OpenAI has launched Sora 2, an advanced video and audio generation model focused on enhanced physical realism, multi-shot consistency, and user control. This model marks a significant improvement over previous versions by effectively simulating cause and effect in diverse scenarios like gymnastics or basketball, incorporating realistic physics. Users can upload their likeness for custom “cameos,” enabling personalized integrations into generated scenes while maintaining control over their appearance.
The accompanying Sora app, currently available via invite in the US and Canada, encourages creative expression with a non-addictive design and various parental controls. Initial monetization will involve payments for extra generations. OpenAI emphasizes safety through consent features, automated moderation, and strict defaults for teen accounts.
Sora 2 could streamline filmmakers’ workflows in pre-visualization and concepting, while addressing concerns over licensing and content provenance. This innovative tool aims to reshape video production and social media engagement through AI-driven creativity.
Source link
