Why Runway CEO Cris Valenzuela thinks AI filmmaking is the future
56:54 Share

Support for this show comes from Atio. Atio is an AI-native CRM built for the next era of companies. Its powerful data structure adapts to your business model, syncs in all your contacts in minutes, and enriches your business with actionable data. Atio also allows you to create email sequences, real-time reports, and powerful automations, all to help you build what matters, your company.

Join industry leaders like Flatfile, Replicate, Modal, and more. You can go to addio.com slash decoder and you'll get 15% off your first year. That's A-T-T-I-O dot com slash decoder.

Support comes from ServiceNow. We're for people doing the fulfilling work they actually want to do. That's why this ad was written and read by a real person, and not AI. You know what people don't want to do? Boring, busy work. Now with AI agents built into the ServiceNow platform, you can automate millions of repetitive tasks in every corner of your business, IT, HR, and more, so your people can focus on the work that they want to do. That's putting AI agents to work for people.

It's your turn. Visit servicenow.com. Support for the show comes from Alex Partners. The market is evolving at a breakneck speed, and we're only starting to understand how disruptive forces such as AI, cyber threats, and tariffs will change the game. The winners will be those who prioritize execution and know when to adapt. And for unparalleled insights, they can turn to the Alex Partners Disruption Index. Stay tuned to hear more about it later in the show.

In the face of disruption, businesses trust Alex Partners to get things done when it really matters. Read more on the latest trends and C-suite insights at disruption.alexpartners.com. Disruption.alexpartners.com. Hello and welcome to Decoder. I'm Neil Apatow, Editor-in-Chief for The Verge, and Decoder is my show about big ideas and other problems. Today, I'm talking with Runway CEO and co-founder Chris Valenzuela.

This one's special. Chris and I were live at an event in New York City last month hosted by Alex Partners, so you'll hear the audience in the background. Runway is one of the leading AI video generation platforms. The basic concept is familiar by now. You start with a reference image, either something you've created using Runway's own model or something you upload. You type in a prompt, and Runway spits out a fully formed video sequence.

What's most interesting to me about Runway is that while AI hype is at a fever pitch right now, there's a little more depth to the company. Chris founded Runway back in 2018, so he's been through some boom and bust periods in AI, and you'll hear that experience come through as we talk about the technology and what it can and can't do.

Back when Chris began to more seriously explore AI video generation as a researcher at New York University, we still mostly referred to AI as machine learning. And you'll hear Chris recount how primitive the technology was back then compared to now. That said, the AI hype really is out of control, and Runway is on the same collision course with creators, artists, and copyright law as every other part of the AI industry. You'll hear Chris and I really get into all that here.

One theme you'll hear Chris come back to again and again in this conversation is that he doesn't see runway as a disruptive outsider to filmmaking, but rather an active participant in the art. He sees runway as a tool that will bring filmmaking and other forms of artistic expression to many more people, not as an apocalyptic force that will hit Hollywood like a wrecking ball.

You'll hear him say they're working with many of the biggest movie studios today. Runway has already struck a deal with Lionsgate, and just this week, the company made another deal with AMC Networks. In that announcement, Chris said that embracing AI video generation would be, quote, a make-or-break moment for every entertainment company.

But coasting up to Hollywood doesn't mean Runway is off the hook in the AI versus art debate. And in fact, Runway itself is part of an ongoing class action lawsuit over the use of artistic works to train AI. Last year, it was revealed that Runway had trained on huge swaths of copyrighted YouTube material, including the Furchas on YouTube channel.

So I asked Chris as plainly as I could whether, in fact, Runway had, in fact, trained on YouTube, and how the industry might survive a world where all of these companies are made to pay substantial amounts of money to creators if even one of these big AI copyright lawsuits doesn't break their way.

I think you'll find our back and forth in this to be pretty candid, and Chris articulated some of his own defenses for how the AI industry has approached this topic and what might happen next. It's Decoder, so we of course also talked about runway structure. Chris has a lot to say about how runway functions as a research lab, and the tension that exists between releasing and refining real products and putting them in the hands of professionals, while at the same time working on new models and tools that might make the current technology obsolete.

One quick note before we start. Decoder is planning for the future, and we want to hear from you to figure out how we can make our show better. You can visit voxmedia.com slash survey to give us your feedback. We'd really appreciate it. Okay, Runway CEO Chris Valenzuela. Here we go. You started a company before the big AI boom. Yeah. We were joking earlier that the URL is Runway ML. Yeah. Because that's what people were calling it machine learning before. Yeah. What's changed since the boom in that approach? Have you had to, like, rethink...

"Okay, everyone understands what training a model is now, and the market for GPUs is more expensive." What are the changes? A lot has changed. I think we started the company in 2017, 2018. ML was the way we referenced the field of AI broadly. I think a few things have changed. First of all, models have gotten really good. It's obvious for everyone. I hope everyone here has ever used an AI model by now. I'm assuming that has happened.

Seven years ago, no one was. I think consistency, quality, just overall outputs of models across the board are getting really good. And that's just changed people's experiences with AI. The second thing I think has become more real is that the value and how useful models are is becoming more evident to many people. And a couple of years ago, it was more theoretical of how it could potentially be used. There's still many avenues where we don't entirely know how it will change things.

We just know it well. And in some others, it's already changed many things. Like learning, education, it's pretty clear that pretty much I think every student out there from now into the future will start using AI models to learn.

But I think that has jumped in. And then competition, of course. Like now everyone is paying attention to this. When we started, there was really no one trying to build. We've had this same conversation eight years ago when I told you we're going to have AI models that can render video in hyper-realistic ways. People thought we were crazy. And now it's like an obvious kind of no direction. And there's a lot of people also trying to solve that same problem. Is it...

