2 ms·
It's worth noting that when photography was new it was viewed much like generative AI is today - as not "real art", "cheating", or somehow otherwise less valuab
by zettabomb 3y ago
It's worth noting that when photography was new it was viewed much like generative AI is today - as not "real art", "cheating", or somehow otherwise less valuable than traditional art forms. Today, after such impactful photos as Afghan Girl (Steve McCurry), Raising the Flag over Iwo Jima (Joe Rosenthal), or Monolith (Ansel Adams), I don't think anyone could successfully argue that photography a lesser art form. I don't know what that means for generative AI but it sure seems to be a similar situation. The old guard feels threatened and tries to invalidate the new techniques, but when those people are gone will it be gradually accepted alongside everything else?
- viraptor 3y agoWhile this may change in the future, there's an aspect that many aren't really aware of - the good AI generated content takes time and skill. If you hang out in some stable diffusion communities, you'll see people iterating and tweaking elements for days to get things just right. There seems to be that impression that you can create art-level results easily. In practice, if someone has a specific idea, getting the right things in multiple steps out of current models is an art skill on its own.