Runway’s technology can make short movies based on a few words of text.
(Bloomberg) — Runway AI Inc., whose artificial intelligence software can create a short video from just a few typed words, has raised $141 million in funding from Alphabet Inc.’s Google, Nvidia Corp., Salesforce Inc. and other investors.
Runway’s new funding, which the company plans to announce Thursday, is an extension of a $50 million Series C financing round in December led by Felicis Ventures. With the new cash infusion, Runway, which has been making AI-based film and editing tools since 2018, has raised about $237 million.
Runway is at the leading edge of a global surge in interest in AI. So far this year, investors have poured billions into AI startups, particularly companies that focus on generative AI, which takes inputs like written prompts and generates new content such as a poems or images. Yet while a growing number of startups have produced AI chatbots and image generators, video is still very much in its infancy and could present a new frontier for AI technology.
Runway’s latest deal values the New York-based startup at $1.5 billion, according to a person familiar with the matter, who asked not to be identified because the information is private. In December, the company was valued at $500 million. Insider earlier reported some details of the fundraising.
In an interview, Runway co-founder and Chief Executive Officer Cris Valenzuela said the company plans to use the funding to conduct additional research into video- and image-generation AI models.
It’s much more complicated to generate a video from a written prompt than to create a single image, à la the systems available from companies such as OpenAI, Stability AI and Midjourney. The difference is a bit like taking a single photograph vs. shooting a video. The intricacies of physics, for instance, must be accounted for when generating a video for a prompt such as “waves lapping on a beach at sunset.” These are the types of challenges Runway is still working on.
“I think the key word for us is consistency – improving consistency across the generation is key,” Valenzuela said. “But also, together with consistency, you need to improve the controllability of the models.”
Runway’s newest product is Gen-2 — an AI system that can spit out a short, silent video in response to written text — now available online, through an iOS app and in a Discord channel. That tool followed Gen-1, which requires users to upload a video as an input source that is used (along with user guidance such as a text prompt or a still photo) to generate a new, silent, short video. The company also makes web-based editing tools for filmmakers and helped create the original version of the text-to-image AI model Stable Diffusion.
Jensen Huang, CEO of Nvidia, said in a statement that Runway’s technology could change how artists work and tell stories. “Generative AI is transforming the content creation industry,” Huang said, and “breathing new life into stories and ideas that were not imaginable.”
More stories like this are available on bloomberg.com
©2023 Bloomberg L.P.