from Cesium. Patrick, how are you today?

Patrick Cozzi:

Hi, Marc. Hi, everybody. Doing great. We have a very special episode today.

Marc Petit:

Yeah, absolutely, very special. Today, we have with us Ed Catmull. Even though Ed does not require an introduction, I'll do a quick introduction before we hear from Ed.

Marc Petit:

Ed was a co-founder of Pixar Animation Studios and President of Walt Disney and Pixar Animation Studios. Previously, Ed, you were the Vice President of the computer division of Lucasfilm, where you managed development in the areas of computer graphics, video editing, video games, and digital audio. You ended up founding three of the leading centers of computer graphics research, the Computer Graphics Lab at NYIT, New York Institute of Technology, the computer division of Lucasfilm, and ultimately, Pixar Animation Studios. Ed, welcome to the show, but please first tell us in your own words about your journey through CG and to the metaverse.

Ed Catmull:

Well, I started in this in around 1970, where I was returning to graduate school intending to study AI or languages. But, my first course was in computer graphics taught by Ivan Sutherland. That course changed the course of my life because now I could see where one could combine art, because I liked to draw, and combine it with technology. The reason I switched to computer science in the first place, incidentally, was that while I was graduated in physics, I wanted to be at the frontier. Now, it was clear that computer science was at the frontier. Incidentally, now it's 50 years later, it's still at the frontier. It's very amazing.

Ed Catmull:

But, Utah had a great program of trying to advance the state of the art. We pretty deeply understood the implication of Moore's Law and having to do with this exponential growth and what the rate of growth meant. With that, I worked on systems for displaying curved surfaces. I also came up with a system for modeling curved surfaces, which wasn't part of my thesis. Later, I published on with Jim Clark and now is the main modeling system that's used in graphics and in motion pictures and effects. But I went from there to New York Tech, wanted to do computer animation. At that time when I was getting my degree and close to graduating and looking for a job, nobody in academia thought that computer graphics, and this goal of trying to make an animated film, was worth pursuing.

Ed Catmull:

You couldn't do it, but it wasn't considered to be a central part of computer science so I couldn't get a job. That's why I went off to a small company around Boston and then was hired at New York Tech, which actually isn't a major center of computer science, but it was to do computer animation. There are a number of people joined us, including Jim Clark for a while. Alvy Ray Smith came, did the first paint program. We did a lot of original work because we had more compute power for graphics and more frame buffer than anybody else. RGB had a full red, green, blue frame buffer at a resolution of basically 512 by 512. It was $180,000. We had two of these. We had more RGB storage for pictures than anybody else in the world.

Ed Catmull:

I'm not saying this so much to reminisce about what it was like in the good old days, as to say that we could see, over the course of what we were doing, the phenomenal change that we were going through. That was the case. The change was happening and we knew it was happening and that it was going to keep on happening, which is still true today, but it's like feeling that change. George Lucas had success with Star Wars. He's not a technical person, but he believed that technology was going to change the industry. I was asked to come in and bring in technology to help change the visual part of filmmaking, compositing and so forth, digital audio, video editing, and then added later games to that. With each one of those, we brought in people who were an expert in those fields and we were growing, we were figuring things out.

Ed Catmull:

But however, in this case, we were now with professional filmmakers and funded by the only person in the entire film industry who thought this was important. For the rest of the industry, this was inconceivable. Actually, the point was they couldn't think of why they should try to conceive of it. That's how far off they were. George was an exception to that. Then due to financial things at that company, while we made a lot of progress in each one of those areas, we needed to get sold. Steve Jobs, who had just departed Apple under unfortunate circumstances, but he acquired Pixar, but he understood that we were different.

Ed Catmull:

The cool thing was that we had decided back at New York Tech to publish everything we did. SIGGRAPH was brand new at the time. We essentially joined a new organization and we made the decision to publish everything. There are a couple of people with the same long-term goals, but they kept their new ideas a secret. From my point of view, we were so far away, the keeping thing secret was a bad idea. The better thing to do was engage fully with the community. We published everything. What it is, because we did that, we were able to hire the best people.

Ed Catmull:

When we got to Lucasfilm, George wanted to change the industry. He supported the fact that we participated in open community and we published every idea. Well, then we go to Pixar and we continue publishing everything that we did. This is what's surprising to people, is Steve completely supported this. He never questioned my decision to publish everything because he knew we were doing something different. He's known to be somewhat secretive, or extremely secretive, but in our case, it was completely fine. Then, of course, by the time we got to Disney, then we were successful. Disney didn't even question our desire to publish everything. But, I feel like this strong bond because my greatest professional relationships come out of this open community of SIGGRAPH.

Ed Catmull:

We struggled to begin with because the computing power that was out there still wouldn't support what we wanted to do. In fact, when we got to Lucasfilm, one of our intellectual competitors was able to buy a Cray-1 super computer, which cost $10 million. We just pulled a piece of paper and then making rough estimates in our head, we figured it was going to take 100 Cray-1 computers to make an animated feature film. That meant we needed to wait. Not only did we need to wait for that, but we also needed the cost of the computer to come down by one or two orders of magnitude.

