Shelly Palmer: Midjourney Launches Its First Video Model

David Holz, founder of Midjourney, emphasizes that their new video generation model, “V1,” aims for aesthetic control rather than photorealism. Based in Nice after Cannes, Holz notes a division in attitudes toward AI: some embrace its transformative potential, while others question its reality. V1 produces 10-second, 24 fps clips from various prompts, showcasing dynamic motion and diverse aspect ratios, yet it operates more as an artistic tool than a realistic video creator. The alpha version is currently private, and there’s no set timeline for public access or pricing as safety is prioritized. Midjourney joins the ranks of OpenAI, Google, and Runway in the competitive text-to-video space, each with unique approaches and capabilities. Google’s Veo 3 is currently the only model available for wider usage, indicating that the race in this field is just beginning. Feedback and discussion are encouraged among interested communities.

Source link