Re:Sora: Creating video from text
by geor9e on Thursday February 15, 2024 @07:45PM
(#39391521)
It's possible that through sheer volume of training, the neural network essentially has a 3D engine going on, or at least picked up enough of the rules of light and shape and physics to look the same as unreal or unity
[ Parent
| on HN ]
Re:Sora: Creating video from text
by samsullivan on Thursday February 15, 2024 @11:04PM
(#39393043)
It would have to in order to produce the outputs, our brains have crazy physics engines though, F1 drivers can simulate an entire race in their heads.
Re:Sora: Creating video from text
by staticautomatic on Friday February 16, 2024 @09:50AM
(#39397772)
I wonder if they could theoretically race multiple people at once like chess masters.