OpenAI is back with yet another surprise in the form of Sora- a new text-to-video generator. The San Francisco company unveiled Sora recently; a tool that makes use of generative artificial intelligence to create short videos on the basis of written commands with the help of generative artificial intelligence.
Well, there have been many video-making tools so far, but Sora somewhere stands apart due to its high-quality videos. The introduction of Sora proves to be a significant step not only for OpenAI but for the future of video generation through text.
Understanding SORA
Generative AI stands as an AI branch that is capable of creating something new. Chatbots like ChatGPT and image generators like DALL-E are some of its examples. An AI system that is capable of generating videos is not only novel but also challenging. However, it is based on the same technology.
Sora stands as an AI text-to-video generator tool that creates videos. These videos can be up to 60 seconds long on the basis of written prompts by means of generative AI. The tool is also capable of generating video from a still image.
As of now, Sora is not meant for the public. As per OpenAI, it is still engaging with artists and policymakers before releasing the tool officially to the public.
Sam Altman, the CEO of OpenAI took to the platform X, formerly Twitter, to ask the users to share interesting prompt ideas. Later, the CEO shared realistic-looking videos. One of the videos created by a prompt was that of a bicycle race on the ocean, and different animals riding the bicycles, shot with a drone camera view.
Potential Weaknesses
Sora is capable of generating incredible and complex scenes in the form of videos. However, as per OpenAI, there still exist some weaknesses. These include loopholes in some cause-and-effect elements and some spatial elements.
Moreover, there also exist some potential risks related to its social and ethical implications. On this, OpenAI stated that it is taking essential safety steps prior to making Sora available widely.
Comments
All Comments (0)
Join the conversation