There are a *functionally* infinite series of images possible from that latent space. And yet not a single one of them are one you would have made without it. (it is not actually infinite but you are never going to see the end)

I also don't think this is isolated to just genAI, a lot of corporate art starts to devolve into everything-but-nothingness. But lets take a diffusion model for instance. The 'model' in question being the space of all possible images given the dataset. Not all actually possible images.

I don't mean to say that all art requires great effort - certainly some great art required very little (I'm a Duchamp fan) - but there was some Process that happened

i have been having this day for like 5 days now

ooh gdc, i went like 4 years in a row before covid, had a good time. its so expensive for me to get out there. cool city to visit though!

i'm gonna try!! hi !!

REALLY? omg :') ... well im gonna make some that are EVEN BETTER . hopefully