Loading summary
Transcript49 lines
- [00:00]
A
Presented by tech domains where tech founders find sharp, memorable names for their tech startups. Hello and welcome Back to Equity TechCrunch's flagship podcast about the business of startups. I'm Rebecca Balon and this is the episode where we bring on industry experts to help us explore a trend in the tech world and dive deep. AI generated video has gone from novelty to creative tool in what feels like the blink of an eye. And nobody understands this better than Runway. They're a startup that's based in New York and they're punching way above their weight in terms of their model releases coming out with models that are competing with some of the most well funded labs like Google and OpenAI. As Runway scaled its video generation models, it learned that its applications are extending way beyond creativity. They are advancing into general world models with applications in use cases like gaming, robotics, and perhaps a more generalized intelligence. Here to talk about this today with us is Runway's co CEO and co founder, Chris Valenzuela. Chris, welcome to the show.
- [01:04]
B
Thank you for having me.
- [01:05]
A
Yeah, really excited to chat to you again. We've been chatting all week. I've been chatting with your team about what you guys are doing at Runway, but I wanted to kick things off by bringing up something you said recently, that AI could help Hollywood make 50 films instead of $100 million blockbuster. Now, I think with, you know, catchy headlines like that, people might misconstrue or they get attached to something like, oh, these AI guys are just trying to disrupt and displace, you know, Hollywood and filmmaking. But want to know what you're actually proposing. Right? Like what is the more nuanced argument than just treating filmmaking like a content factory?
- [01:43]
B
Yeah, that's a good question. And I agree sometimes headlines do tend to be a bit more click baity. I said that in the context of a lot of other things within a conference. And I think of course that feels like much more interesting to publish, which I agree, it's actually not that far from what I was actually arguing. And my reasoning here is kind of actually pretty simple, to be honest. It has to do with the way media and feature films, or just content in general, has been made for the last couple of decades. And it's a process that's filled with a lot of technicisms and processes and budget constraints and time constraints. It's a process that's driven a lot by approvals, by convincing the right people. The industry itself, I would say, is in a tough spot. The Hollywood and filmmaking is mostly because it's getting Just way too complex to approve and make anything. And I'm not the only one who's argued this. I think many people have said this before. You need at least a couple millions just to make a movie, and then at least a couple more millions to promote the movie. And so unless you're willing and able to fundraise or convince someone to give you at least 50 to $100 million to make a movie, then it's not probably going to happen. And of course there's exceptions and this is not a jurisdiction for everyone. But I would say in the large part, I think most people agree with that sentiment and just go to LA and you'll see that all over the place. And the reason there's a lack of interesting projects or a lack of diversity of films is in part because of that. Yeah, what I think is happening now and the new ingredient in the mix as well, what happens when you have a technology that allows you to fast track many of those stages and steps. Right. So it will help you iterate way faster on pre visualization, on pre production, and of course on making the actual frames for the movie. And that's already our scene today. Like AI is getting deployed at scale in studios and also people and independent filmmakers. And what you start to see kind of trajectory wise, is a place where if you have a really good idea, if you have a really good script, if you have something the world wants to hear, you might be very soon in a position where you can tell that story and you can tail it for a minimal fraction of the budget that otherwise will require to you something else that feels like probably at that stage, like insane that we weren't that way. And so my summary of that was like, well, yes, hopefully with $50 million or $100 million, you're going to be able to make a thousand films or hopefully many more. The constraint would not be a technical one, it would be a storytelling one. Do you have something good to tell? If you do, then like technology will help you to it in the first place.
- [04:26]
A
Yeah, you know, I think that like the concern is among many people is that this will just create more AI slop. Right. And as someone who's a writer, I've worked on film sets. There's like a fear, there's a visceral fear that a lot of that will go away, a lot of that industry will go away. That collaborative nature of creating films, on the other hand, I'm not sure if you saw I got an ad for this today, there's a new scary movie Coming out like this, like the. The satire films, scary movies, There's a second practical Magic. Practical Magic is my favorite movie. I don't know that I need to see it again. My point is that there's all this existing IP that keeps getting produced, and other amazing stories are being sidelined because, I don't know, whatever industry mechanisms say that we need to keep producing existing IP with the same actors you've seen for 30 years. And. And that's the only way we're going to get a return on our investment. I'm hopeful that tools like yours can change that. But, you know, before we go into it, I want to take a step back, because Runway, there are other companies in this space, right? Luma is a competitor. OpenAI is making video generation models. So is Google. And one thing that I find interesting about you guys is that you don't have that typical Silicon Valley background, right? You co founded Runway in 2018. You and your co founders, Alejandro and Anastasis, came out of NYU's art school program. Can you talk a little bit about that?
