Jordan Shaw

Over the last few months, I’ve been feeling pretty nostalgic for the late 90’s and early 2000’s skate videos like Baker Skateboards 2G, Flip Skateboards Sorry, Girl Skateboards Mouse, P.J. Ladd's Wonderful, Horrible, Life, Alien Workshop Photosynthesis etc etc… It got me thinking and wanting to experiment with some Skate GANs and see what happens when you pump in ~25,000 videos stills from originally released on VHS videos and train a neural net. Not quite what I expected, and not quite the quality I was hoping for but these clips are enjoyable. I’m still experimenting with different data sets and editing down the footage to see if I can get cleaner and more “skate clips” style outputs. For now, I find these pretty fun and the audio definitely helps set the vibe from back in the day you’d watch newly released skate video with your crew. I’m happy that for now, we all still gotta skate and machine learning algorithms won’t be generating any new skate videos for us in the near future. Instead, we can count on them creating abstract imagery that vaguely resembles classic skate videos but in latent space.