@haitchfive on Nostr: The implication of GameNGen is that, if a difussion model can learn to be a Doom ...

The implication of GameNGen is that, if a difussion model can learn to be a Doom engine, it can learn to behave and present like any other type of graphically interactive software.

Based on this method, it's not impossible to think that one could create a videogame without any programming at all, but providing all the necessary input data, imagery, and animated behaviour for a model to learn.

Mind you, with enough patience, you could draw it all on a notebook.

#GenNGame #ai #ml #difussion #DifussionModel #gamedev #indiedev #videogame #videogames