Please welcome to the stage Paramount CTO Phil Wiser and Runway co-CEO Cristobal Valenzuela.
Everybody. Everyone, it's me again.
Yeah, sorry. Just for a minute and back. Well, I'm really excited to be here chatting with Phil. I think Phil, we know each other for quite some years now and there's a lot of conversations even yesterday we were like chatting about a lot of things hopefully we'll be able to chat today. I think how you've been thinking about AI and how things are heading, where things are going to go.
And actually, I don't know if you guys know, but Phil has a rock band and he's like really really into music. And so, yeah, it's pretty good rock band. You should put on your music video at some point.
That's my only qualification as a creative. We'll talk about that.
But you've been in the intersection of both technology, media, art for quite some time. You started a company, you sold that company, you worked at Sony, you're at Paramount now. And you probably have seen waves of transformation come and go with time.
How do you think this wave of where we are right now compares to the ones you've experienced before?
Great. Yeah, and first of all, thank you for letting me come and talk to you. Secondly, congratulations on all that Runway's done and we'll talk about that a bit because we first spoke a little over 3 years ago. You know, I saw this opportunity and we'll dig a bit deeper into why, you know, for me anyway, my intuition, you know, sensed that this was the one. But yeah, at that point, you know, Chris had an idea and was building out some very, you know, basic ideas around video generation and the way the companies executed. We'll talk about it. Like that makes the difference in the market. So, you know, congratulations on that.
And just building on the background, just so you have some context of what I'm talking about and the perspective I'm bringing, as Chris mentioned, I've started a couple of Silicon Valley internet media companies over the years. So, I come at it from outside of the media industry as well as inside the media industry. And my music experience was starting the first company to sell music over the internet, which really caught the internet 1.0 wave. But then I got to go inside of a company, Sony Music, and watch them deal with it. And it was not pretty. If anyone lived through the Napster era and saw what happened inside of big companies that didn't see it coming or just resisted the change, it's not a pretty outcome.
And I think that is, you know, hopefully people are learning from those changes. But then getting to, you know, your question around this technology trend, look, I'm going to put this in the top 10 technology trends of all time. And I would even go as far as top five. And when I say top technology trends, I mean like going back to fire.
It's a good frame of reference to have.
It is. But I think it's important because we all look back and say, 'Well, is this more like the cloud or is it like the internet or maybe like mobile?' I think those are important, but those are all, you know, computation sub-revolutions, right? And this is really building on top of that. And I say going back to like things like fire and the printing press because they have a couple of values. Like those big changes rapidly decrease the cost of something that humans do. Whether it was at that point, you know, writing things, or it was manual labor with a steam engine, this is rapidly decompressing the cost of, you know, some forms of knowledge work.
The other thing is, you know, what are you externalizing and what are you making better for humans? And in this one, a lot of that knowledge work not just being compressed and frees us up to do more. So I think, you know, that shift is one. The other is the economic impact. Like this there's a reason there are multi-trillion-dollar investments happening around this right now because there is a, you know, a lot of people that are doing financial analysis way beyond me that see the overall impact. So I do view this without being hypie or hyperbolic, as what, you know, one of the greatest trends of all time.
Was there a moment where you felt like it was beginning to become that trend? There was a, like was or is more a continuous evolution of over time of things that pre-act in terms of
It was actually 10 years ago. Like when the basic transformer architecture was introduced, when we had models that could actually look at images and recognize them. And I actually built an AI model. I was at Hearst and we had a lot of fashion companies. And even, you know, a little over 10 years ago, we built a model to go in and just look into our image archive and recognize brands, different styles of clothing. And when we could do that very simple bit, it could see that that was going to be transformative. I actually held an AI summit at Hearst in 2017.
And brought in all the leaders of AI at that time from Google, Amazon, and others to kind of wake up that, you know, very diverse media and information company to what's possible.
But what I didn't see was timing. That's always the hardest part. So
Yeah. So I guess how do you when do you know exactly when you're experimenting or testing versus now when you're like moving into production and value? Is there how do you distinguish between those two things?
Well, yeah, when there's a new technology that comes in, like distinguishing what's going to be, you know, substantial and lasting versus what's going to be ephemeral. I look at this First of all, consumer adoption is like the best metric. But even consumer adoption can be ephemeral. So I then look, you know, a bit deeper than that and say, well, what type of dependencies are there for broad adoption on that.
And I'll go back to internet music as example because when the internet came, you know, I was like, 'Hey, we're going to sell music over the internet, democratize music creation and distribution.' But, you know, I had to look at all the dependent technologies. At that point, computers could not play back an MP3 file in real time, which is like, 'What's that mean?' They weren't powerful enough to play back an MP3 file. So, you had to look at compute, storage cost, internet bandwidth, how you can compress all like all those dependencies to see was going to actually get a scape velocity as a new technology trend. I think the same is true here. There's a lot of things that had to be in place and part of it was just compute infrastructure that's happening right now.
But, I would say like of all the people I've met and dealt with over time, Steve Jobs was the greatest at this. Like because he would sit and wait until all those dependencies lined up and he saw a track record. Like there was a track record of some level of consumer adoption, a track record of some technology barriers being knocked down and then he would come in, snap it into an amazing product and launch it. So, I think we haven't gotten there yet, right? I mean, you could argue ChatGPT was that moment. I'll say it's not. I mean, but there's a future in front of us where there's going to be amazing products that are that next iPhone moment.
Interesting. I guess you just feel a little bit like you're spending you are a technologist that hard working in media. So, you have like maybe I think of you as like you have these two brains of art and science like working together in some way. Like but it seems like maybe what you're describing is just like always tension between like short-term focus of what the industry is working right now and I think media at large is going through a major transformation, consolidation, and change. So, you have this short-term like medium-term momentum and then this long-term bets that are sometimes just hard to see and maybe you're saying like we need to project and kind of measure and see how do you spend your time between those two things? Between like what works in the industry right now or what seems to be the focus and like what's ahead, do you see them at odds or not?
I don't, but there's a couple of things you hit on there. Like one, when you're trying to drive long-term strategy to big company, I use this kind of breaststroke metaphor of people that swim, you know, you pop your head up, look where you're going, and you pull really hard. That's how I've always operated. Like you take that moment to get your bearings and say, 'What do I really believe in this moment is going to be important for the next 6 months, year, that's going to be a tack toward a 5 to 10-year trend, and then just go super deep and push.' And then pop your head back up and see where you are. And sometimes that's when you kill things like Sora, right?
But I had to say it. Like I was Was I the first one that did it today? See? But that's how I operate, certainly even in big companies and also in startups that I'm running, because I feel like if you're always just floating on top, you don't really get a sense of deep intuition of what's working with your teams.
So I definitely like lead from the front and the back at times. But I think you have to have both if you're going to be successful in a big company or you're trying to build an emerging company. They're not at odds with one another.
