OpenAI unveiled its groundbreaking text-to-video tool, Sora, last month that can generate realistic-looking 1080p videos. Currently, it is only available to a handful of filmmakers and creators who are testing it to find out vulnerabilities and fixing them, ensuring it is safe before they roll it out to the public. In an interview with the Wall Street Journal, OpenAI’s chief technology officer Mira Murati informed that they are aiming to launch Sora to the public before 2025 and that it could even happen within “a few months”.

OpenAI’s aim for Sora is serving creators with a tool that they can “edit and create with”

Image: OpenAI

Unlike the video showing Will Smith eating spaghetti that was unveiled just one year ago, Sora generated content in “hyperrealistic” (except for some weird representation of hands and fingers) and OpenAI CTO says they won’t serve this tool to the public before they ensure it’s safe to do that. Notably, Sora-generated videos will also be watermarked just like many other text-to-image generation tools.

When asked about how this tool might affect the jobs of creators in the future, CTO Murati said that their aim is to serve this text-to-video generation model as a tool to help creators in their work instead of replacing them.

“You know, the way that I see it is this is a tool for extending creativity and we want people in the film industry, creators everywhere, to be a part of informing how we develop it further and also how we deploy it. And also, you know, what are the economics around using this model when people are contributing data and such.”

Speaking of data, when asked about what data they have used to train Sora, she didn’t go much further but just said it is “publicly available and licensed data.” It could be videos from YouTube, Facebook, Instagram, and similar platforms.

Notably, Sora doesn’t include audio in the generated clips and OpenAI reportedly has plans to integrate audio in the future. Although, it might not be Sora but an even better version of it with a different name. Just like DALL-E, you will need to pay to be able to use the model and OpenAI CTO Murati said that the aim is to

Meanwhile if you are wondering how to use Sora AI, know that the access is still limited.

Related: