When to implement ai modeling?
Will it charge in the future?
I haven’t found any features that are better than "Midjourney " and “Stable Diffusion” yet.
The goal isn’t to be better than them. But the good news is we are training the model over time, so it’ll always get better!
It’s to make sure what you generate is 100% copyright free and owned by you (at release obviously, during beta it’s another story). It’s also to create the other channels (normal, height etc) needed to make that texture useful in a game.
If you think about it, it makes sense, they have way more training sets than us as they don’t limitations on what is being generated, unlike what we want to achieve.
I tried Muse textures. They don’t provide me anything better than I can find in the asset store already.
Muse textures only work for very simple geometry, like a brick wall. Forget texturing a car or an entire house.
It’s nice that the results are 100% conflict free, but I don’t see where I would ever use them.
That’s okay. We are improving the model all the time.
Hopefully we can provide a tool that will be helping your use cases! For texturing a car and a house, we have a different tool that will do that in research/development!
Think about this being one tool out of many that are coming