In Cringe Video, OpenAI CTO Says She Doesn’t Know Where Sora’s Training Data Came From
Wondering what data OpenAI used to train its buzzy new text-to-video AI? OpenAI CTO Mira Murati seems to be wondering, too.
NGL between the title and language of this article and the title of the related articles I saw at the bottom of the page, futurism.com is cringe as fuck. It's bad that she didn't answer and the way she gave her non-answer was bad, but I don't need a website to interpret the video for me.
The original source at the WSJ is much cleaner, and whatever else you might think about their site/content/agenda, at least it was a pretty squeaky description of what's going on in the video:
Sora, OpenAI’s new text-to-video AI model, can create realistic scenes. In an exclusive interview, WSJ’s Joanna Stern sat down with the company’s CTO, Mira Murati, who explained how it works but ducked questions about how the model was trained.
What an annoying headline lol
I once had a serious fight with a friend about what constitutes public data.
She looked one of her medical providers up on social media, and found out he was into bdsm or some such, and it upset her.
I felt it was wrong what she had done. It's his private life and has nothing to do with his work including his work with her. Whereas she saw it as public data that is totally OK to access for anyone.
Just because I post something here doesn't mean it's socially acceptable to look me up. And if you do at least hide it because it's creepy af.
i think if you post it online on anything other than a DM, you are consenting for it to be public. even DMs are barely private.
you dont want people to know, don't post it. its even in most platforms TOS....
Agreed, people get surprised their public data is public, I pretty damn well know my profile pic could end up on everyone's phone the moment I set it as profile pic