- [05:53]
B
Sure, yeah, I'm happy to. Also, to the point before, I think, on, like, the quantity of content, I think that we've seen that before. I think people sometimes forget this is actually there's a law for this called Sturgeon's Law. It's like the most of the things that we see are low quality and small percentage of it are good quality. But my point is that we've seen that before. The field goes back to the 40s and maybe even before that. But deep learning and neural networks working weren't really a thing in 2016, 2017 and 2018, when we started working on some of the ideas of Runway. And so, in a way, when you speak about competitors and others building here, when we started, there was no competitors because there was no industry altogether. And so there's no one we could like, compare to or look or inspire or think that we were, like, competing with. And I think that helped us shape a little bit of this independence and unique way of thinking about why we do this in the first place. And you're right, we started this as I would say, artists and engineers. We come from a background, the three of us, from a background that mixes art and science. We were either engineers with an interest in art or an artist who wants to become an engineer. And for a long time, if I'm honest, it was kind of strange and weird, and we didn't feel at place in many different times and moments because it was hard to classify us you guys were in art school, but you're software engineers, so you're publishing papers on AI, but also you're like speaking about impressionism. It just doesn't fit. And I think that in a way has become maybe our biggest strength over time is that, yes, it doesn't fit. That's exactly the point. The point is that if you want to make good products and good technologies and you want to create your own destiny, maybe don't live your life by labels, but like create on your own label. And I think with the company itself has embodied that culture over time. And so it comes from a place of care. This is what we have dreamt about doing for years. And so the fact that we can do it as a job feels also like a dream.
- [07:54]
A
Yeah. And you know, another thing that I think is important to point out is that you're based in New York, headquartered in New York. You have offices everywhere, of course, but New York, strong. I really appreciate that as a New Yorker, always fun for me. It means I can visit you in the office, which I did. Lovely office, by the way.
- [08:08]
B
Thank you.
- [08:09]
A
One thing I noticed while I was walking around looking at the way people work is everyone seems to be, I mean, it seems really flat, like your, your desk is amongst it with, you know, all of your other workers desks. And it feels like there's filmmaking people next to engineers and like, are you. How do you consider yourselves? Are you a creative tool, an AI lab? Both, Yeah.
- [08:29]
B
I mean, you think too much about the best way to describe like the label of a company, but if you look at it like very practical. We are a deep research lab, an AI research lab that's pioneering and building like large models. So yes, you could think of us as a large research lab. We also create products and create like exciting tools. And I think we're a software company in some domain. We make movies, we like making shows, we like making just interesting pieces of content. I don't know how you can classify that as a studio, creative agency sort of artist. Again, I think the sweet spot for us is what happens when you have all of those things in the same place. To your point, you can go into the office and you literally have someone who's working Hollywood and visual effects for the last 10 years sitting literally centimeters apart from a research scientist working on pre training on large auto regressive models. And those people, funny enough, can speak the same language, which is not that common, and they can speak that same language technically and from a value perspective, because some of our researchers and engineers are artists, they're musicians, they're photographers. They also make films. And some of our artists and creatives are also engineers. So it's a nice combination of things. And I think, yeah, it's a. I think we've built a pretty special place from a cultural standpoint.
- [09:51]
A
You know, one of your early investors told me that you guys refused at the beginning, at least. I don't know if you're still doing this, to hire people with a PhD at first just because you were like, no, we're doing things differently here. Is that true?