Ed Catmull:

For us, it was develop, participate, and learn, and accrue the knowledge along the way. Ultimately, obviously that paid off. We maintained good relationships with Lucasfilm and the special effects group there. There's a pretty phenomenal journey all the way through. I can go on for a long period of time. I don't know if I've answered your question or not, but let me let you ask some questions.

Patrick Cozzi:

Ed, it's so cool to hear your journey directly from you. We want to do some geek questions, but before I jump into those, I wanted to ask you about leadership and culture. I'm a big fan of your book, Creativity, Inc. Thank you for writing it. At our company, Cesium, it's actually required reading. Everyone who works here has read it. I was wondering if you could share some of your advice with the community here on how to build a great culture and a culture of innovation within a company?

Ed Catmull:

Well, I think one of the things that's foundational is that the people who come into most companies really want to have an impact. They want to do something good in the world. Not just in terms of the technology here, but we all realize there's some fairly existential threats that we face as a society. I would say that when I started, when I was in school, there was the existential threat of the Cold War. We still have those and people want to make a mark. They want to do good.

Ed Catmull:

You start off with the assumption that people want to do something, but then you get into a company and they deliver products. Regardless of the intentions, the need to deliver the product on time and on budget kind of trumps the values of developing the people and supporting them. It's really important to keep that balance. That is, as you go into a hard project, you always should make the assumption that people want to do better. You need to support that desire, which means helping them see, at all times, and at the hardest of circumstances, that we care about their professional growth, them taking care of themselves. If we do that, then we win in the short term and we win in the long term, and they win in the short term and in the long term. There is a way of balancing it.

Ed Catmull:

But I will say, at most places, there's something about that rush to get done where taking care of people goes down a level. We have to watch out for that. It's kind of built into the system, into human nature for leaders and the need to deliver. You have to figure out, okay, what gets in the way? Why doesn't this always happen? You focus on that, then, what I’ve found, is the people rise to the occasion.

Marc Petit:

Is there one thing that you remember vividly about working with Steve Jobs and how did the Apple culture and the Pixar culture crossover?

Ed Catmull:

Well, the interesting thing to me is that I think I've worked directly for Steve longer than anybody else. I watched this arc. The real story is more like the hero's journey. The unfortunate thing is when articles and books are written about him after he passed away were frankly kind of rushed because they wanted to get out very quickly. But, as there was interviewing of people before he died, even though Steve was fine with us talking with people, there wasn't anybody that worked with Steve who was going to actually psychoanalyze him or talk about him while he was still alive. Here's the view that I experienced.

Ed Catmull:

That is when I first knew Steve, his behavior towards people was not very good. I mean, there was a reason that there were these stories about him. I saw that too. I will say that his getting removed from Apple was a huge blow to him, clearly, and was basically kicked out for those reasons largely. The first version of Pixar, when we were selling hardware, because there was no animation business and we weren't far enough along Moore's Law to make it economical, we were selling hardware. That business failed. This is a failure for Steve. That was a difficult one. Interestingly enough though, even though with that failure, he stayed with the team and the team all stayed together. It's pretty amazing.

Ed Catmull:

But with NeXT, NeXT didn't succeed as a company. It was very difficult for him. It was like failure number three in his life. He also got something from that, which was that he did put together a great operating system based on Berkeley Unix, which is now the basis of the operating system you have on your Mac and on the iPhones. So, he’s got that to work with. But along the way with these failures, he learned a lot. He's extremely smart. But what he learned was to be more empathetic, to not feel that he had to win 100% and hit everything out of the ballpark. He learned that it was extremely important to have relationships and treat the other people as partners. Once he made that switch, that he realized that, he changed as a person.

Ed Catmull:

I should also note that he got married to Laurene and he had some children. All this he could put together and he became an empathetic person. Usually you don't think of people of as learning empathy, but he did. People can learn empathy. The way he treated people completely changed. After that changed, everybody stayed with him for the rest of his life.

Ed Catmull:

The reason I tell this story is that the focus is on that sort of exotic, exciting, bad-boy behavior at the beginning, with the implication that that kind of person was the one who made the great Apple. It's not true. It was the person who went through the wilderness and learned from what he did and figured it out, because he's so smart, that this was now the person who returned to Apple. It was that changed person who made Apple into one of the greatest companies ever and most impactful companies ever. That was a result of learning from the mistakes and becoming a better person.

Ed Catmull:

This is what I saw as we got to the point where we made Toy Story. We built the company. The culture at Pixar was a unique culture and very different than Apple's culture. He didn't try to make us the same, but he learned from it and he could easily go back and forth between the two. The truth was Apple was in his DNA, but he wasn't trying to impose his DNA on the group at Pixar....

