As it is stated here Gen-3 Alpha is the first of an upcoming series of models trained by Runway on a new infrastructure built for large-scale multimodal training. It is a major improvement in fidelity, consistency, and motion over Gen-2, and a step towards building General World Models.

Based on the results we see, we should be excited about upcoming Text to Video, Image to Video and Text to Image tools that will be available for us to use. You can take a deeper view on their products here and try runway for free here. I don't know about you but I can't wait to check the upcoming models they have to offer!

For those of you that have not yet got the chance to know Runway, it is founded in 2018 by Cristóbal Valenzuela, Anastasis Germanidis and Alejandro Matamala-Ortiz. As Paul Drews and Emily Zhao writting at Salesforce ventures. The initial idea came out of Cris’ thesis project at the Interactive Telecommunications Program at NYU, where he met his co-founders while researching applications of machine learning models for image and video use in the creative domains.

Informed by their own experiences as artists, the Runway co-founders set out to answer the question of how a well-built digital tool could simplify the approachability of complex ML models and circumvent the need for deep technical background to give better access to state of the art machine intelligence to artists and designers. The mission of Runway was to democratize access to machine learning so more people can start thinking of new and creative ways to use those models.