OpenAI’s Sora is coming, and the promise is UNREAL
Building Startups by Ajay Yadav
The ‘No-BS’ BS Newsletter
OpenAI’s Sora is coming, and the promise is UNREAL
Hello and welcome back to the Building Startups Newsletter!👋
Sam Altman took X by storm last night by introducing Sora, OpenAI’s text-to-video generator. 📹️
Keep reading to be blown away by a few of Sora’s videos.
Everything You Need to Know About Sora
✅ Sora is a text-to-video model. “It is able to generate complex scenes with multiple characters, specific types of motion, and accurate details of the subject and background," OpenAI’s statement says, adding that it can create multiple shots within a single video.📼
✅ We’ve recently seen similar text-to-video generators like Pika and Runway ML, and while both of those tools are impressive as well, I personally think that Sora will have an edge with the quality of content it is generating.
✅ Sora can generate up to a minute long videos while maintaining visual quality and adherence to the user’s prompt. :
✅ The model is currently only available to red teamers (domain experts in areas like misinformation, hateful content, and bias) to assess important areas for harms and risks, and a few visual artists/designers/filmmakers as well to gain feedback on how to improve the model’s efficacy.
✅ OpenAI has also acknowledged that the current model has weaknesses. Sam and his team have ensured that that they will be undertaking several important safety steps ahead of making Sora available in OpenAI’s products:
We’ll be taking several important safety steps ahead of making Sora available in OpenAI’s products. We are working with red teamers — domain experts in areas like misinformation, hateful content, and bias — who will be adversarially testing the model.
We’re also building tools to help detect misleading content such as a detection classifier that can tell when a video was generated by Sora. We plan to include C2PA metadata in the future if we deploy the model in an OpenAI product.
Introducing Sora by OpenAI
Sora in Action
Adding to the hype Altman created with Sora’s announcement, he also asked people on X to reply with absolute random video prompts they’d want Sora to execute on.🤖
Soon there were thousands of replies to his tweets with prompts people wanted to see executed on Sora. The results?
Immaculate🫰🏼✨
SEE FOR YOURSELF if you haven’t already⬇️
Sora’s creation for a prompt by Kunal Shah
Notice how the video also gets the text right and how the shots change, all through a single text prompt👇️
Remember how we’ve struggled with hands in AI generations? Not anymore, it seems👇️
It’s very exciting to think of how the video creation landscape will change post the commercial release of Sora.
But another key takeaway is the increasing importance OpenAI, and AI companies overall, are placing on safety and avoiding misuse of generative platforms.🛟
Are you excited to try Sora out? I can’t wait.
That’s all for today’s newsletter, see you in the next one. Stay curious 🧠
Be kind, share this newsletter with friends through the link below :)