OpenAI's latest text-to-video generative AI model, Sora, has captured global attention with its ability to produce lifelike videos. Mira Murati, the Chief Technology Officer (CTO) of OpenAI, confirmed in an interview with The Wall Street Journal that Sora will be accessible to the general public "this year," with a potential release in just a few months. The announcement comes after OpenAI unveiled Sora in February, showcasing its capabilities through a series of generated videos, marking a significant advancement in generative AI technology.

Presently, access to Sora is limited, but according to Murati, this is soon to change as the platform gears up for broader availability. Additionally, Murati revealed plans for Sora to incorporate audio generation in the future, enabling users to create even more compelling content directly through the AI interface. OpenAI aims to enhance the tool further, possibly allowing users to edit AI-generated videos, though Murati noted that video generation requires substantial computational resources, indicating a pricing model akin to DALL-E, OpenAI's text-to-image generator.

Regarding the data used to train Sora, Murati refrained from divulging specifics but mentioned that it primarily relied on publicly available or licensed data, including content from OpenAI's partner, Shutterstock. Notably, videos generated by Sora will feature a watermark at the bottom, distinguishing them from human-created content, a measure aimed at addressing concerns surrounding misinformation. With several countries enacting stringent regulations to combat misinformation, such as restricting election-related queries on platforms like Gemini AI, the inclusion of such features underscores OpenAI's commitment to ethical AI development. Stay tuned as OpenAI's Sora prepares to revolutionize content creation, promising a new era of AI-assisted video generation for users worldwide.