Ed Catmull:

At Pixar we had put together originally the best group of people in computer graphics, but we knew right from the beginning that by itself was incomplete. Trying to make movies is not a technical adventure. It's a combination of technology with the artists and Steve understood that also. As you know, he's always had a great design sense. He could appreciate that the design sense to the artist was of extreme value. For me, this is the outcome and that's the real story of Steve Jobs, the hero's journey.

Patrick Cozzi:

Great insights, Ed. Ed, you've invented or co-invented some of the core techniques in computer graphics, texture mapping, AA, subdivision surfaces, and rays and RenderMan. I'm thinking way back to 1974 where the z-buffer was described for visible surfaces as something that was ridiculously expensive and then what happened over time, right?

Patrick Cozzi:

Moore's Law, as you said, added a lot more transistors. Parallel algorithms were discovered and it became the standard visible surface algorithm. Now we have real-time ray tracing. I'm curious. From your perspective, have you ever been surprised about the speed of this evolution of graphics?

Ed Catmull:

The thing about that speed is because we could see it. We could also feel it. In a matter of a few years, you see everything would change and we had every reason to believe that the advance of technology and Moore's Law was going to continue. Now it lasted longer than any of us thought. Originally, Gordon Moore changed his timeframe because Gordon Moore is a very smart man.

Ed Catmull:

He knew that any exponential thing has to come to an end or has to transition, but it just went longer, even his second prediction. It's just this odd thing in that the increased power meant that more people were using it and reducing costs with increase in the utility. The economical forces behind keeping it going forward became very strong. The chip companies involved in this were now motivated to invest in new technology and new fabs in order to make them.

Ed Catmull:

You saw this cycle and people didn't know how long it would go. The result is that it's the change in the world which made the chips continue at this speed and the fact that they continued also impacted the world because of the incredible usefulness in all sorts of areas because of the lower costs and it's still happening. We've come down to the point where you've got the complete computer in your pocket.

Ed Catmull:

You're going to have the complete computer in your watch or your ear or embodied somewhere on you or your glasses. There is every reason to believe that will continue because of other things that are happening in terms of sensors and power, lightness in that way even though Moore's Law itself is no longer the driving ability here because we have reached the limit. Most people in computer science would say that we've reached it, but the other things continue to change at a higher rate.

Ed Catmull:

One of the areas where it changed at an even greater rate than Moore's Law has to do with computer graphics and what's happening with the GPUs and NVIDIA is one of the main drivers behind this. It's still happening. It's going to happen for quite a bit of time in the future. It's hard for people to conceive of it because we didn't evolve to have a mental construct around exponential change, but it's what we're seeing.

Ed Catmull:

For most people, we don't have a gut feeling of what it means for this process to continue. It's one of the reasons I think that what's happening with the metaverse is going to change because you can still see two to three orders of magnitude of improvement in image quality available in your local devices at a cost effective rate. When you increase at that rate it's going to make the user experience and the ability to interact with it get a lot better.

Ed Catmull:

When people say, "Well, okay, I don't know if this is real or not. The experience isn't good enough," the experience may not be good enough, but all of the forces aligned with it are enabling to advance it to the point where it will become good enough. It's not just toying around with it. It's going to happen.

Marc Petit:

If we put the bar as a Pixar movie, you want to take out your crystal ball and tell us how far are we from having a day to day experience in the metaverse that would feel like the latest Pixar movies?

Ed Catmull:

Presumably, one wants to cross some threshold where you feel like, “I'm really enjoying what I'm seeing and I feel like I'm interacting with somebody else.” Now we know in the game world a lot of people feel like they're interacting with a character. Plus, their friends are participating in this too. The substitution for it is such that only some people get into it and a lot of people don't. It's just the truth, right?

Ed Catmull:

Not everybody feels like they're lost in it and I'd be one of those. I go in there and I see other people in it and it's a crude avatar and one that I don't feel necessarily compelled to go into, but whatever that avatar is or whatever that representation of the other person there is going to get a lot better. This is going to happen.

Ed Catmull:

All of the forces, the economic forces, many of which serve other purposes, are coming together and they're in support of this direction. Frankly, the games industry itself is large enough that it's one of the driving forces in order to push the advance of the technology. I mean it's a big industry. It's getting bigger and it's going to cause more people to put resources behind making it better.

Marc Petit:

Let's switch gears and go back to the topic of creativity. Pixar has one of the most sophisticated production pipelines that we've seen. It can output amazing results. I've always wondered how much that complexity supports or hinders the creative process. I know artists are constrained to thrive, but how did you feel at Pixar with this problem?

Ed Catmull:

It's true that artists understand constraints and the constraints change over time. In the case of technology, the artists want certain things. Whatever the constraints are, frankly they're pushing on them. They know they're there and that often those constraints require them to be more creative about it, but they're still going to ask for things to change where the boundary is.

