How Runway is pushing synthetic video toward reality

In the space of just a year, text-to-video AI has improved hugely in the quality and usability of its output—so much so that synthetic video seems less like a gimmick and more like an eventuality. The New York-based startup Runway has made huge strides on the quality of video its models produce based on text prompts and other input.

In February 2023, Runway released its Gen-1 model, which could apply the composition and style of a sample image to an existing video. In March it announced its Gen-2 model, which can create a totally new four-second video based on just a text prompt. By August it added a new option to create longer, 18-second videos.

The company updated Gen-2 again in November, this time with an obvious leap forward in picture quality and smoothness of movement. Anastasis Germanidis, Runway’s CTO and cofounder, has said that his company’s north star is an AI system that can generate a two-hour film with high picture quality.

Runway now faces stiff competition from OpenAI and its Sora tool; it’s likely both companies will continue breaking through research barriers on the way to that goal, and this could have huge implications in Hollywood, Silicon Valley, and beyond.

Explore the full 2024 list of Fast Company’s Most Innovative Companies, 606 organizations that are reshaping industries and culture. We’ve selected the firms making the biggest impact across 58 categories, including advertising, artificial intelligence, design, sustainability, and more.

https://www.fastcompany.com/91033528/runway-most-innovative-companies-2024?partner=rss&utm_source=rss&utm_medium=feed&utm_campaign=rss+fastcompany&utm_content=rss