Your ability to actually do the work, was it constrained by the amount of compute you had at the beginning? Is it just scaling laws that have brought you to where... So scale, that's one of the main things I think we've realized as an industry, that scale matters. I guess the lesson that we've seen over time is that if you just scale computing, then it models them to work really well. I think at the beginning it wasn't that obvious, and I think it became more obvious over the last couple of years.

And then more compute, yeah, definitely helps. But more compute and more data and also better algorithms. So it's not just one single ingredient. It's not just if you get more compute, suddenly things get better. I think it's a combination of different things. Put this in a practice for me. When you guys first started,

How long would it take to render a frame of video versus how long now? So when we started, you couldn't. Like, that's a thing. Like, the first thing we ever did was a text-to-image model. There were images of, like, 256 pixels wide. And they were, like, if you've ever seen a Mark Rothko, like, painting, like, very abstract, that's the closest it could get, you know? So if you wanted to render a face or a house or whatever, it was, like, in the range of color, but it was, like, very off.

And so we went from that pixelated, very low-res image to 4K content of 20 seconds long of very sophisticated movement and actions. And so I think it's the realization that at that time, video was not even the scope of what you could think possible.

And then over time, it became really feasible. And now I think we kind of joke where we've consistently moving the goalpost, where the feedback we get from runway is like, great, Chris, you can generate, I don't know, that bouncing ball in Mars. But you know, in Frank 27, like the ball direction is slightly off. And I'm like, great. That's a great piece of feedback because I will solve it. But also like you don't realize that like a year ago, you just didn't thought this was possible. Yeah.

One of the reasons that I see the big platform companies so invested in video generation in particular is they're pointed at the advertising industry. You mentioned you have advertising clients. Mark Zuckerberg, it's not even subtle anymore. He's like, I'm going to kill the advertising industry. He just says it out loud. I think he says it at Stripe Sessions a couple weeks ago or last week. And his pitch is like,

You don't even have to do anything. Just come to us and tell me any customers you want and maybe some ideas what your product is. And I'll generate video advertising and I'll stick it in the feeds and you just watch the money roll in. This is a very Mark Zuckerberg way of thinking. But that is the first big market where you see, okay, we're going to bring the cost of

of making the ads down and that will result in some return. Is that where the demand is coming in for you as well? I think that's a very appealing concept and world for many people who have never had the chance of doing ads in the first place. There's many businesses out there that just can't afford to buy an agency or work with an agency to get a production team to shoot a AAA film or ad.

And I think part of it is like, well, if you can actually help others do that, I think that's great. It definitely raises the bar for many because now anyone can do it. I think it's less about killing the ad agencies. I think that's an overall simplification. I think it's more about you reduce the time it takes to make something. The cost of making any piece of content will dramatically or hopefully go to the cost of inference.

And so if you're good at making things, at conceptualizing ideas, you're going to have systems that can aid you in generating whatever you need. But you still need to have a good idea, right? And so you still have agencies, you still have talent and creatives, but perhaps the time it takes to make things is just going to be dramatically reduced. And that hopefully opens the door for many other folks to do this kind of work. I mean, I think...

Mark wants to call the ad agent. He's a very aggressive human being. But the reason I ask that question that way is I see so many of these products and so many of these capabilities, and they haven't yet connected to business results. For a lot of companies, there was a study from IBM last month maybe, it said 25% of the AI investments they had seen in companies had returned on that investment. It's a low number. Everyone's trying stuff and figuring it out.

I get it in advertising. I understand, right? That's just a cost of acquisition of customer. Have you seen places in film studios or other places where just bringing the cost down is worth the investment? Yeah, absolutely. I was just with a call with a studio right before this and we were going through like a script that they wanted to like test Runway for.

And the typical, I don't know if you guys have ever worked in film, but like you develop the script and once you get the script, the common thing to do next is like a storyboard. And so you basically take the storyboard and someone spends like a week or two weeks just drawing like, this is like for a scene, like or a couple of scenes, not for an entire film, but

It's really expensive and like time-consuming So when they were reading me through the script where they need our help with our runway I was on the side generating like the storyboards on the fly and by the time they finish the storyboard was done and so they I think the first thing was like they couldn't realize or fully understand what was going on because they've they never have work at that velocity at speed and For them speed is also cost if you have to complete

compound the time it takes to make all of those servers by hand, and they have the screenwriters doing it on real time, then it shrinks the time the whole project gets developed and worked on. So you have all these moments and gaps where AI can really just help you accelerate your own work, specifically in creative industries where still things are very manually done. I actually want to ask you about that because I know you think a lot about the creative industries and the act of creativity.

The counter argument to that is the gap between the screenwriter and a storyboard artist and the time it takes to communicate and translate is where the magic happens. Having the AI collapse that into a mechanical process as opposed to a creative process,

actually reduces the quality of the creative. How do you feel about that? Yeah, I don't think I fully agree with that. I think part of it is, I think sometimes we're over-obsessed around the process of how we make things. Like the goal of the screenwriter is to get the ideas that he wants in his mind or his world out there.

and the most obvious ways you work with this set of technologies and tools around you, if you're able to do it faster, I think that's great. You can iterate on concepts faster, you can understand your ideas faster, you can collaborate with more people, you can make more. One of the big bottlenecks of media these days is that you have people working on one project for four years, three years.

then you might actually work on it and the student might actually decide to kill it for many different reasons. So if you think about it, you spend your entire life, or like four years of your life, working on a thing

that never saw the light of the day because it happened to be killed by a weather reason. And I think the idea would be, well, you don't have to work on one project. Hopefully you can work on many more. So that's also the quantity prospect of it becomes, I think, a component that we should consider because right now we're bounded by the way we're working and it's very slow and it's very constrained by all these processes. If you can augment that, then people can start doing more and more and more. I think that's great. Is that kind of the model for you? Is it that quantity will drive the business?

