Pushing the boundaries of artificial intelligence in the technology world, OpenAI introduced the ability to create video from text with its new model called Sora. This model can take users’ text-based instructions and turn them into realistic and imaginative videos. This innovative feature offered by Sora heralds a new era in the world of art and design.
Sora can produce videos up to one minute long from texts, without compromising visual quality in the process. It can animate imagined scenes in detail by adhering to the instructions given by users. For example, she can bring to life various scenarios, from a stylish woman walking through the streets of Tokyo to giant woolly mammoths walking through a snow-covered square.
One of Sora’s most notable features is his ability to simulate the physical world and movement. This is part of OpenAI’s overall goal of empowering AI to understand and simulate real-world interactions. The model not only understands users’ requests, but also understands how these requests may have an impact in the physical world.
OpenAI points out some limitations regarding Sora
However, like every technology, Sora has some limitations. In particular, it may have difficulty simulating complex physical interactions or accurately modeling certain cause-effect relationships. They may also have difficulty with things like tracking a specific camera movement or blending spatial details like left and right in the scene.
OpenAI is taking significant steps to ensure Sora’s security. For example, it is developing tools to detect that videos produced by the model do not contain misleading content, and plans to include C2PA metadata if Sora is used in OpenAI products. It also applies existing security methods developed for DALL·E 3 to Sora.
Sora is currently being rolled out to assess harms or risks in critical areas, with access to “red team” members and creative professionals to provide feedback. This early sharing is important as OpenAI begins working with outsiders and gives the public a glimpse of what AI capabilities are on the horizon.
Sora has the potential to be a powerful tool in the hands of creative professionals and artists. This model offers the opportunity to create new forms of expression and enrich storytelling in the world of art and design. However, this technology needs to be used responsibly and be wary of possible misuse. OpenAI aims to collaborate with policymakers, educators, and artists around the world to understand the positive use cases of this new technology and create safer AI systems overall.
Source link: https://www.teknoblog.com/openai-sora-nedir/