Ed Catmull:

That just happens and it's a good thing in terms of advancing the tools that are given to them. Early on in particular in our pipeline, it's an iterative process whether you iterate on the story, you iterate on the lighting or the modeling or the making of the characters and you iterate on the animation. Early on, that iterative process was pretty slow. We had to make crude models of the lighting. We had to greatly simplify the background on the characters so the animators could work in real time.

Ed Catmull:

In fact, initially they would do something and then wait a little bit and then they'd come back and see it in real time. As a result, there's something else we should talk about, which is the impact of GPUs, but it's gotten to the point where animation and lighting and so forth essentially happens in real-time. The closure, the final image, looks to what they're going to see when they're iterating, then the easier it is for them to make the decisions and go through this process.

Ed Catmull:

That's having an impact on everything and there are other things we do to help with the iteration. Only in the last few years we started to use motion capture to assist with doing certain kinds of mocking things up. Motion capture is used and has been used for many years effectively in live action special effects. For the kinds of things that we are doing, since they're caricatures, then we have to do a lot of subtle things that most people are only subconsciously aware of.

Ed Catmull:

That's where a great animator, because they're very astute observers and put things in there that are these subtle signals that motion capture frankly doesn't capture. Basically, it's to bring it into our pipeline. All of this is to use with other techniques to help the storytelling process and on something like Finding Nemo or Inside Out where they're highly caricatured, then the motion capture doesn't even make a lot of sense.

Ed Catmull:

You have to do everything using some sort of initial simulation of their characters. You have the dynamics of a fish or the dynamics of something. Then, that's a tool for the animators to use to work on things. There are always some challenge with each one and basically you have to make a list. What's important for the story? What isn't? What am I going to do for it?

Marc Petit:

Hopeful that real-time tools will make a big difference in terms of the story development process and that iteration speed?

Ed Catmull:

They already have and the process will continue. Already today, for instance, the quality coming out of game engines is greater than what we had for Toy Story. If you compare the two, you can see the difference. The story quality on Toy Story was still extremely good and I would say that the process of working out the story hasn't really gotten faster. That's more of a thought process.

Ed Catmull:

After all these years, the amount of time it takes to make a film is basically the same. The difference is you can make an inexpensive film much more cheaply because you can do it so fast, but if you shortchange the story, what you have is a high-quality looking image that’s not a very good story. We've tried all sorts of things and basically all that matters is that we not screw up the story.

Patrick Cozzi:

Ed, in your intro you mentioned Pixar's contributions to the open community through all the publications. Pixar has also contributed lots of open source projects, particularly USD. I was wondering what your thoughts are on USD as the HTML of the metaverse and how do you think Pixar could handle this opportunity given that the scope is so much bigger than movies?

Ed Catmull:

The philosophy we've had besides participating in and publishing in the community and appreciating what we've gotten out of the community, our notion is also the we're a pretty small company and that we can't do everything. Our animation system, for instance, is extremely rich and complex. The focus is on the things that are unique to what we're doing.

Ed Catmull:

If you take something like Paint as an example, the first RGB Paint was done by Alvy Ray Smith back at New York Tech. We had the most advanced frame buffer when he built the Pixar image computer. We had the knowledge and the expertise to build a paint system, but our view at the time was, which I think was correct, was that painting is so important that people do have to have it.

Ed Catmull:

Companies are going to build something around it. If we want to pursue our goal, we should not do a paint product. We should use what other people are going to do in any case. Now there are certain areas where we use tools, but there has to be systems for defining surfaces. That's the subdivision surfaces. We decided to make them open because we weren't building all of the modeling tools or things around it.

Ed Catmull:

We are better off if we actually open up what we've got so that people can build on it because we don't have the resources to do it all ourselves. As we tie these tools together, the way we define the models and describe the scene was also the kind of thing where you just can't do it all yourself. We are better off making it open. Then, we get challenged to how to make things better.

Ed Catmull:

Incidentally, we did the same thing with RenderMan. Originally, it was a standard for describing scenes, which 19 companies got behind to support. Then, as it came out, we're still early in this process. Even though we made it an open standard, what we found was we were the only ones really, not the only ones, but for the most part the only ones actually doing a high quality render behind it.

Ed Catmull:

While it was still available, we were producing the render. Now the group that was selling the product is located in Seattle. What that meant was they weren't at the studio because if all these smart graphics people were at the studio, they probably would've gotten subverted by the needs of production. Having them located separately protected them in what they were doing.

Ed Catmull:

The other instruction was, since they were selling to the effects companies, was that they not even asked for permission to add features because we wanted them to take care of what the needs were of other people. Now the advantage of this was that the special effects industry was asking for things that we weren't asking for. Their turnover in projects was much faster than our turnover in films.

