It's surprisingly hard to get AI models to watch and comment on actual video content, even if they are made for the task.
The developers of OpenFold3 have released an early version of the tool, which they hope will one day perform on par with DeepMind’s protein-structure model.
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback