On Tuesday, OpenAI announced the release of Sora 2, an audio and video generator that succeeds last year’s Sora. Along with the new model, the company also launched a linked social app called Sora, where users can generate videos of themselves and their friends to share on a TikTok-style algorithmic feed. While the app is currently invite-only, OpenAI has shared impressive examples of videos created with the new Sora 2 model. The company says the model is better at following the laws of physics, making the videos more realistic. For example, in a video of a basketball player missing a shot, the ball will rebound off the backboard instead of spontaneously teleporting to the hoop.
The ‘Cameos’ Feature and Social Safety
The Sora app comes with a unique “cameos” feature that allows users to drop their own likeness into any Sora-generated scene. To use their own likeness, users must upload a one-time video and audio recording to verify their identity. This feature also allows users to share their cameos with friends, giving others permission to include their likeness in videos they generate. OpenAI says the social app, which is a “major AR-sized bet,” is the best way to experience the magic of Sora 2. The Sora iOS app is available to download now and will initially roll out in the U.S. and Canada. The company says it hopes to expand quickly to other countries.
The launch of a social platform with a feature like “cameos” will require significant user safety measures. OpenAI, which has struggled with similar issues in its other products, has included parental controls for the app that allow parents to override infinite scroll limits, turn off algorithmic personalization, and manage who can direct message their child. However, this type of access can easily be abused, as a user could still generate deceptive content that could be used to harm another person, a persistent problem with AI-generated video.
A Free and Experimental Launch
The Sora app is free at launch, which OpenAI says is “so people can freely explore its capabilities.” The company says that at launch, the only plan for monetization is to charge users to generate extra videos in times of high demand. To curate its algorithmic recommendations, OpenAI will consider a user’s Sora activity, their location (attained via their IP address), and their past post engagement and ChatGPT conversation history, though that can be turned off. This is a contrast to Meta, which announced just last week that it added a video feed called “Vibes” to its Meta AI app that is purely composed of AI-generated videos.
What The Author Thinks
The launch of a social app built around a powerful generative AI tool is a high-stakes gamble for OpenAI. While the “cameos” feature is innovative, it also creates significant risks related to disinformation and the abuse of personal likeness. By entering the social media market, OpenAI is taking on the same safety and moderation challenges that have plagued platforms like TikTok and Instagram, but with a new and more powerful tool for generating highly convincing, and potentially harmful, content. The success of the app will depend not just on the quality of its AI, but on the company’s ability to navigate the complex social and ethical issues that come with a platform where anyone can generate a realistic video of themselves and others.
Featured image credit: Heute
For more stories like it, click the +Follow button at the top of this page to follow us.