I think quantity leads to quality. The more you make, like as an artist, the more you make, the better things you will do. Like every artist has probably never drawn once and be like, oh, sorry, I'm a master. Like Picasso painted hundreds of thousands of paintings. And many of you have never seen all of that. You just see like the 1%.

The same for musicians, people are there playing every single day until you hit something that actually works. And I think tools should be like that. They should be able to augment how you work so you can do more. And then you're the one choosing what you're doing. But look, I started the company because I always wanted to make films. And I grew up in Chile and I've never had the means of even buying a camera in the first place. I got my camera when I was 27 years old. It was pretty late.

And part of it was very expensive. I couldn't afford Adobe software, which was very expensive back then. I think I probably wouldn't have become a great filmmaker, but would have been great if I had the chance of telling the stories that I had in my head. And I think it was a technical barrier that prevented me from doing so. And now we have kids in every part of the world using Runway and making those ideas, which I find just fascinating. It's great.

How does the pricing of Runway work? Where does your revenue come from? What's the model? It's very simple. It's a subscription. You just pay for the product and you get access to different parts of it. They have a free tier, so you can also just use it for free. And then we work with schools. So there's a course at NYU, the NYU Film School, that teaches students how to use Runway. So instead of going to film school and giving you a camera, they give you Runway.

And so we're doing that with a few other schools as well. And so for all of those, we just give access for free. Are you going to-- the studios you're partnered with, do they pay a lot of money? Are they subsidizing it for users? No, I mean, for businesses, yeah, we charge. I mean, students can pay. But also, they pay because it's useful. If it helps you do something, then sure, the value, it's worth it. Are you profitable yet?

No, we're growing and I think a part of what we're doing is just investing in research more than anything else. What's your runway to? I think again, we've been working, I've been obsessively working on this and I would say over the last 12, 18 months, it's just like the models have gotten to a place where you can actually do very good things with it. I think there's always an optimization function that companies have to run, which is do you want to optimize for whatever is working now or do you want to keep on growing?

I think for us, it's like we really want to keep on growing. There's a lot of research we can invest on and a lot of areas of growth that we can keep on going. So I think the tension right now has always been like, do we want to optimize for this or what's next? I think we want to lean into what's next. I think there's a lot of things we haven't actually fully discovered that we could do that we want to do. We need to take a quick break. We'll be right back. Support for the show comes from Alex Partners.

Disruption is the new economic driver. The days of predictable business cycles are over. For over 40 years, Alex Partners has helped companies develop winning business strategies amidst uncertainty. One of today's greatest challenges? The rise of AI. As AI reshapes the tech landscape, Alex Partners is committed to helping your company thrive.

In their sixth annual Alex Partners Disruption Index, a global survey of 3,200 senior executives, 65% of executives believe AI and machine learning provide positive opportunities for their companies. And 62% of CEOs expect significant business model changes in the next year. In the face of disruption, businesses trust Alex Partners to get straight to the point and deliver results when it really matters.

Read more on the latest trends and C-suite insights at disruption.alexpartners.com. Disruption.A-L-I-X partners.com.

Support for this show comes from Pure Leaf Iced Tea.

You know that point in the afternoon when you just hit a wall? You don't have time for self-care rituals or getting some fresh air, so maybe you grab a beverage to bring you back. But somehow it doesn't do the trick, or it leaves you feeling even worse. What you need is a quality break, a tea break. And you can do that with pure leaf iced tea, real brewed tea made in a variety of bold and refreshing flavors with just the right amount of naturally occurring caffeine.

With a Pure Leaf iced tea in hand, you'll be left feeling refreshed and revitalized with a new motivation to take on what's next. The next time you need to hit the reset button, grab a Pure Leaf iced tea. Time for a tea break, time for a Pure Leaf.

We're back with a live conversation with Runway CEO Chris Valenzuela. For the break, we were talking about the growth of Runway and its business model, especially considering how nascent AI video still is today. But now I wanted to ask Chris the decoder questions, especially about when to invest in research versus when to actually pull the trigger and release a product.

One question I ask everybody on Decoder: How is Runway structured? How do you organize the company? It's very lean. Someone thought the other day that we were a thousand people, and I thought that was the best compliment they could give me. We're like a hundred people or so. It's very flat, very focused on autonomy more than anything else.

And then very focused on what we do is less of like objectives. And we don't don't believe in objectives. We have a way of working that's where we just set boundaries and where we want people to like do research or explore. Because a lot of what we do has never been done before. And so if I tell you how to get there, I'm probably wrong because like we've never done it. So it's research. You have to experiment and fail. And so what we do is we set the constraints or the boundaries and where we want you to experiment.

And that's the best outcomes of research we've done, have been about setting the right kind of boundaries and then people, let people go, you know? Let people like work on its own and figure out on its own how to do it. So you're like full holacracy, no org chart?

I mean, there's some work charge in some way, but not people like collaborate. We have a studio, an internal studio with creatives and producers and filmmakers working along research. And those people are sitting on the same table, speaking the same language, and they come from different backgrounds, but they've managed to just collaborate together. And so, yeah, that's one you want to provoke. One of the reasons I'm interested in asking that question, particularly of AI companies at your size,

is there is the deep connection to the capabilities of the model and the research that's being done and the kinds of products you can build. I haven't seen a lot of great focused AI products. Runway actually might be one of them. But in the broad case, there's ChatGPT, which is just an open-ended interface to a frontier model. And then we're going to see what happens.

