OpenAI Launches Sora, a Social App for AI-Generated Videos: OpenAI is stepping into social media with a new app called Sora, designed for making and sharing AI-generated videos with friends. The free app, available starting Tuesday by invitation, is powered by an updated version of OpenAI’s video-making software, also called Sora.
Like the original Sora released in December, users can create short video clips from text prompts. But the new version lets people see videos made by others, making it more social and interactive.
Create Your Own AI Avatar and Voice
One of the standout features is the ability to make a realistic AI avatar and voice of yourself. You can then use this avatar in videos created by you or your friends, as long as you give permission.
A Tough Crowd
Even though ChatGPT has been a huge success, with over 700 million people using it weekly, OpenAI still hasn’t made Sora a household name. The company faces competition from big players like Google, Runway AI, and Midjourney, which also make AI video tools that can produce clips quickly and, in some cases, cheaply than starting from scratch.
Stepping into Social Media
With this new app, OpenAI is taking a big step toward creating a social media platform. This puts it in competition with TikTok and Meta Platforms, which recently introduced AI-powered video feeds like Vibes.
This move could open new revenue streams from advertising and also bring more attention to OpenAI’s technology.
“A lot of us, even internally, were kind of sceptical of the idea of an AI-generated feed,” said Thomas Dimson, a software engineer at OpenAI, during a briefing with journalists this week.
The cameo feature, which lets users insert themselves into videos, made staff think the feed could be “good for the world,” he said. “Over time, I’ve noticed that a lot of the other social platforms have actually drifted away from this idea of connections and friends.”
Improved AI Video Generation
The updated Sora is also designed to fix common problems with AI video generators. OpenAI says the new software is better at making videos that follow the laws of physics, something many AI video tools struggle with. It also aims to follow users’ prompts more reliably across multiple shots.
“We really think we have something which feels like it might be the ChatGPT moment for video generation,” said Bill Peebles, who leads OpenAI’s Sora team.
Peebles explained that the latest software can handle more complex visuals, like a person doing a backflip on a paddleboard in water, complete with realistic fluid dynamics and buoyancy. The app can also automatically put together individual scenes without requiring the user to manually edit them.
Audio and Multilingual Support
Sora can generate audio that matches the videos, including sound effects, background noise, and dialogue in multiple languages. These features could make it especially appealing in Hollywood, where OpenAI has been trying to gain traction with mixed results.
During a press demo, the company showed a news-style clip of Dimson drinking ketchup from McDonald’s dispensers. Another AI-generated video looked like a perfume ad for Sora 2 with the tagline: “The new fragrance from Sora – fresh, clean, unapologetic, for whoever you choose to be.”
Safety and Authenticity
As Sora’s capabilities grow, concerns about its impact on the filmmaking industry and online authenticity are likely to rise. OpenAI said it will not let people create videos featuring public figures unless those individuals choose to make a cameo. It also won’t allow videos based on someone’s image.
All videos made with Sora are watermarked so viewers know they’re AI-generated. The same applies to exported videos, and screen recording is disabled to control sharing.
Availability
The Sora app will launch on Apple’s iOS first, with plans to expand to Android later. The updated Sora software, Sora 2, will also be available on the sora.com website.
ALSO READ: iPhone Beats Android To Massive Google Home Upgrade
