I find Stable Diffusion and related LAION projects to be extremely interesting, less so for their actual technical features (the wonder of image generators, while still extremely cool, has worn...
I find Stable Diffusion and related LAION projects to be extremely interesting, less so for their actual technical features (the wonder of image generators, while still extremely cool, has worn off a little bit) but more so for being fully open source and modifiable.
Stable Diffusion in particular has experienced an explosion of development and applications in just a week (well, more than a week now): with image editor plugins, animation support, resource optimizations, advancements to draw-the-rest-of-the-owl, etc, etc. I think these modifications are incredible for just one week.
And to be entirely honest: it's reassuring to me that state-of-the-art AI does not lie entirely at the beck and whim of a small group of Silicon Valley techbros.
I find Stable Diffusion and related LAION projects to be extremely interesting, less so for their actual technical features (the wonder of image generators, while still extremely cool, has worn off a little bit) but more so for being fully open source and modifiable.
Stable Diffusion in particular has experienced an explosion of development and applications in just a week (well, more than a week now): with image editor plugins, animation support, resource optimizations, advancements to draw-the-rest-of-the-owl, etc, etc. I think these modifications are incredible for just one week.
And to be entirely honest: it's reassuring to me that state-of-the-art AI does not lie entirely at the beck and whim of a small group of Silicon Valley techbros.