Do you think that as you get bigger, the products will get more focused? Or do you think you still need the connection between the team building the model and the product teams themselves? I think the connection of product and model helps the product team understand better what's coming. So you need to understand that the way tech has used to work was in much lower cycles of R&D. Now research tends to move in very fast cycles. And so the issue with product, and I think product is one of the hardest things to do right now because

you scope the kind of like area of product you want to work on, you design it and you start building it and by the time you build it, it's kind of obsolete.

You've basically lost six months of work or whatever long it takes you. And so product needs to behave like a research organization. The way we tell our team is, look, we have research scientists working on research, but everyone in the company is a scientist because everyone is running experiments. So before you spend too much time doing something, run an experiment. Build a simple prototype, understand if it's worth it, and then check with research if the thing that you're working with, it's going to become useful or it's going to be there or it's going to actually not get submerged by the next generation of models.

What happens a lot is that our customers are coming to us with specific questions like, "Hey, the model does this, but it doesn't do this. Can you build a specific product for that?" And we could. We could build a product just for that, or we could wait for the next generation of models that would just do all of that on the fly. So that's the tricky part because you're always trying to catch up, play catch up. And I think companies that understand research are in a much better position than just companies that are trying to catch up.

There's a comparison I keep making here that you're not going to like, but I'm going to make it anyway. I started covering tech a million years ago, gray hair and the beard. And when Bluetooth came out, everybody knew what the product was going to be. Everybody saw that headsets...

Every real estate agent in America had like a giant Motorola headset. And it's like, oh, you want AirPods. And then the standard was just not ready for another decade. And then Apple had to actually build a proprietary layer on top of the standard to make AirPods. And that took a full decade. It was just not ready. And there was a real dance there between what do we want to build? What's the product? Can we build it? And does the technology support our goals? Yeah.

You're kind of describing the same dynamic. And the thing that gets me about what you're describing is, well, the model's just going to eat the product over and over and over again. How do you even know what products to build? Yeah, it's very hard. Because everyone can see the AirPods, right? Everyone's like, the computer is going to talk to me. It's going to be fine. Yeah, but I think that's more than just the computer will talk to me. I think there are parts of how it will talk to you and when it's the emotion. There's a lot of...

product that goes into research back, you know? And the thing is, like, no one really knows, to be honest, what the future product experience would look like. Because a lot of the interactions we're having, we just never thought we could have. And so you're only going to realize by having people use it. I think that happens a lot in research where, like, researchers spend, like, so much time red teaming and doing all the work, and then you put it out, and in, like, two minutes, just someone, like,

figures out how to use it in a completely different way. And actually, I think that's great. It points to the fact that I think the previous generation of software was based on this idea of you choose a vertical and just went there.

Then the next generation of software is based on like, you choose a principle of how you want to operate in the world and you build models towards that. Our principle is every pixel or more pixels that you will watch and consume will be generated, will be simulated. That's the principle we're operating. And therefore you can go into many different products based on that idea. And so it's kind of the difference between choosing a vertical and choosing a principle in which you want to operate.

But right now, as you're deciding what products to build, you're getting market feedback from users. You have studios using the tool, agencies using the tool. You've got to make some decisions, right? We do. Where are we going to fill the gaps with the product and where are we going to wait? How do you make those decisions? We focus a lot on research and focusing on understanding what's coming and what's worth building. I think there's always a trade-off, specifically with startups, where if you spend too much time working on the wrong thing, it may actually kill you.

I think we listen to users, but sometimes users don't really know what they want. They know the problems really well, but they can't articulate the exact solution for it. So you don't find yourself building exactly for what they're describing because they can't describe the thing that they don't know it's coming. And so it's, I don't know, I think it's kind of like art, I guess. You become just really good at intuition and being like, okay, that thing, even if it could be a great deal now, we're not going to do it right now. And so I think companies overall build...

intuition and that's just experience of doing it enough times and then say no. You have to say no a lot of times. Customers come with great ideas, but just say no. Not because you don't think you can solve for them, but again, because it will trap you into the wrong thing for the wrong. This is the other question I ask everybody broadly. How do you make decisions? What's your framework? How to make decisions? What kind of decisions? All of them.

I think there are different decisions. There are decisions that are like much more long-term and irreversible and decisions that are much more reversible. I think we're very much of the idea that, again, run experiments and like be willing to understand if you're wrong in your assumptions. If you need to make a decision, like do it because it's like you're confident it will work. And if it doesn't, you can like

change your mind, you know? Sometimes product decisions are coming from that taste component to it. I think overall taste has become a good way of like directing the company, I would say, for how we operate in like marketing and how we hire. I don't think there's one particular framework, but just overall the idea of like taste and intuition has become clear in how we make decisions. Do you think you're going to have to change that as you hit the next set of scale? Right? At 100 people, you can be like, just listen to me. Yeah. At 1,000 people, maybe not. Yeah.

There's a thing we keep referring around this idea of a company-company. We don't want to be a company-company. A company-company is a company that behaves like a company because that's the way companies behave. And you're like, no, don't do that. Be a company that's focused on solving a problem or a research constraint or a user need. Don't focus on the things that are superficial that you're supposed to be doing because you're a company.

And I get paranoid about that because the moment you lose that and you get into that, you're dead. You're going to stop innovating. You're going to focus on the wrong things to optimize for. And I think it's just culture maybe reinforcing this to the team. I still interview everyone in the company. I'm still pretty much involved in how we make decisions on product. The moment you step back is probably the moment like things are just... Organizations tend to...

seek like a slow velocity if you're not constantly like pushing them all the time. Do you think there's going to come a point where the split between

the capabilities of the underlying model slow down and that you have to put more into product? Maybe, but I don't think we're close to that. Even if we stop now research, we decide collectively industry-wide, let's stop research. I think there's like 10 years, 20 years of innovations that are just there, you know, latent, waiting for someone to discover them. I don't think we're at that point yet where you can say, hey, this is enough.

