Cristóbal Valenzuela on a Culture of Shipping, the Artist as CEO, and Running a Full Stack Company
Cristóbal Valenzuela is a developer, artist, co-founder, and CEO of Runway, a research-driven company building the next generation of creative tools powered by artificial intelligence. With Runway, you can—among many other things—generate novel videos with text, images, or video clips. You’ve probably seen output from Runway in advertisements, viral videos, or Oscar-winning films like Everything, Everywhere, All at Once. Cris studied and met his co-founders at NYU’s Tisch School of the Arts ITP program, where he eventually worked as a researcher after graduating. Cris and his co-founders purposefully structured the team as an applied research company. By seamlessly merging technical expertise with artistic vision, they set out to reshape what's possible in the realm of creativity.
The following quotes are based on an interview we did with Cristóbal in May 2023.
Applying machine learning to creativity
Around 2015, a weird and surreal image generator captured my heart: Deep Dream. It was a computer vision program created by Alexander Mordvintsev from Google. Deep Dream identifies and enhances patterns in images, creating dreamy and psychedelic-looking images. The algorithm was originally created for interpretability or explainability in deep learning, aiming to shed light on the internal workings of the neural network and provide insights into the features and patterns of neural networks. But it was swiftly repurposed as a creative tool. I quickly became fascinated with the idea of exploring how novel computer vision and graphics ideas could be used to build creative tools that were unthinkable or impossible to imagine before. That ignited a curiosity and personal fascination to understand and eventually build completely new types of creative tools for artists.
An artistic approach to a technology company
Within our team, you'll find a remarkable convergence of talents—an ensemble of individuals that straddle the realms of engineering and art. Many of the people in our team are artists in their own right, from painting, ceramics, and music to poetry and creative coding. We have assembled a team of artists, researchers, engineers, and innovators who are passionate about their work and their craft. The expertise comes from the ability to merge art and science smoothly together.
This shared foundation has proven to be a unique asset for Runway. It's easy to get swept up in the buzz surrounding AI. Yet, amidst the noise, it is important not to lose sight of a fundamental truth: these tools, at their core, exist to amplify human creativity. That is the essence that truly matters. And who better comprehends the nuances of human creativity than the artists themselves?
I met Anastasis and Alejandro at NYU ITP. ITP is a unique program inside the Tisch School of the Arts aimed at exploring the frontiers of technology from an artistic point of view. The best way to think about ITP is to consider it an art school for engineers and an engineering school for artists. It’s a place where you can have deep learning researchers sit right next to conceptual performance artists.
The three of us have a mix and blend of backgrounds in engineering, research, design, art, and business. Since we met, we immediately found a common ground that binds us: a shared understanding of how to build things.
Being an immigrant founder
Being an immigrant has been both a blessing and a curse. On the one hand, there are a lot of logistical challenges. It's a journey filled with time-consuming processes, obstacles, and bureaucratic hurdles—each requiring its fair share of patience and perseverance. We've navigated a labyrinth of paperwork and barriers just to secure the opportunity to work on Runway inside the US. We had so much figuring out to do at the beginning. We care deeply about building a diverse and inclusive workplace which also includes having people from different cultures and supporting immigrants from various countries.
On the other hand, understanding how much effort it takes to be an immigrant builder means you cherish the time you can devote to building the company. The ability to figure things out, has become a mantra inside Runway. Do you want to build something hard that you’ve never done before? Great, it’s very simple: just figure it out.
Being a full-stack startup
At Runway, we embody the spirit of a full-stack AI research company. From the deepest depths of research and development, all the way up to crafting delightful product experiences, user interfaces and branding, we traverse the entire spectrum. This approach gives us visibility and control over every aspect that requires adaptation. Whether it's at the model level, infrastructure, or product development levels, or even in our collaborations with film studios and our own film festival.
Of course, there are limits to our full-stack capabilities. We won't be venturing into the realm of building GPUs, for instance. Certain functions can be outsourced or supplemented with off-the-shelf solutions without compromising our core organizational structure. But this full-stack approach allows us to have precise creative control over every aspect of the company. From research to marketing - for example, we even created a small creative agency that handles everything from branding to video post-production. This internal team builds and shapes our brand's narrative. This creative team has done an incredible job at crafting unique narratives and stories and that’s largely because they get to see how research works and understand the technology behind our products.
I often get asked what marketing agency we use because folks like what we are doing. Many investors and other companies reach out, asking us to connect them with whatever agency we use. I got so many of those requests that I eventually told folks we used an agency called Pasarela and started sending people to their site. Pasarela stands for Runway in Spanish, and the site redirects to our careers page.
How Runway is run (and organized) differently
We structure our company around ensembles—small teams dedicated to specific mandates or goals. Each of these goals aligns with a larger master goal we set at the beginning of each year. We prioritize flexibility and adaptability by frequently rotating people across ensembles whenever a major goal requires focused attention.
Measuring the success of each ensemble varies depending on the context. Some ensembles have more straightforward metrics, such as growth or revenue. However, other ensembles, like our alignment and safety ensemble have unique challenges in how to measure their goals. Creating a new model, for instance, isn't something easily quantified by a KPI since the model itself can reshape the very nature of the KPI.
We don't rely on any specific framework; instead, we place our trust in our collective institutional knowledge. We believe that truly disruptive innovation cannot be achieved through simple 2x2 matrices. It's the focus and the team, not the process.
Processes and frameworks are changing often. Very often. We’ve been at this for years. Our accumulated knowledge runs deep, encompassing insights into building models, conducting research, and establishing effective organizational structures.
"The core strength of Runway is not a product—it's the way we build products."
How to react when Bigtech comes after your market
It’s important to remember that the core strength of Runway is not a product—it's the way we build products. We operate in a competitive market that will undoubtedly become more crowded over time. It's inevitable that others will attempt to recreate what we have already accomplished. But if you're afraid of that, you're missing the point.
Runway is not a product. Runway is the collective effort—the humans behind the organization building something. I read a wonderful quote from Elad Gil exactly about this point that summarizes how I like to think about competition: “Companies that innovate early keep innovating. The ones that innovate late never innovate again.”
The future of AI-driven creativity
I hope I don't sound like a broken record, but the first thing to remember is that we are still very early on in this journey, both from Runway's perspective and from a general research and industry standpoint.
Secondly, generative models are already changing everything and will continue to do so. The best movies are yet to be made, and the best stories are yet to be told. There is an abundance of stories and creativity that we haven't had the opportunity to experience because the necessary tools were not available. What excites me the most is the pivotal moment when we place these incredibly powerful and expressive tools into the hands of more people. The stories we will hear and the journeys we will embark upon will be truly extraordinary.
Think of AI as a new type of camera—a remarkable system that can unlock new dimensions of storytelling. However, just like having a traditional camera doesn't automatically make you a filmmaker, simply having an AI tool won't instantly turn you into a master storyteller. It's crucial to understand how to wield this tool, how to harness its potential, and how to shape its output. Like any other tool, you need to have an iterative process with it to truly comprehend the craft.
Looking ahead, there may come a time when it no longer matters to specify if something uses AI or was created with a generative model. We're moving towards a world where generative models seamlessly integrate into our lives, much like the internet has become an essential part of our daily routines. You don't announce that you're logging onto the internet for work; it's simply a natural part of your workflow. Similarly, AI will become so ingrained in everything we do that we will no longer refer to it as "AI" but simply as tools—tools for human creativity.