Friday 14 February 2020

mouthy hamster

Our programmer friend, author and AI-minder Janelle Shane (see previously) took a different approach to the holiday medium that arguably machine-learning could most easily access and influence—the sadly unavailable chalky candy-heart—explicitly not attempting to have her neural network try to caption them but instead only seeding the task with a list of the original (and impressively varied) three-hundred and sixty-six messages to one’s sweetheart and no other context. Here are just some of the results but be sure to visit the links above to see more and learn about the methodologies behind machine learning.