On October 9th, Bill Peebles, head of OpenAI Sora, announced on the X platform that the app, dubbed the "AI version of TikTok," had surpassed one million downloads in less than five days, a growth rate even faster than the company's phenomenal product, ChatGPT. Notably, this achievement was achieved while Sora was still invite-only and available only to users in the United States and Canada. Currently, the app has topped the Apple App Store's free app charts, demonstrating its impressive market potential.
Sora, released at the end of September, features a core feature that allows users to generate new video content using real-life video clips of themselves or their friends using AI prompts. Users first upload a short, real-life video, called a "cameo," which is then used by AI to create content. The app also features social sharing, allowing users to share their "cameo" with a specific group or all Sora users, authorizing others to use their image to generate videos. This unique interactive model has quickly attracted a large number of users.
Technically, Sora leverages OpenAI's powerful Sora 2 video generation model, capable of processing complex dynamic images and supporting a variety of styles, including realistic, cinematic, and anime. It also features synchronized audio generation. OpenAI emphasizes its social nature in product design, positioning it as a tool for interaction and creation among friends, which is why it adopts an invitation-only system. The app's feed dynamically adjusts content based on the users' following, and regularly collects user feedback to continuously optimize the experience.