- [10:04]
B
Maybe it wasn't like that. We refused to, like, hire people with physicists. It's more just generally, I think we tend to not look at credentials as much as maybe others might have. Like, some of our best people might not have any of the credentials you traditionally look for. And that's kind of the point. Like, I think talent is equally distributed around the world. And if you happen to get into Stanford somewhere else, like, might not necessarily mean a lot about how much you can do and how much you can execute and build or learn. I think we want to embody that like, it's a meritocracy. Like, we judge you by the outcomes of and what things you can make, not by who you are or where you went to school or who you vacation with, who you know, maybe, because, like, that's where we started. If we judge ourselves by who we knew and where we were and our last names and our titles, then we wouldn't be here. And so it's a representation, I would say, of this very hard work driven meritocracy that we built internally as well.
- [11:00]
A
And I know that that has also had its effects on, you know, how you've had to be lean as a company and how you've had to, you know, start revenue generating pretty early. So, like, really the revenue generation was happening alongside the research. But I do want to talk a little bit about the research, right? Because I was, you know, in the course of researching you, I was looking at some of your, you know, the first video gen models that came out. And, you know, of course, for the time they were impressive for the time, like two years ago, whatever, four years ago. And it just reminded me of that Will Smith eating spaghetti thing. Like, we've come a long way in video generation, right? And one piece of research that I believe came out of Runway was about dual process image generation, right? Having this kind of like, judge vision language model, assessing AI image generators in real time. Is that something that you guys were
- [11:55]
B
working on a couple of things, I guess on the, on how far we've come. Yeah, we've come very far. I mean I think the first ever video model we put out that maybe you've seen some of the outputs was around 20, 23, so like two and a half years ago, which in AI landfill like three decades ago. And it was such a big deal that it was in the front page of the New York Times. We had a whole article if people want to search for it. It's like everyone was amazed that you could render this incredibly low res, incredibly bad by today's standards, like video output. But it was new, it was totally new. No one thought you could do it. And so there we are and today we're able to output like incredibly hyper realistic 4K footage with a level of realism and cinematography that you will have never imagined by just two years ago. Yeah, I think the industry and the field has gone really far and the idea for us will be like, well, hopefully in three more years we'll look back at today and the models we have today will feel like the model we had two years ago. And I think we're heading towards a trajectory very soon from a research perspective. We do a lot of really interesting research. We publish a couple things. We don't tend to publish all of it for mostly competitive reasons. We do have research in the video domain and image domain. We publish also things in biasing. So how do you make sure that the models, maybe that's the one you're referring to. It's how do you make sure that the models can become a bit more self aware of how they're depicting things in the real world? Because the models are trained in such a way that they might reflect how humans see the world. And so if you want to have a more diverse set of outputs of a model, then you need to improve the way you train the model. And that's something we've spent time and spend significant amount of resources to make sure the models are both from a product but also from a graded standpoint and diverse standpoint interesting enough for people to actually use.
- [13:48]
A
Zooming out a little bit. So Runway has become the go to for filmmakers video editor. As you've raised hundreds of millions in funding. Pitchbook says somewhere close to 86060 million at a 5.3 post money valuation. Your last round, I remember covering that and it was with the angle that you're, you're trying to go deeper on world models. Can you explain why there's this push we're seeing just generally from video to world models. Like how do those two meet? Because Google trains theirs separately. They train VO separately than Genie. So I'm kind of curious how you see them moving together.
- [14:26]
B
Yeah. To the point of not having necessarily own your own destiny that I was speaking before we spoke about this idea of general world models back in 2022, 2023. And so relatively early within the lifetime of what we now consider a common trope of AI, which is how do you build these general world models or world models? And I think we build our own intuition as to what that actually means and how we think about them. Maybe an observation is, I think generally world models have become like a catch all term to describe anything. I saw a legal company recently describe their work as a world model. Then I'm like, okay, you're kind of stretching it a little bit.
- [15:03]
A
So how do you define a world model?
- [15:05]
B
So for us, it's always been around this idea that you could train models that can understand the physical world in perhaps a similar way that we humans do. And there's an argument to be said that the best path for that is you need to show the model or models like enough pixel or video data. And if the models can understand how videos should be or how videos are created in the real world, they're naturally understanding or becoming aware of things that occur and happen in the real world. Physics, gravity, cohesive reflection, cause and effect. Those are things that we don't tell our models to learn from. We don't tell a model, hey, when you throw an object, the object should move at this velocity and with this set of vectors. And this is why the model just implicitly learns about it. And so in a way, video has become the most obvious representation of how you can train AI systems that can understand, act and behave in the world in similar ways that we do. And that's ultimately, I would say a lot of the holy grail over time is you don't want models that can describe reality. So you think about language models as models that can describe reality because they train on text and taxes or representation, like an obstruction of like reality that we humans have created. Training on video alone is a bit more pure because you can simulate reality. You don't have to describe it, you can just simulate and unlearn from that simulation. So that I think for us has been a much better path and philosophy of research than just fully focusing on language alone.
