Advertisement

OpenAI has developed an amazing text-to-video tool named Sora that can generate realistic video clips based on simple text prompts. However, there has been a question since the launch of this new tool – what data OpenAI used to train the model.

The CTO of OpenAI was asked In an interview if they used YouTube videos to train the model, and she could only say “I’m not sure about it.” Later on, the COO also refused to answer whether the model was trained on content from YouTube. Nonetheless, there are reports claiming that OpenAI has used YouTube videos to train Sora.

Google

In the latest news, Google’s CEO Sundar Pichai has responded to the matter saying he will sort it out if the allegations turn out to be true. According to a New York Times report, OpenAI used over a million hours of content from YouTube just to train Sora.

“Look, I think it’s a question for them to answer. I don’t have anything to add. We do have clear terms of service,” noted Sundar Pichai upon being questioned whether the matter is a violation of Google’s terms and conditions. “And so, you know, I think normally in these things we engage with companies and make sure they understand our terms of service. And we’ll sort it out,” he added.

Reportedly, The New York Times has already sued OpenAI for using their copyrighted content on AI training. However, he didn’t reveal how he plans to sort out the matter.

In an ideal scenario, a creator should know and own the right to opt in or out of the use of their content by someone else. The AI training requires a ton of data and it’s likely to be collected from the web, but not without permission. In fact, upon being questioned whether OpenAI used content from YouTube, the company’s COO hinted that alongside building a tool that can detect AI-generated images they are also working on developing “a content ID system for AI that lets creators understand as they create stuff where it’s going, who’s training on it, being able to opt in and out of training”.

Related:

Comments

RELATED ARTICLESMORE FROM AUTHOR