OpenAI, known for its AI language model ChatGPT, has ventured into the realm of video generation with the introduction of its latest creation, the Sora AI model.

Following the viral success of ChatGPT, OpenAI aims to revolutionise video content creation using artificial intelligence technology.

Sora, unveiled by the company on Thursday, operates similarly to OpenAI’s image-generation AI tool, DALL-E. Users can input a desired scene or provide still images, and Sora will produce high-definition video clips accordingly.

The model can extend existing videos or fill in missing frames, showcasing its versatility in video content generation.

Significant expansion

The move into video marks a significant expansion for generative AI, following the success of chatbots and image generators in various consumer and business applications.

However, concerns regarding misinformation have escalated, particularly with the rise of AI-generated deepfakes, which have seen a 900% increase year over year, according to data from Clarity, a machine learning firm.

Competing with tech giants like Meta and Google, who recently announced their Lumiere video-generation AI tool, OpenAI is positioning Sora to be a leading contender in the market.

Startups like Stability AI and Amazon have also entered this space with their own video-generation models.

Video length

Currently, Sora is capable of generating videos up to one minute in length.

OpenAI, backed by Microsoft, aims to achieve multimodality by combining text, image, and video generation within its suite of AI models.

Brad Lightcap, OpenAI’s COO, emphasized the importance of multimodality, stating, “The world is multimodal… the world is much bigger than text.” He highlighted the need for AI models to encompass various modalities to better reflect human perception and interaction with the world.