- [16:36]
A
Are there certain things that LLMs do today that you think could kind of be replaced with just with video? Every founder struggles to find the right Domain name. Most are taken, the rest feel like compromises. But you shouldn't have to settle. If you're building in tech, your domain should clearly reflect that. And there's no better name than tech. That's why startups like Nothing tech and 1x tech are already using tech to define their identity. From day one, it's clean, credible and says exactly what you're building. Secure your tech domain today from any registrar of your choice.
- [17:13]
B
Yeah, I mean like visual reasoning and visual planning. So that's why some of the most interesting compelling use cases of world models is in the robotics side having a system understand how to navigate a very particularly complex action in the world. The way you can think about it is I'm going to have a language model. Just do vision and try to reason about it with language. Another one could be more like we do. Like when, when you're tasked to do things in the real world, you're not, you don't have an internal monologue all the time about how things work. Like sometimes you drive and you don't even remember how you drive. You just like you drove, you know. But you have an internal implicit understanding of the world into intuition. Right. So but because you've, you've seen enough times, you've drove in enough times, you've seen it that your brain is kind of hardwired to do it. And there's, there's not going to be said that there are a similar way to train artificial networks and systems that can do that same kind of behavior. But now you feed it videos and instead of language alone, have you started
- [18:13]
A
training single models on multiple modes like voice, text, video, all at the same time?
- [18:19]
B
Yes, we train what we call omni models. So omni models are models that have kind of the combination of those things together? Yes.
- [18:26]
A
Okay, so for world models, the near term use cases that everyone's talking about, I mean others in the field, I mentioned Google before, there's world Labs by Know Fei, Fei Li's company, they're pretty focused on, on world models. It seems like the trajectory is entertainment use cases first. Right. Like gaming, interactive entertainment and then robotics. And I know Runway has, has started doing a robotics push as well. Is there something beyond that?
- [18:53]
B
There is, I mean we have, we have like three main kind of like product, I would say focus. One is, I call it linear media. So linear media is videos and films and short stories and ads and just content generally. Then there's like non linear media. Nonlinear is real time. So what happens when can create this kind of videos But I don't have to wait. I can just like render them in real time. So this is real time. You're asking me a question, I reply back and I'm aware of your, your video. You're aware of mine. You can ask me to do things, you know. Yeah.
- [19:22]
A
So you actually, before you move on, talk a little bit more about that because you talked up. You guys recently released Characters, right? As a.
- [19:28]
B
Correct.
- [19:28]
A
As a product for real time video and I'm seeing this crop up a lot. I recently was at Humor Luminex, I interviewed Pika. They are doing real time video agents. Right. So I won't go into it in too much detail but it's essentially a way to create, I think for influencers a real time video agent that's like a maybe a twin of an influencer that other people can interact with. Interesting. But then you know, I also interviewed Descartes and they're doing, you know, they're using real time video for things like virtual try ons. So there's a lot of e commerce use cases and I'm curious what your seeing for characters so far.
