16-2-2024 (SAN FRANCISCO) Microsoft-backed OpenAI has announced the development of Sora, a cutting-edge software designed to generate minute-long videos based on text prompts. The company disclosed this breakthrough on Thursday, showcasing Sora’s potential applications in red teaming for AI system flaw identification, as well as providing a creative platform for visual artists, designers, and filmmakers seeking feedback on the model.
According to a statement from OpenAI, Sora has the capability to generate intricate scenes featuring multiple characters, specific types of motion, and accurate details of both subjects and backgrounds. The software can produce multiple shots within a single video, adding a dynamic dimension to its video generation capabilities. Notably, Sora is not limited to video creation; it can also animate still images, showcasing its versatility.
This development comes on the heels of OpenAI’s successful launch of the ChatGPT chatbot in late 2022, which garnered attention for its proficiency in composing emails, writing code, and even crafting poetry. The emergence of Sora marks another stride in OpenAI’s mission to push the boundaries of generative AI technology.
Meta Platforms, the parent company of social media giant Facebook, has also delved into AI-powered image generation with its model Emu. In a bid to rival tech giants like Microsoft, Google, and Amazon in the rapidly evolving generative AI landscape, Meta enhanced Emu last year by introducing two AI-based features capable of editing and generating videos based on text prompts.
While Sora represents a significant advancement, OpenAI acknowledges that it is a work-in-progress. The company cautions that the model may occasionally struggle with spatial details in a prompt and encounter challenges in following specific camera trajectories. OpenAI is actively working on refining Sora and is concurrently developing tools to discern whether a video was generated by the software.
