Observation: Experimental video made using Runway Gen-1 Video to Video

I took short video clips with my phone of two friends hovering around a third laying on the floor and Runway Gen-1was able to transform them into this. I added some dialogue and sounds and edited the clips together into this short video. You can see how the machine learning Runway Gen-1 extrapolated simple everyday items in the background and created the water, buildings, lights, harbor just from my text input. There was a chair that came out as a chair, but you could see the AI trying to fit the shape into a person.