Sora is the next step to reaching AI for generating manufacturable things

[stepalibre]

The ability of an AI to understand physics well enough to generate videos can help in generating CAD models. Its understanding of context across many frames is what I think is needed for generating precise technical objects and real world things. I can't wait to try it!

Not using tech like Sora alone, but pairing it with a CAD engine, CNC machines and 3D printers.

This is huge!



This is essentially object tracking.

Sora: Creating video from text

openai.com

Video generation models as world simulators

We explore large-scale training of generative models on video data. Specifically, we train text-conditional diffusion models jointly on videos and images of variable durations, resolutions and aspect ratios. We leverage a transformer architecture that operates on spacetime patches of video and...

openai.com

 

[stepalibre]

 

[stepalibre]

Can you generate and/or use photogrammetry with this technology? If so, then you should be able to generate far more accurate 3D scenes/models and test them in the AI generated environment making even better 3D CAD to AI rendered image or video.

It's not the video it's the data you can extract from the video that's the bigger deal.