- [20:08]
B
So first of all it's great that more companies are paying attention to this. I think it helps the ecosystem focus more. I think similarly with world models there's like this. It's hard to agree on what it means and how it can be used. It's general purpose and people might have different ideas and things. If you want to build influencers around it, I don't think that's necessarily bad. But that's one way. There's many other ways of doing things. I think for us it's mostly been about building software or models that can feel as human as possible in the best possible way we can. And so you can think about learning. For example, one of the things that is a huge gap in terms of what models can offer students of any type of age or form or kind of content in terms of helping you learn through things like the ability. And this is proven in many studies that one of the highest predictor of success of a student is having a tutorial, having someone that can be there as entering or personalizing kind of your, your education. And of course like it's impossible to do that at scale. Like how are you going to get everyone in the world a tutor unless you're like extremely rich. Right. Well that's the role of technology. Like there's a path where anyone can like have a tutor that's there 247 that understands the preferences that's never going to get tired, that can resay the things over and over again. And I think a lot of what we're seeing right now with characters and some of our models are heading towards that direction of this personalization aspect. And of course you can think about personalization in many different ways. And that's why I call it nonlinear because you don't begin and end in a very specific, predetermined way. It's an open ended world. This conversation can go anywhere we want to take it. It's open ended, it's non linear. That same ability wasn't really possible with the technology until very recently. You couldn't just open your computer and navigate an open ended world because that was not possible until very recently. And so I think a lot of the interesting use cases and applications start to emerge as this becomes like more accessible to more people.
- [22:08]
A
And you know, you guys have your, what is it, the Innovators lab and also your builders program. So you're putting this out into the world. So I think startups can experiment and kind of show you where the use cases are, right?
- [22:22]
B
Yeah, correct. I think it's recognizing that it's a fairly new technology. I mean it's relatively like weeks, months old. And this is for me, from an artistic, creative standpoint. If you look at the history of some of humanity's largest and biggest cultural and scientific revolutions have been people tinkering with things. You give them stuff, you give them technology and you have instead of the creator, figure out all the things that could do, you give it to someone, right? So think about, I mean, cameras. Cameras were a scientific breakthrough supposedly meant to be used to study human motion until some people just thought it was good to make theater with it. And then like, then here we are, right? Like it's a consequence of just tinkering. And so we want to promote more of that experimentation and tinkering specifically with early stages of new models.
- [23:08]
A
Is characters available via API?
- [23:10]
B
Yes, you can use it via an API.
- [23:12]
A
Is there anyone you wouldn't let use it? Like I imagine my first use case that comes to my dystopian brain is this is like a companion thing, right? So people can like just have their virtual girlfriend that they, you know, people are already doing this, right? So like now they want to face.
- [23:27]
B
But why do you think that's dystopian necessarily? I could see that. I could see they're going to feed being like, I guess more of a bad thing.
- [23:34]
A
But like, I think technology has isolated humans in a lot of ways, right? We have the loneliness epidemic That a lot of people point to social media as the cause, which is ironic because social media was meant to bring us all together, but really it's made a lot of us feel more isolated. I think that if you have an always on companion that agrees with you and doesn't push back and is just like everything, you want it to be a fantasy, why would you bother with real people? Some people. Why would some people bother with real people? You know, I like the conflict, so,
- [24:06]
B
so I, I can see that. But I actually kind of disagree the premise, to be honest. I don't think, I think reality is way more complex than it being explained by one single variable. I don't think whatever issue we have in society at every single point, I think it's very rarely explained why one thing alone and one thing on its own, it's a combination of factors. Has to do with the cost of living, it has to do with birth rates, has to do with globalization, it has to do with technology. Many different things that are coming together, I think sometimes find the easiest, find a culprit and find one thing that feels like easy to target, which is like, I don't know, technology. And I think AI is a little bit of this because AI, it's such a general purpose technology, it can be used for so many different things. It can be anything you want, literally can be. Take any shape that you want. And so in a way it becomes a mirror of our own state of mind. Because if you're, I don't know, in a bad state because you think the economy is heading in one direction, AI, like if you think Hollywood is cooked because people can get paid AI, if you think, and of course it's not about that. There's way more things that explain both of those things and many things. But I don't think there's one thing that will explain everything in our society and I don't think AI will also be that one thing. And maybe more specifically to your point, of course there's misuses of this technology in real time and characters and having this. I agree it could go in bad ways if people want to take it that way. And hopefully our position is like we'll collectively try to prevent that from happening. The score is going to be people who are going to find ways of misusing it. But I think we're all largely the outcomes and the impact of the positive side of things will overweight and overcome any other kind of negative outcome of it.
