© IE Online Media Services Pvt Ltd
Tags:
Journalism of Courage
OpenAI stunned the world with its latest text-to-video generative AI model, Sora, which could generate lifelike videos. The company’s Chief Technology Officer (CTO), Mira Murati, in an interaction with The Wall Street Journal, has now confirmed that Sora will be available for the general public “this year” and it “could be a few months.”
Back in February, OpenAI introduced Sora to the world by posting a few videos generated using the platform, making a major stride in the world of generative AI. Currently, only a limited number of users have access to Sora, which, according to the CTO of OpenAI, will soon change.
Murati has also confirmed that Sora will “eventually” incorporate audio generation, allowing users to generate more compelling content straight from the AI bot. The report also highlights that the company is looking at further improving the tool and might even allow users to edit an AI-generated video. It was confirmed that generating a video is ‘much more expensive’ in terms of computational capability, and the company is planning to price Sora similarly to DALL-E, OpenAI’s text-to-image generator, which used to cost $15 for 115 credits post the initial 50 free credits in the first month and 15 free credits every subsequent month. OpenAI currently offers DALL-E 3 access to ChatGPT Plus subscribers, which costs $20 a month.
When asked about the data used to train the model, the CTO said, “I’m not going to go into the details of the data that was used, but it was publicly available or licensed data.” She also confirmed that OpenAI has used data from its partner Shutterstock. Videos generated using Sora will have a watermark at the bottom, which helps users distinguish between human-created and AI-generated content. Several countries have started to impose strong laws to prevent the spread of misinformation, which includes the restriction of election-related queries on Gemini AI, and more.