Ed Catmull:

In many ways, they are pushing the state of the art in imaging faster than the main studio was. Because we were meeting the needs of other people who were changing faster, then by the time we needed something we already had it because it was meeting somebody else's needs. They continue to do that. For me, it's pretty cool and for us that's the challenge. If you do something and you make it open source, then other people are going to ask for things that push and you want to always put yourself in a position where you're being pushed and challenged.

Ed Catmull:

It's the solving of the problems that actually requires you to be creative. That's why I define creativity to be broader than just what people think of in terms of the arts or the sciences, those kinds of skills, or writing, right? It's really solving problems. You've got a hard problem. What do you do? How do you make it safe for people to solve the problem and be engaged in it? Basically, you want people to be challenged and take on something difficult, whatever field that's in.

Patrick Cozzi:

Very cool. Ed, can you share with us, how do you think big companies like Disney approach open source? Do you think they see value in it as well?

Ed Catmull:

Well, they vary in what they are. Disney itself is not really known as a technology company. They have several groups within the company who understand technology. So the notion of open source is kind of a new one to the corporation. And the question is, "Okay, what's the gain?" So I can define what the gain is easily for Pixar and for Disney animation and for Lucasfilm, because Lucasfilm was doing open source material even before Pixar was. It's easier to explain the gain. For the corporation, they kind of say, "Okay, how does this help us?"

Ed Catmull:

Part of it is actually trusting what studios are saying, because there are a lot of very good technical people in various places in ABC or streaming or ESPN. It's basically making the mental transition over to what it means in a world where technology continues to have an increasing part of most parts of the pipeline. Now, I'm not there anymore as you know. I retired from Pixar, so I hope it goes the right direction. I was completely supported when I was there. The great thing about it, is like "Okay, we should do this." I said, "Okay, let's do that.

Marc Petit:

Yeah. The contribution of Pixar through USD has been quite enormous. Let's go back to creative processes. We're seeing the most important industries convert to the same set of real-time tools in movie making, production, game development, live events. Animation is starting to embrace the same real-time tools. Let's assume for a minute that we achieve that dream of sharing assets. You create a world and that world can be used as is, for a game or for a movie. Does this mean for the creative process, does this open ways to do different kinds of entertainment? How do you fuse game mechanics and story arcs? What do think people create there? You've been leading the Braintrust, the most creative team. You think that there would be a step forward in terms of what you can build?

Ed Catmull:

Well, there's certain things that go in a step by step process. You were talking about the Braintrust. We were fortunate in that we started off with six people who were phenomenally creative, and they were very supportive with each other. But as we grew, we added people to it, so the dynamics of the group changed and the processes had to change as we grew. It was a step by step process and recognition of how things went. We were trying to be very introspective at times. We weren't always introspective. You're doing the job, you're not just navel-gazing, but every once, you step back and say, "Okay, is it working as well as it used to?" It basically advanced in the way it worked. We learned a lot about it, but at the same time, it wasn't like the original group, but then things always change. That's life.

Ed Catmull:

But it was a step by step process. When I say step by step, none of this is ever smooth. It goes and jumps along the way. The same is true with technology. There's an underlying trend driven by this exponential change that's going on, but there are step by step programs that are happening that aren't necessarily, in fact, frequently, aren't recognized by others, but you reach a threshold where basically things have changed so much that it results in the world changing. If I take the example of computer graphics, well, it was like the background thing. It happened in this community with a few people trying to use it in this industry. But basically, it wasn't until, I'd say, it was starting in 1989, but really in '93 when Jurassic Park came out. Then in a few short years, there was a rapid change of the industry that took place.

Ed Catmull:

It looked like this overnight thing that was disrupting and changed everything, but as a result of the building up over time. So you really have two things that are going on as the underlying things that are supporting it and changing and solving the problems creatively. The other is after they've crossed a threshold when there's a sudden change and there's a major impact that takes place. One of the problems I have with even the word disruptive technology is that it kind of implies that somebody came up with something that changed everything. Well, actually there was something that took place slowly, and then it crossed a threshold. It is in fact disruptive in that sense. I don't normally use the terminology and I don't like it because it ignores the process of things growing exponentially and people solving problems until they solve enough that it impacts things.

Ed Catmull:

When things change by an order of magnitude or two orders of magnitude or more, then it's always going to change things. But I would also change them in unpredictable ways. There are a lot of things that took place that I could not have predicted, but I just learned long ago that I want to do something, but I can't predict the outcome and that's okay. It's actually pretty exciting. That's the path we're still on. As we bring the technology into different mediums, then it alters what the storytelling potential is. What's the storytelling potential in VR or in games? Because they're different. And we saw this with AR when the Oculus came out, people don't recognize that AR was invented over 50 years ago with Ivan Sutherland when he at Harvard. There's been work at places like UNC, where they've been working on it for many years.

