OpenAI, the developer of ChatGPT, said on the 30th (local time) it will launch a social media (SNS) app that lets users create and share videos generated by artificial intelligence (AI).
Named after the video-generation AI Sora, which OpenAI released in Feb. last year, the app lets users produce AI-generated videos and share them with friends. Like existing social media platforms, the Sora app provides an algorithmic feed tailored to users' interests and interactions.
Videos can be up to 10 seconds long, and created videos can be shared to the feed within the Sora app. It is similar in format to other short-form video feeds such as TikTok and Instagram Reels. Brian Nowak, an analyst at investment bank Morgan Stanley, said, "The newly launched Sora app directly competes with existing social media platforms such as Meta, Google, and TikTok."
The Sora app is based on OpenAI's latest video AI model, Sora 2, released the same day. In the app, users can create AI avatars and voices that look and sound like their real selves and place them in videos within the app. OpenAI said Sora 2 can precisely align images and audio in videos and was designed to reflect user prompts across multiple scenes more accurately.
The company said it can also generate complex visual scenes. For example, it can generate a person doing a backflip on a paddleboard while accurately reflecting fluid dynamics and buoyancy, and it automatically stitches individual scenes together so users don't need to edit them themselves.
Bill Peebles, Head of Team of OpenAI's Sora team, said, "This software could be 'a ChatGPT moment' for video generation."
Bloomberg News said, "With this app launch, OpenAI has taken a big step in building social media products," and added, "At the same time, it could open a new revenue stream through advertising and increase the technology's visibility."
OpenAI first released the Sora app on the iPhone operating system (iOS) and plans to roll out an Android version later.