Well, I think this changes everything. What do you think? #3D #AI

What's the workflow? It's quite simple actually: @krea_ai just launched the open beta of their Live canvas. In here you can paint and transform that using AI, OR you can screencapture anything on your computer and transform that.. or even use your webcam. In the video, I use Blender for the basic models. Other tools are making similar systems, among them @freepik, who also does screen capture, and @LeonardoAi_, which, currently offers a paint to ai tool without the capture capabilities. There's also several Real-time Latent consistency spaces on Hugginface that does this, among them this one that I used previously: After I found an image I liked in Krea, I sent that to @Magnific_AI for the refined images. Krea also has a built in upscale/enhance model but I find that Magnific offers a better model, although Krea has the benefit of being free currently. Here's the upscaled/enhanced versions. Imagine once we're able to have full coherency on models like this..

the amount of tools we have going into 2024, it's like we just graduated with a 3d Max / Mya degree from a 5 year college.

Scary stuff

Haha 😂

I think we need a refine complete scene back to 3D button in the ui 😉😁

I agree and we probably will have quite soon!

maybe it would change it if you focused on tools that assist artists instead of doing everything for them. Like you know areas like unwrapping, weight paints, or retopology could greatly benefit from this if you focused on that.

Blender default cube is changed forever

I think metaverse will be our new reality soon.

This is a new kind of renderer. It's offline-ish one now, but we will see a 3D game rendered this way soon enough as well.