- [25:54]
A
And are we talking just broadly about
- [25:55]
B
AI though, or just broadly about characters in general again, if my daughter can have a 247 personal tutor that can teach her anything she wants when I'm not there. I found that amazing. Like, I wish I could have that. Like, I remember myself at school, like having questions like, I don't know when to answer this. I'm gonna have to wait, I guess, until like next day when I can ask my professor if I have time to help me do a thing. And now I learn stuff like all the time. I don't have to, you know, so that's where all making us better. And I think it's hard to speak about that or like focus on that because it doesn't. It's not as interesting as like, oh, this focus on the dystopian things. And so sorry for pushing you, but
- [26:33]
A
a little bit, I think it is as interesting. We could probably have this philosophical debate all day, but I don't want to get. I don't want to get wrapped up in it. I want to briefly. So you said you had like three different buckets, right? And this is the nonlinear one, focused on video generation now, real time video generation. The next bucket is the world models, which we've already talked about. What's the third bucket?
- [26:53]
B
So no, it's more think about world models as the foundation on which you build like the bucket. So world model is our technology. It's how we build this. And we've been building it for the last seven years, right? So that's our basic first pillar. And then you have what we call linear, which is films and media and like linear stuff you read and it's the same beginning to end and you're going to watch it again. Nonlinear media, which is what we just described, characters and avatars and world building. And you can explore things as you go. And the third one is what we call physical AI, which is what happens when you start deploying this base model not only on storytelling and nonlinear, but also on physical systems like robotics. And so we're basically building towards those three main kind of dynamics and not
- [27:33]
A
just for training robotics, but actually being like the software of the robot itself.
- [27:37]
B
Right. So a version of that. So yes, you can create synthetic data, you can augment data that you already have, and you can also test how good or how effective the policies of models of robotic systems are. So you do like a. It's called sim to sim or simulation to simulation in a way. But yeah, it helps you basically create better systems that can navigate the real world.
- [27:56]
A
Now you guys started out, you know, obviously wanting to give filmmakers superpowers Right. And your company has grown so like the remit has grown so much so fast. And I feel like that is a side effect of just the research. Right. And like what you've learned watching video generation scale. I'm curious, what does it take to get to a state of profitability given how new this technology is, how much research it still requires? The cost of compute.
- [28:26]
B
Yeah. So one note, I guess we didn't start necessarily with trying to make filmmakers more powerful. I think it was mostly about how do you build the original thesis and still is, how do you build AI models that can augment human creativity? And human creativity can take different shapes and forms. And I think that's still at the center of what we do. That's our DNA. If you're successful at building tools and AI system, talking about menhuman creativity in film and entertainment and learning and robotics, many other things, if that's valuable, people are going to be willing to pay for it. And that's what we have seen so far. We've seen the largest growth of runaway just over the last quarter. Every week has been a consistent set of new record in terms of revenue growth and sales growth over the last couple of weeks.
- [29:09]
A
And is this enterprise or prosumer or where is most of the growth coming from?
- [29:14]
B
It's enterprise and prosumers. Yeah, I would say the focus has been mostly on that side of things and businesses are getting ready and they actually find value in it. Again, if I went back and we have this conversation seven years ago, people were not finding value because the models were not useful and no one really cared. Now I can show you in 10 seconds so many valuable things if you're an advertising company or a brand or an agency or a studio. And so now you put budgets and you're willing to spend money because it helps you. So it's not anymore ethereal. It's very practical and the challenges are around compliance and like how are you storing, how are you kind of collaborating with people? What prompts are you? It's like a lot of other things that had not to do with like kind of models. The thing it's like, well, how do you actually deploy this in companies? And for us that's how you build a sustainable business that you have to just have those two things work hand in hand.
- [30:07]
A
Well, we're just about out of time. Where can our listeners connect with you online?
- [30:11]
B
Probably on Twitter. I tried to share as much as I can on Twitter and of course on our website. I mean if you want to learn about Runway. Just go to runwaymail.com and we'll have a lot there for you.
- [30:20]
A
Well, thanks again so much for joining us on the show. To our listeners, you can find me on Everywhere. I'm Everywhere. I'm on x, I'm on LinkedIn, I'm on Bluesky and you can find Equity at Equity POD on Bluesky and X. Talk to you next time. Equity is hosted by TechCrunch senior reporters and produced by Teresa Loconsolo with editing by Cal. Subscribe on YouTube or wherever you get your podcasts and find out what's next at techcrunch. Com events. Thanks so much for listening and we'll talk to you next time.