00:36:39 · Sachin, Anastasis Germanidis, Adam Miller
[00:36:39] S1: Where does filmmaking now sit? When I think of Runway and you walk around the office, you see all these wonderful films on the walls and, you know, it feels very much like at the ethos of the company with the filmmaking festival you launched. Where does now with this kind of expanded directions filmmaking sit? [00:36:59] S2: Film and storytelling is very central to Runway and will continue to be. I think we have this motto that the best films are yet to be made. I think that's... We're still... Like, we we still are at the, you know, day zero, very early days of how, the craft of storytelling and the craft of filmmaking will change as a result of those models. I'm also very excited about all this new form of media that are gonna emerge the moment you you have, interactive video models and the way the video... And and the the moment the video models improve even further. So imagine there's all these kinds of new media that, exists between film and games and theater, that are yet to be invented. I think we're very excited to see how our, kind of world models, interactive video models are gonna help to craft new kinds of new kinds of forms of storytelling. [00:37:56] S0: And you've said before that you think language models won't solve a lot of the fundamental problems we have in science and technology. With world models, do you think we'll be able to solve a lot more problems, whether it's in science, mechanics, any types of other manufacturing even?
00:38:13 · Anastasis Germanidis
[00:38:13] S2: Yeah, and we're seeing already some initial work towards that with, how we're using our world models and physical AI and robotics, I think but I think the there is just so much room to further scale world models, the way, like my my vision, I think where we can, grow in terms of our models and our research is there is always observations of the world whether it's, you know, from whether it's real world human scale video to any kind of data that we gather from scientific experiments, across every scale of of the world that we haven't trained a single model on all of those, modalities. And the moment we're able to do that, those models can help us run experiments inside the world models much faster than we would in in the real world. The bottleneck to scientific progress or a big bottleneck is less on building this, genius artificial scientist. It's more about running experiments quickly and getting signal from the real world about what works, what doesn't work. So being able to use world models, whether it's video based world models or world models that are trained on biological modalities or physics simulations in order to be able to run and kind of multiply the number of experiments that you can run, think will have a very significant impact to, how kind of scientific progress continues.
00:39:53 · Adam Miller, Anastasis Germanidis
[00:39:53] S0: Thank you for the interview. That was incredibly inspiring. And I think, like, your company is, for the last few years, just blown us away with all these video releases. And I'm sure there'll be new releases, especially with the world models that would just blow our minds as well. [00:40:07] S2: Yeah. Thanks, guys. [00:40:08] S0: Thank you.