Runway also enters the arena of advanced generative AI models for video. He presented Gen-3 Alpha, which for the moment is not yet available to the public but which was shown in some clips.
Gen-3 Alphasuch as Sora by OpenAI and the recent Dream Machine by Luma AI, will allow you to create coherent videos with a simple textual description of the scene you want to reproduce. We remind you that “coherent videos” mean those films that respect the correct movement of the subject between one frame and another and represent physics in a way similar to reality.
Gen-3 Alpha was trained on video and images and, when available, will serve as the basis for all Runway services, i.e. Text to Video, Image to Video and Text to Image, and existing control modes such as Motion Brush, Advanced Camera Controls, Director Mode.
Runway was the first startup to make a name for itself in video generation through its web platform, but the arrival of Sora has raised the bar significantly, even though OpenAI’s model is not yet open to the public.
The incredible qualitative explosion of AI-generated videos. Will we still be able to believe what we see?
Go to further information
The clips published by Runway of Gen-3 Alpha they definitely show higher quality than Gen-2. In our opinion even by several orders of magnitude (we recommend scrolling through the videos and not limiting yourself to the first clip, which sometimes isn’t even the best).
As for security, Runway says Gen-3 Alpha will be released with a new set of safeguards, including a new internal moderation system and the C2PA protocol for encrypting details about the origin of digital content.
Runway has given no indication of the public availability of the new model and, for now, provides a contact form for companies that want to customize Gen-3, i.e. obtain more coherent and stylistically controlled characters and satisfy specific artistic and narrative requirements.
