1 Like
If anyone wants to play with these visual generative models, then Draw Things is by far the best all-in-one app[1] available for phones, tablets, or desktops (Apple Silicon only, uses Stable Diffusion). Lots of very cool features (like controlnet poses and many more), and the UI lets you explore this utterly vast visual landscape with ease.
There are commercial online systems like midjourney and bluewillow, but being able to run the models directly on your machine is a big plus.
With some smart prompting, you can get totally amazing images out of drawthings. Personally I am more impressed with visual models than language models (which have been a hot topic on the forums), and as a visual neuroscientist it is really fascinating how visual knowledge is synthesised so comprehensively. I’m particularly fascinated by photographic portraits, we are way past uncanny valley and it is hard not to imagine these people existing somewhere in the world…
[1] there are more complex tools like invokeAI, A111, imaginAIry but these usually require more geek skills to install and use.
3 Likes
Hi nontroppo,
The images on drawthings.ai have triggered a range of possible images related to climate change that I am a little bit hesitant to actually have rendered.
Despite the trepidation, thank you very much for the reference.
Thank you!
scrive
A highly detailed landscape painting of a wizened Scrivener called scrive with a long white beard, using a MacBook and contemplating the environment and climate change, in the style of Caspar David Friedrich
1 Like