Ed Catmull:

And people using Silicon Graphics boxes and then various other things to do it. What the Oculus did was it got rid of the lag. That's all it did. Now, this was important viscerally because the fact that you could move your head and the picture didn't keep up with you always told your brain that something was wrong. Now, you still had some of the impact of it, but it really wasn't one that grabbed onto you. But when you had the Oculus, you got rid of the lag. Now, what that did, because it crossed a particular perceptual threshold, was it started the hype machine going.

Ed Catmull:

Virtual reality has applications as well as its AR in a whole number of areas, but probably the hardest one to effect is storytelling itself. But all the hype is about the storytelling. Suddenly, people are jumping into this new medium of storytelling. Well, frankly, with one exception. All of the early versions of VR and storytelling sucked. It wasn't very good at storytelling. Now, that doesn't mean it won't happen. It means that, okay, that's not the first good application for it, so we have to figure it out like anything else. How do you develop the language of it? How do you tell good stories?

Ed Catmull:

The hype gets ahead of the reality and then people are getting disappointed. This happens over and over again. The hype machine kicks in, people are getting disappointed because it doesn't meet what some people published or made out. In any case, it's a common thing. For me, that's where we are with the metaverses right now. The hype machine is in high gear. Now, it's still coming. But since the hype machine got so high, then people are saying, "Oh, that's a lot of hype." Well, it is a lot hype, but that doesn't mean it's not going to happen.

Marc Petit:

Yeah. Just a quick question before I throw back to Patrick. The emission industry, Pixar and the Braintrust was kind of the most influential group. Who do you think has the influence today of the animation industry in storytelling?

Ed Catmull:

Well, I think that for animation, I think that I'm actually proud of fact that Pixar still takes on challenging projects. In fact, the way we think about it, because we had to think, "Okay, what kind of movies are we making?" And as you know, with some companies, they make a lot of sequels. The thing is sequels are easy to market, and therefore they make money. After you make several of them, they can become cheaper to produce. So there's a natural tendency to do that, but if you keep doing that, it's creatively bankrupt. In our case, it was clear that we were going to make some sequels. What we said was, "Okay, a third of the films will be sequels and we would only make them if the director, the one who originated the idea, thought it was a good thing to do." That was imperative to us, is the person, whether it was Pete Docter or John or Brad Bird, they had to like it. In the case of Incredibles 2 for instance, we waited a lot of years until Brad was ready to do it.

Ed Catmull:

The other thing is because we worked that way, sequels are no easier to make at all. There really is no economic advantage in terms of the cost of making a sequel. About a third of our films are films that a director comes in that we know is experienced and will do a good job and comes up with a good idea, so we make it. About a third of them are that way. The other third of the films are very high risk. They would fail the elevator test. In this industry and a lot of industries, the elevator test basically says that you come up with an idea and you can pitch it so cleanly and carefully and explain why it's a good idea that in a short period of time, you can explain to the decision makers that you should do it. So in our view, a third of our films should fail that test.

Ed Catmull:

There is no way you can explain to somebody in a few minutes or a few hours or a few weeks or a few months about why it's a good idea to make a movie about a rat that cooks in Paris. Or an old man, he grows up, and he marries a childhood sweetheart. They can't have a baby, they get older and she dies. And he's so depressed that he ties a bunch of balloons to his house and he floats away. All right. That doesn't pass the elevator test, but it requires you, because it's challenging, to be more creative, because you have to come up with something that people are going to connect to. Taking on the challenge now is a real jolt to the system. Even though they don't necessarily work on the film, the company takes pride in the fact that we take high risks.

Ed Catmull:

So in our case, just recently we made this film Turning Red. All right, that's not an obvious idea we should talk about. I said, "Well, that's really cool." That's not the sort of thing that others would do. I'm a big believer in that. Okay, yeah, you do some things, because you know you want to make money and they're going to be easier to market, but you want to make damn sure they're good. For that middle third, well, okay, we really trust the person and it's a great idea. They should make it. We're going to support it. It's still going to be hard. It's no easier. They're all hard. And the other one is like, okay, if you solve this problem, it's going to be a great movie. And they do! The entire time that I was there, we completed 20 of 21 films. Only one movie did we actually abort after we're into it.

Ed Catmull:

But in truth, we switched directors, because we still wanted to make the film, but the director was Pete Docter. Pete Docter said, "Well, as long as I'm restarting, I have another idea which I'm pretty passionate about and it takes place inside the head of a little girl." Well, okay. What we value is the passion in the film director. That's what he wanted to do, and if that's where his passion is, we had to support. That's what we wanted to support. He delivered a great film as a result of that. I think Pixar still does that. That's in their DNA, and it's very important. Also, I think it's great because they're isolated from the rest of the entire corporation. That identity as a culture that takes risks is very important. I think it's important for companies to think about that. What's the culture? And do the people own what they're doing and feel responsibility to the point where if the leaders are screwing up, they're kind of pissed? We want them to hold us accountable so that we do a good job.

