this s*** is highkey scary, like mkbhd said it's just been a year since the Will Smith spaghetti video

It’s time for us all to die

Enjoy life now before simple reality and truth get molested and controlled by manipulative forces

nah man we've gone too far

In before this thread is moved to G&G

Need a central AI/ OpenAI advancements thread

This technology being available to the laymen means the government have had this for years and are way beyond this already. Stay woke

This is so impressive considering it was designed with the help of GPT-5 and this is the worst it will ever look. The real leap will be having all of this run and interact with user inputs in real-time.

onlyfans model might be out of a job if they ever release a version of this that allows nsfw prompt

It's so over

Youd be overestimating the governments

ehh i doubt that, they really dont need this to push narratives

wtf wonder what else they got under the hood

need this for p*** asap

Underrated things from the technical paper [Link]

• Sora can animate any current image OR video you have with prompt. Not just text prompts.
• Sora can make a video inferring happened before or after a given image/video using a prompt. It understands what is in the image and animates it appropriately.
• Sora can stitch two separate videos together. Sora will "dream" up everything inbetween based on your prompt.
• Sora can add special effects and new sceneries to existing videos by text prompt or blending videos together. (pic below)
• Sora can generate videos where the camera moves around, it understands that objects have a dimension to them that needs to be preserved.
• Sora can create animated images within images (tweet). You could prompt every screen to show a different video with a specific story of your liking.

Imagine the feedback loop when you blend AI videos with other AI videos. I'm thinking about the implications of what this would mean if they manage to do it in real-time some day. Video games would be done. Special effects companies would be done. All you need is to sit down and imagine a prompt and it will create a world for you.

it's all coming to life

YES! Remember those visualizers in QuickTime/Windows Media Player? I want that with Sora. Let AI color in every song I'm listening to, the lyrics and the melody could be the prompt.

Early

Im glad I got job security before AI could replace me

This is my life mission, to make the worlds best music visualizer

Jail time

That is a great idea. I’d love to see a beefed up visualizer for modern times
Sort of like that: Jitter can be used in Max Msp to generate visuals from audio