Because I think there's just too much space to grow and have models to think that. Like, we just released a model two weeks ago. And I'm not kidding. Every day, I open our users and Twitter and Instagram, and there's a new use case. Like, now, just before coming here, someone was using it to, like, basically for, like, clothes. So you can try on anything. Like, you basically go to any shop online, any e-commerce site, put a photo of yourself, and just, like...

see yourself wearing that in hyper-relief manners, like I just never thought you could do it for that. And you can. So, yeah. I was talking to Kevin Scott from Microsoft, CTO of Microsoft, and he's made the same point in a slightly different way. He said there's more capabilities in the models we have today than anyone knows what to do with. I agree. And to me, it's like,

well, then we should start building products that like make sense. But then the tension is, well, the next generation of models are just going to eat my product. Yeah. When does that get stable so that anybody can like,

Make some products that are good. So here's the great example. That's a great distinction between verticals and principles. If you think about a vertical, then you'll choose a solution and you build towards that. If you think about a principle, you should assume that many of the things that you're trying to build in the product will eventually become futures of new models. So therefore, your product should be many layers ahead if you want to spend time on it. And so the principle should be, for example, an image generation.

Zero shot. So zero shot means you don't have to do, if you want a model to do something, you don't have to train it. You need to just show it examples. Therefore, you can widely expand the range of things models can do if you have the right examples. So maybe a good idea would be find, collect examples of things you can teach models for. And then it changes the way you can approach product. So I think that distinction between principles and verticals is relevant for that.

One of the big trends in the industry is the cost of every new model is getting exponentially higher. Sam Altman is touring the capitals of the world being like, can I have a trillion dollars? And maybe he'll get it. You never know. He might get it. Are you on the same cost curve where every new model is that much more expensive? Do you have a trillion dollars? Is the answer yes? If you have one...

AI tends to move in two waves. There's the expansion wave and the optimization wave. Expansion is like, well, we're discovering what we could do. And if you think about the models from two or three years ago, yeah, they were expensive. Now, most of those models, you can train them on your laptop because models have gotten to a stage where you can optimize them. And there's one thing engineers love.

It's optimizing things. And so if you tell them, like, here's a thing that works, optimize it, people will go very hard on it. And like for some models that are two or three years old, now that's the case. They're very cheap to train from scratch.

I think there are new models that are still in the expansion phase. We haven't figured out exactly how to optimize them. But we will. But I think what happens is the same thing. If you spend too much time optimizing them, the trade-off is you're going to stop working on the new expansion. And I think most companies these days are betting on expanding. So they're betting on paying more for the sake of expanding that and not falling behind, rather than trying to optimize and reduce the cost of the thing that works. Where are you?

I think we're, again, on the expansion side. Having the ability to expand that, having the ability to innovate on that, it's way harder. And then having the ability to just catch up and play optimization

game is easier. And I think our bet is like, well, this is the advantage point where you can keep on moving things and just keep pushing boundaries. The big platform companies, Microsoft, Google, Amazon, OpenAI has a deal with Microsoft, right? They run their own hyperscalers. Is that a competitive threat to you? Is that an advantage to you?

Well, Goo is an investor. So we work closely with them. Again, there are different functions of businesses. If you're a hyperscaler, you're probably in the business of optimizing things. You need to make things cheap and scalable for everyone. It's a different function from a research lab who is building these new things. And so, again...

Probably it's good to pair the two because you have a good research lab with a good optimization. Then there's transfer that you can make technology-wise that will allow companies to just run on the things, sell it, and then get feedback while the company, this other part of the company is working on the next thing, which is where we are. If Google's an investor, are you running on GCP? That's correct. So do you just let them buy the NVIDIA H100? Do you worry about that at all?

Yeah, Vue and Vitas are the same investor. The AI industry is full of this, by the way. It's very obvious. Well, I think it's people who have seen this and I think you want to provoke this. And you want to provoke... Many of the things we're discussing now weren't that obvious eight years ago until many people started to make the right bets on it.

And I think, again, depending on where you are, it might be a good function to partner with people who get it and who want to work with you long term. And I think the people we work with can help us get to that point. I think NVIDIA as an investor is one of those things about the AI industry which is very funny. They're investing in the applications that drive the usage of their chips in all these places. And maybe some of them will pay off and maybe they won't. That's the nature of investing. But at some point, everything has to add up.

to actually deliver a return for Nvidia. Do you feel that pressure that this big runway has to be a big enough business to justify all of the infrastructure expense? I think the justification comes from the value you see customers and the adoption that you see. I think that's the way, that's how you see AI in products go from like zero to many millions of revenue in like a couple of weeks, months. Something that was like unseen before. And it's because it just, it's such a different experience, it's such a different value.

that if you're ambitious about it, I think, yeah, it will definitely get there. And we're already seeing this.

Still, video, for example, is very early. Gen 4, our latest model, is literally a month and a half old. And so most of the world has not even experienced it yet. So it's also a distribution problem. How do you get to everyone out there who can use it? Are you at millions in revenue? We're, yeah, more than that. Do you have a path to billions in revenue? We hope, yeah, over the next couple of years. I'm asking, all these companies have to generate billions in revenue for all these investments. I think they will. Many will. I mean, again, if...

Think about it from first principles. If you're in the business of ads or movie making, if you're in the business of movie making, you're spending hundreds of millions of dollars to make one movie.

If I can take that process and help you do it for a couple million, then all the delta, I can literally charge you for whatever delta I'm helping you improve there. And hopefully, I can charge you way less so you can actually do more. And if you expand that, then you're also not helping them, but you're expanding the window of who can do that thing in the first place. Because if you think about professional filmmaking, it's a very small industry, mostly because it's very expensive.