Marc Petit:

Actually, we should jump to this question. You referenced cooperation with the Walt Disney Company. Very interesting company. They're deep in the physical world with theme parks and cruises and the digital world with movies and games and TV shows. Walt Disney Imagineering, trying to tie the physical in the realm. They seem ideally positioned to benefit from the metaverse. Do you agree with that? Do you see very specific opportunities or even challenges for Disney as the metaverse grows importance?

Ed Catmull:

Well, I know they're trying a few things, but I don't know that much, because while they're working on it, they are, which is true with a lot of companies. They're fairly close to the chest on it. So I don't know either what's going on other than they're saying that they're going to do it. Well, in the case of filmmaking, when Bob Iger was made the CEO, it was an amazing story. While Steve Jobs had met him, he didn't really know him. Steve Jobs had made him the CEO, and then immediately Bob Iger calls him up and said, "I'm now the CEO, but I've noticed from the theme parks that most of the new content comes from Pixar." Because we had a contractual relationship, they did the marketing for our films and the distribution. But actually coming up fairly soon, we're going to be at the end of our seven film deal with them. Bob said, "The problem is that our animation group says they've just about fixed their problems, but I don't think they have. The only group that has actually figured out how to run an animation system is you guys. So, we'd like to talk about acquiring you."

Ed Catmull:

Now, that really impressed Steve, because Bob started off by saying, "I have a weak hand. Can we talk?" And that honesty basically was... Because it was the start of things. They developed a close bond, a very close friendship. And then we entered into a relationship where Disney acquired Pixar. And then, Steve set up something where basically I would be the President of Disney Animation, and I had a dual report to the head of the studio and to Bob Iger. We also put together a steering committee to watch over Pixar, which included John, but also the head of the studio. Bob Iger and the CFO of the company, which was an incredibly powerful thing to do, because what it did was it let Pixar grow and protected Pixar.

Ed Catmull:

I would say that throughout this whole thing, the entire 13 years there, Bob Iger was incredibly supportive, and Pixar was like this sort of island within this organization. But it also meant going back and forth between Pixar and Disney Animation. My approach to Disney Animation was, those people are very good. It's just the philosophy, how they make films, needs to change. The assumption was everybody wants to do well, so let's put in place something like Pixar, but it can't be the same and shouldn't be the same because they had different personalities. And we don't want it to be a clone. I set up the rules that neither studio could do work for each other. We enforced that because it was challenged a few times. You can't do work for the others.

Ed Catmull:

Now, you can talk with each other. You can beg, borrow and steal technology, but you don't have to. The pipeline at Disney was completely different than it was at Pixar. The only thing we put in place really was to have the equivalent of a brain trust, which they call a story trust. And then it was this awesome thing to watch these people flower and do things. You end up with Big Hero 6 and Zootopia and Frozen, which were phenomenal films because now these people were able to do things on their own and really add creatively. The other thing, because the logical thing is if one company acquires another, is to say, "Well, let's merge them together within the same business."

Ed Catmull:

There are people in Disney who asked that question. Why aren't they merged together? My view was, which I could articulate, was that in an area that's changing quickly, we can change faster if these two groups have different opinions about how to do things, and if they don't feel an existential threat from the other. And the result of that setup was, on their own, they organized something they called “Dis-graph.” This is self-organized. They get together. Then they included Lucasfilm later, but it was a way of sharing information, but they knew that it was their choice as to whether or not they tried to do what the other was doing, because they liked the idea.

Ed Catmull:

That open sharing helped both studios grow faster and better. It was really wonderful to watch. I'm just a believer in that. Is how do you have it so that these groups can both... They want to have a sort of, in one sense, the competition of they want to do better than the other, but they like each other and they're friendly, and so there's no threat between them. There's often when I see something, I'd say, "Oh, that's really cool. I should tell so-and-so at the other studio. Or have you talked with them?" Only to find they've already talked. They weren't going through channels. They just pick up the phone, talk to the right person at the other studio.

Ed Catmull:

This whole system was protected by Bob and put into place. That was a result of the acquisition, but because it was very successful, Bob then felt embolded to buy Marvel. And Marvel was successful, so then he went ahead and he bought Lucasfilm. Now, in this case, there's a clear strategy of getting good talent in each of these areas. In the case of Marvel, people look at... they think it's Marvel and so forth, but he picked up this phenomenal creative leader in Kevin Feige and the team under him.

Ed Catmull:

Most people aren't aware of this team under him, but they're amazing. We worked with them on Big Hero 6 because Big Hero 6 is essentially based upon a little-known Marvel comic. It's the only one where two of their people joined in on our Storytrust sessions down at Disney. They were awesome. He says, "He's got a great team."

Ed Catmull:

That's what they picked up, was this amazing team of people. Okay. So, now you look at other areas. How do they move into the metaverse and how do you actually get the right teams? Either they're grown internally, or from other places, and support them. Other than the positive experience they received from Pixar, Marvel, and Disney, I don't know yet on the metaverse side.

