While generative AI models continue to advance, OpenAI has also officially entered the field of social platforms and announced the launch ofNew short video platform "Sora", with its latest Sora 2 model as its core, allows users to directly generate audio and video content, and is described by the outside world as the AI version of TikTok.
Unlike traditional short video platforms that rely on users to shoot or upload videos themselves, Sora's content is almost entirely automatically generated by AI. Its interface design is similar to TikTok, using vertical screens and sliding browsing. However, the video length is currently limited to within 10 seconds. Whether it will be gradually extended in the future like TikTok cannot be confirmed at the moment.
AI-generated portraits and interactive mechanisms
Notably, Sora 2 not only generates images but also synchronizes audio, allowing users to create more complete virtual avatars. OpenAI has also implemented an identity verification feature. If a user opts in, Sora 2 can generate videos using their likeness. The platform automatically sends notifications when others use their likeness through "cameos" or "remixes," even if the content hasn't been publicly released. This enhances transparency and security regarding the use of likenesses.
Copyright and content restrictions still need to be clarified
Since the production of the film involves a huge source of materials, Sora is bound to face copyright-related challenges.Wired website reportThe platform will reject some restricted generated themes, but the strictness of the regulations is still unclear.wall street journalIt points out that content copyright holders may need to "opt out" to prevent the material from being used for generative training by the Sora 2 model, which also means that copyright management will be a key issue that OpenAI must deal with next.
Meta also launched a similar service "Vibes"
Meta recently announced the launch ofA new short video app called Vibes, all platform content is automatically generated by AI, and it was first launched in North America, Australia, New Zealand, the United Kingdom and some European markets. This means that Meta not only continues to deepen its presence on social platforms, but also hopes to open up a new stage for AI technology through Vibes.
Why enter the social field?
It's widely believed that OpenAI's launch of Sora not only demonstrates the multimodal capabilities of the Sora 2 model, but also likely reflects its focus on the market opportunities presented by the US government's fluctuating stance on TikTok. If TikTok faces stricter restrictions in the US, Sora could capitalize on this opportunity and attract existing short video users to AI-generated content.
On the other hand, adding social attributes to AI models will obviously help expand user reach and further bring more people into the OpenAI ecosystem.
Invitation-only, only available on iOS platforms in the US and Canada
Currently, Sora is only available in the United States and Canada, and is initially available only on iOS. This is an invitation-only platform, meaning users who want to try it out will have to wait for official releases to gradually expand. This will limit its spread in the short term, but it will also allow OpenAI to observe user behavior, test content moderation, and platform operations before gradually expanding to a wider market.
summary
The launch of Sora marks OpenAI's move beyond simply providing AI tools and its direct entry into the social entertainment sector. In the fiercely competitive short video market dominated by TikTok, Instagram Reels, and YouTube Shorts, Sora's differentiated approach, centered around AI generation, may open up new opportunities. However, its ability to balance creative freedom, content security, and copyright protection will be crucial to its success.