Well, if I have something that makes it cheaper, then I can expand their definition of who can get in the industry in the first place. And from a market perspective, that's great because you have many more people who can do something that they never thought they could. We need to take another quick break. We'll be right back. This message comes from Rinse.

These days, you can do a lot from your phone. Book a vacation, trade stocks, but you can also make your dirty laundry disappear and then reappear washed and folded with rinse.

Schedule a pickup with the Rinse app and before you know it, your clothes are back, folded, and ready to wear. They even do dry cleaning. Sign up now and get $20 off your first order at Rinse.com. That's R-I-N-S-E dot com.

Support for this show comes from Pure Leaf Iced Tea. You know that point in the afternoon when you just hit a wall? You don't have time for self-care rituals or getting some fresh air, so maybe you grab a beverage to bring you back. But somehow it doesn't do the trick, or it leaves you feeling even worse. What you need is a quality break, a

A tea break. And you can do that with pure leaf iced tea. Real brewed tea made in a variety of bold and refreshing flavors with just the right amount of naturally occurring caffeine. With a pure leaf iced tea in hand, you'll be left feeling refreshed and revitalized with a new motivation to take on what's next. The next time you need to hit the reset button, grab a pure leaf iced tea. Time for a tea break. Time for a pure leaf.

This episode is brought to you by Amazon Prime. From streaming to shopping, Prime helps you get more out of your passions. So whether you're a fan of true crime or prefer a nail-biting novel from time to time, with services like Prime Video, Amazon Music, and fast, free delivery, Prime makes it easy to get more out of whatever you're into or getting into. Visit Amazon.com slash Prime to learn more.

We're back with a live conversation with Runway CEO Chris Fonsuela, discussing how Hollywood is reacting to the pressures of both AI video and TikTok and YouTube. I also wanted to know what he thought might happen to both the AI and entertainment industries when some of these landmark AI copyright lawsuits work their way through the courts.

The film industry is really interesting. It's under a lot of pressure. So much pressure that HBO Max just keeps renaming itself every six months to get whatever attention it can. It's great. It works, I guess. But fundamentally, they're competing with TikTokers and YouTubers, right? Netflix knows this. Netflix knows that YouTube is the biggest competition. The cost to make a YouTube video is already...

only percent of a cost to make a Marvel movie. And that has basically put the movie industry under a ton of pressure. Do you think AI can actually shrink that gap? Yeah, I think... And keep the quality high? Yeah, so I think that's the point. I think the last frontier of content, the last frontier was low-quality content anyone can make. I think that's TikTok and YouTube. Like, there's billions of people out there making everything. The difference between that and, like, a high-production studio is the quality of the content.

the output, how good the output of the pixels and the videos is. And that for me is mostly a technical barrier. It's not a storytelling one, it's not an ideas one. It's like, yeah, making like a high-end science fiction movie, it's really expensive because you have to hire many people and work with this software that's very expensive. So the last frontier I would say for us, and I think many media companies are thinking about it, is like billions of people making high-end content.

And that is the one that I think if you're in the traditional business of media and you haven't realized that, you're probably very scared. Because then you will compete with anyone, anyone in any part of the world with a small budget and very good ideas can make amazing things. And we're seeing this already. The last Academy Award for animation this year, I don't know if you've seen it, it's a movie called Flow. Very small budget, I think less than like $10 million.

And it's just a very good group of people working with great software. And they won the Academy Award against $100 or $200 million productions. And it's just because you have very smart, talented people working with right software tools. So the flip side of this is those studios are also jealously protective of their IP. That's the thing that they monetize, right? They window it into different distribution channels, into different regions. They sue pirates who steal it on BitTorrent.

You've trained on a lot of this content, right? There's reporting that Runway trained on a bunch of YouTube channels, including The Verge's, by the way. There's your trillion dollars. This is, in my mind, the single greatest threat to the already exorbitant cost structure of the AI industry is there's lawsuits everywhere that might say you have to pay all those creators. Have you thought about that risk?

It's part of how we analyze and how we work. We've worked with different studios and companies on understanding how to train the models for the needs that they have and what they want to do. It's still crucial for me to help everyone understand what these models are actually doing. Because a lot of the assumptions that we get around AI video is that you type in a prompt and you get a movie. And so if you're a studio, and this is actually true, I get, now less often, but I used to get a lot of scripts in my inbox, in my email.

where people were like, "Hey, I'm a producer or a writer. I've been working on this show. I have the whole script done and it's great. I heard you do AI videos, so here's the script. Make my movie." And I've realized a lot of people thought that what AI video meant or AI pixel generation or making videos with AI meant was that you type in a prompt and you get entirely the entire movie that you thought you were in again.

And the reality is that, no, it doesn't work like that. Like, it will probably never work like that. You're still pretty much involved. You need to tell the model how to use it. You need to tell the model the directions and the inputs you want to use. Like, I think part of it is that perhaps most of people's experiences with AI over the last 12 months have been through chatbots.

So the idea of AI has been condensed to this idea of chatbots. So if you have a chatbot, you have AI. And those things for me are summarizing a huge field into a very simple, oversimplified concept.

And so when you think about like copywriting, you think about creating things. I think for me, all the weight still is in like, what are you making? And you're so much in control. And these are not tools that will make things on your own. You are the one deciding how to make them in a way. And so you have to be responsible on how you use them. That's basically the point. But,

But to train the model, you need to ingest a huge amount of data, right? The two things that make the models more effective in an expansion mode are more compute and more data. Have you thought about whether or not you're going to have to pay for the data you ingest to train the model? Yeah, so again, we've done partnerships to get data that we need in particular ways. But again, it's really important to understand that these models are not trying to replicate the data, right? I think the common misconception is that, and I get this, where people thought that you can type a scene of a movie and you get the scene of a movie in runway.