Marc Petit:

I mean, kudos to, back to your Disney feature animation. You turned over the studio without changing the personnel and just by empowering the existing people. We haven't seen this very often in the history of corporate takeovers and mergers and acquisitions. I just want to highlight that point.

Patrick Cozzi:

Ed, I have one last question before Marc goes into the closing questions. I'm going to avoid using the word disruption here. You're a highly sought after investor and advisor, and we're curious if you've come across any innovative or game-changing projects or companies recently?

Ed Catmull:

Yes, but I can't talk about it. There are some things that are secret. (Laughs)

Patrick Cozzi:

Fair enough.

Marc Petit:

We usually close the podcast with two questions. One is to seek inspiration for further episodes. Is there any topic around the open metaverse that we should have covered today, and we have not?

Ed Catmull:

To be honest, I can't quite think of any, although there's one thing I wanted to mention. That is, there's a lot of hype now around blockchain related issues there. Clearly, it's coming in, it's evoked a lot of hype and so forth, as have other new technological changes. But if I look at the long term trends, the one thing that is very clear to me is that everything about the advance of the technology leads clearly towards supporting the metaverse.

Ed Catmull:

I don't think the same thing is true with the blockchains in terms of security, safety, or keeping out the bad guys. And the reason is that, as the compute power increases by two orders of magnitude, the implications for the other things aren't all aligned to support that. Regardless of whether things might think about it, say, "Okay, what are the forces that are aligned?" In the case of computing and security and things like that, it's more destabilizing than anything.

Ed Catmull:

That's not obvious to most people. They're talking about the problems of today. We have to think about what happens when things change by two orders of magnitude. It is moving stably towards a better experience in the metaverse. It's not necessarily true that it implies stability in the other, in every area. It was just one of the things, just trying to put it in terms of a long term context, because people don't usually think of what happens when something changes by two orders of magnitude. And the two orders of magnitude in computing changing is coming.

Marc Petit:

Yeah, that's a very good point. And our last question, is there any person, group of people institution that you want to give a shout out to today?

Ed Catmull:

Well, I think the only thing to say is, as you know, I was around in '92 when John Carmack was coming out with Wolfenstein 3. It was like the beginning of going from 2D to 3D in games. And then NVIDIA was started in '93. The amazing thing to me was that, the big thing that NVIDIA did was to decide to make their iteration cycle in the chips six months. That was not the natural cycle, and they maintained that discipline.

Ed Catmull:

So, what you had was a company who was building chips, going into the games industry, but at the time they were doing this, the people making the games wanted better quality. The SIGGRAPH community had been working for years on the issues of lighting and modeling and dynamics. So, here's an industry that was coming up with the very things that were desired by the gaming world.

Ed Catmull:

NVIDIA would go to SIGGRAPH because we've got the algorithms, which are improving every year, and they then put them into the chips, which they sell to the games industry. But they're not going into boards, which are going into Linux and Unix boxes, so that the researchers in the open community have got better tools to advance their work, all of which they publish at SIGGRAPH. You've got this really interesting, like a powered cycle that is accelerating at a rate which is greater than Moore's Law.

Ed Catmull:

The unexpected event that took place there was it reached such a level of performance, the people realized it could be used for other things, one of which was neural networks and then deep learning. Neural networks and deep learning, which has a major impact on all of computer science, is an unexpected consequence of this cycle between the GPU companies, the game companies and SIGGRAPH.

Ed Catmull:

Nobody was in charge, and of course there was dynamics with each one. Nvidia wasn't the only GPU maker, but they were the ones who started this cycle going, and still sort of like the masters of the cycle, but then these other things happened as a result. This original thing I experienced where computer graphics wasn't a central part of computer science and sort of was questioned within computer science, it turns out that games and GPU and SIGGRAPH were a major contributor to core contributions in the area of computer science. I find that fascinating that it took place with this unexpected outcome.

Ed Catmull:

Now you've got something else which is changing at a rapid rate, and the machine learning will have an impact on the metaverse as well. So, all these things are tied together in this really phenomenal, interconnected network of ideas and thoughts and, you know?

Marc Petit:

Absolutely. Well, thank you very much, Ed, for all of this. In terms of shout out, I would shout out to you in 1972 when you digitized... Was it your hand?

Ed Catmull:

My left hand.

Marc Petit:

It was probably, I mean, you did probably know at the time, but this is kind of, if we had to date the metaverse, I think you're probably one of those moments where it was one of the first original computer graphics movies. So, thank you so much for all your contributions. Been 50 years now, and hopefully we get to see even more progress, and technological and social progress thanks to the metaverse. I want to thank you for being with us today. Thank you so much.

Ed Catmull:

Well, thank you.

Marc Petit:

Well, thank you very much. Thank you, everybody, for listening to the podcast. We welcome your feedback as usual. Topics, ideas, hit us on social. Thank you very much, everybody. Thank you again, Ed.

Ed Catmull:

Thank you.