Purpose made text to video models are already pretty much perfect but there are no open source ones right now.

This is like typing on a keyboard one handed compared to the efficiency that near future models will allow.