These are not databases, so they're not storing the data, they're learning. They're students. They're learning about data, they're getting patterns within that data, and they use that to create something net new. And so the argument I think that it's really important to consider is that these systems are creating net new things, specifically for videos, they're creating net new everything, pixels. And the way you use them should be a responsible way, of course.

And the models are not trying to store anything. So that for me is like distinction because it changes the argument of how you think about training models in the first place. If you think about them as databases, you're going to have a set of different assumptions and use cases and concerns if you think about them as general purpose tools, like a camera. I always speak about running as a camera, you know? Like a camera allows you to do anything you want. It's up to you how you want to use it.

You can get in trouble for using a camera, or you can make great film by using a camera. So you choose. It's shockingly easy to get in trouble for using a camera. Yeah, I know. But I think, look, I grew up in Chile, and there's a lot of films I didn't manage to see. And the way I saw it was I bought them both at street corners.

And I don't know if you ever saw one of those where people would stand in the theater and just record the thing. Yeah, I mean, that was a bad use of cameras. But I think the overall assumption of society was like, let's not ban cameras. Let's actually have a norm in theaters where you can't do that. And if you do, you're going to get in trouble. I think we all agree it's like, yeah, that's a good thing to do. That argument is weaving its way through the legal system right now. There's lots and lots of court cases.

The last time we went through this, it was basically Google that won a bunch of court cases about building databases, maybe different. But Google is like a friendly, young company that slides in the office and wore beanies when they got to work. And the inherent utility of Google search, like very obvious to every judge. The inherent utility of YouTube, which got in a lot of trouble, was very obvious to every judge. And they sort of like horsepowered their way through it. Like they had to pay some money to some people and they had to win some cases. They had to invest a lot into litigation.

And they won because they were cute and they were Google. And it was a very different time. Tech companies broadly are not thought of as young and cute anymore, right? Like no one thinks of Meta and Amazon as Google as like adorable companies that should build the future in the way that they were at the time. Have you thought about the risk that...

They might lose these cases and what that would do to your business. Because this dynamic you're talking about, whether this is a non-infringing use, whether there's broad utility here. Yeah, this argument goes back to the Betamax case in the 80s. Like, it's all there. But it doesn't have to go the way that it always did, right? Like, judges are just a bunch of guys, as we've discovered here in America. They just make decisions. What if it doesn't go your way?

It's hard for me to have an opinion on every single case out there. I think it's more complex than that. I think Google has been a great impact in the world at large. I think it's hard to disagree on that. I think the world has gotten way more expansive. Information has become more accessible to many. I think that's hard to disagree, right? I think there's definitely new challenges with every new technology. I don't disagree with that. I think it is.

You are putting really powerful technology in the hands of everyone, which means everyone, right? So there's use cases around AI that you should be preventing and you should try to make sure you have systems of regulation and safety on top. I think every company is different. So I think it's hard to... One thing I've learned about tech, and I mentioned this as an artist. I went to art school and I started working on tech mostly as a way to...

have my vision around how art should work with tech. That was my idea. And so I still consider myself an outsider to tech. And I think one thing I would consider is that not everyone operates in the same way. I think not all companies are the same. Companies tend to be different in how they operate. And I think there are different ways of managing through this change. And it's hard for me to group everyone in the same and say, yeah, all tech companies are basically doing the same.

Let me write this the right way. You train on YouTube, right? You train on YouTube channels? We train on a variety of different data sets. And so we have teams working on image and video and text and audio. We don't disclose how we train our model because that's unique to our research.

Did you train on YouTube? We have a variety of different data sets that we use to train our models, depending on the task. It's not about do we train on this and that. We have agreements with different companies. We have partnerships with others. The way we train is very unique to us. It's very competitive there, so we're never probably going to tell how we do it because it's very unique to how we train our models. Yeah.

YouTubers own the copyrights to their videos. If it comes out that you train on YouTube and hundreds of YouTubers come ask you for money, at whatever rates, is the financial model of Runway still tenable? I guess it goes back to what are these models doing, right? Well, I'm saying if OpenAI loses its case against The New York Times and training on The Times content is found to be infringing, the floodgates will open. It is not clear if OpenAI will win or lose. If Meta loses its cases,

against the book publishers. They're not doing great in the past couple of weeks. The floodgates are open. If those floodgates open, is your business tenable? No, I think, again, summarizing the entire AI industry as like chatbots and what one company is doing, I think it's a mistake. I think, again, video and media works very differently and there's a lot of other considerations. I think a lot of the assumptions around how AI works

that I've seen around video, it's like having an opinion around cell phones in 1992. You're just probably very early on seeing the impact about how that technology will change industry and probably you've never experienced it before.

And so I think part of what I think is going to happen over time is that a lot of these ideas around concerns around copyright and other considerations will start to change as people understand how this actually works. Like, I'll give an example. I was in a dinner with a producer of a major show you probably have all seen. And he was like, I'm very anti-I. Like, OK, why are you anti-I? It's like, well, because it works like this and it does this. And I was like, no, it doesn't. Like, let me show you how it works.

And then we showed you how it works. And he was like, yeah, now I'm on board. It took me like 25 minutes. And I think he was very adamant of his position of being very against AI because I realized he just had the wrong expectations about what it did. And I think it was a minute of like, okay, let me show you what it does. It's like, you've never experienced this before. We forget this, but we all had to go through training to send our first email. Like people were just, they knew how to send emails and you have to go through it. You just don't understand it. And so you

You start using it, you understand the limitations of it, understand the constraints of it, and then you start using it. I think a lot of the hard takes on AI these days, I think, are based on just the wrong expectations and the wrong assumptions of what it actually does. That gap between how artists feel about AI and then how much they actually use it seems like it's getting bigger every day.

Shows up on our site at The Verge. By the way, The Verge is built on the very foundation that I was right about my opinions about cell phones in 1992. But we see it, right? People read the articles. I talked to

product people at other companies, Adobe, for example, the usage rate of generative AI in Adobe products is like 100%. Generative fill is used as often as layers, which means everyone uses it every day. And then the audience is like, I hate this, make it go away. And there's just this gap. It's a moral gap. It's a psychological gap, whatever it is. There's a gap between how people are using it and how they talk about it and how they feel about it, particularly with creatives, particularly with artists.

I know you spend a lot of time with creators. How are you closing that gap? Is it possible to close that gap? You know, I don't see that gap that often. In film, there's the idea of below the line and above the line. If you speak with a VFX artist, someone who's actually moving the pixels in a screen, they don't have weekends. They've never had a weekend off. Because when you're in a project, it's a very tough timeline with very small budgets. And the director comes with notes, and you have to do the notes, and it's a Friday, and...

there goes your weekend. You're going to be working on pushing those edits every day and you're doing it by hand. And so if you have a tool that allows you to do it faster, of course, you will use it. It's great. It will get you where you need to go faster. I think the gap there is not as big as some people might think because the actual creative minds, the producers, the editors, the VFX artists are already embracing this and it's very valuable. And I guess I'm not surprised about your stats and number.

I think still the above the line, the people who think about creatives as like, oh, like have never had the experience actually working and sitting in might have a different assumption of how it works. Again, I think part of it is just like, we need to show you how it actually works.

And something we do is, we have a film festival here in New York, by the way, if anyone here wants to go. We've done it for three years now. It's in the Lincoln Center. It's a major event. It gathers filmmakers from all over the world. We started the festival with 300 submissions. This year we got 6,000 submissions. We worked with the American Cinema Editors, which is one of the guilds of the editors. We worked with Tribeca Film Festival, so we have industry partners.

And it's a great way of understanding how it's actually being used in real production use cases and how valuable it is for not only the insiders, but also the new voices. I think part of the gap is you need to go to a film festival to experience it, and you probably get a sense of how useful it is. The concern from that class of people that we hear all the time is, this is great, it made everyone's life a little bit easier, it also put half of us out of work.

Do you see that as a real threat or as a real outcome? I don't. I understand the concern, but I think the obsession should be on people more than jobs. Like we used to have people that pressed buttons in elevators. Like that was a job. I don't know if you guys remember this. That was a job. There was a job of people throwing stones to wake you up before alarm clocks were invented. I think no one is saying we should protect people

people throwing rocks because of the job, we should have alarm clocks and the person who's throwing rocks to wake you up should be taught how to do something else. And so you focus on the people and how you upskill and upgrade and learn and teach people to new things rather than like, hey, let's keep this thing because we need people pressing buttons in elevator and that's a job. I think that has happened in Hollywood many times. When Hollywood was at the beginning, it was silent. There was silent movies.

talkies came around, it was a major breakthrough where you can actually have sound in movies. The industry revolted. Charles Chaplin, one of the biggest advocates against films with sound because he said that sound would just kill the essence of filmmaking. And an argument that they had was like, who's going to pay the orchestras that are paying in the theaters, right? Well, it's true. Yeah, we don't need orchestras in theaters anymore.

But also the technology gave birth to an entire new industry of artists like Hans Zimmer. I mean, that was the beginning of an entire new industry given by technology. I think this is for me very similar. We're like, yes, we're going to lose some jobs. Our job should be to train those people to do new things with technology. Last question. If you had to spin that all the way out.

You're successful, the AI industry can pull this off, the models get the capabilities you want them to look. What does the film industry look like 10 years from now? It looks very much... Is it just TikToks? That's like what I like, are we just going to do Quibi? No.

I mean, if someone likes making that, I don't think there's nothing wrong with it, right? I think there are many independent voices out there that have never had the chance of telling their stories because they don't have the means to tell it. Our vision of Runway is that the best stories have yet to be told. We haven't heard from the greatest stories of the world because maybe they just weren't born in LA. That probably is the case, right?

And so I think we're going to see a much more democratized version of film. We're going to have a version of storytelling that's for everyone. And the bar for it will be the ideas. It won't be who you know in the industry or how much money you have. It'll be like, how good is the thing you want to say and how good are you telling it? Well, Chris, this is amazing. You're going to have to come back in the coder scene. Of course, yeah. Thank you for having me.

I'd like to thank Chris for taking time to speak with me and thank you for listening. I hope you enjoyed it. I'd also like to thank Alex Partners for hosting this conversation. If you'd like to let us know what you thought about this episode or really anything else, drop us a line. You can email us at decoderattheverge.com. We really do read all the emails. You can also meet me directly on threads or blue sky and we have a TikTok and an Instagram. Check them out. They're at decoder pop. They're a lot of fun.

If you like Decoder, please share it with your friends and subscribe wherever you get podcasts. Decoder is a production of The Verge and part of the Vox Media Podcast Network. Our producers are Kate Cox and Nick Stapp. Our editor is Hursa Wright. The Decoder music is by Breakmaster Cylinder. We'll see you next time. Support for the show comes from Alex Partners. Now in its sixth year, the Alex Partners Disruption Index explores what the best performing and fastest growing companies are doing differently as they anticipate, shape, and respond to disruption.

How are the world's top leaders identifying ways to integrate AI into their businesses? Learn more on the latest tech insights at disruption.alexpartners.com. In the face of disruption, businesses trust Alex Partners to get straight to the point and deliver results when it really matters.