Reboot time! Let's go back to the beginning: I've been interested - since the beginning of this little chat - in ONE thing. Accelerate modeling/animation process to the speed of drawing a sketch. Anything else - absolutely anything else - is unrelated. That includes self-driving cars, voice recognition, google translate, siri, and even blue brain project. Also I'm not interested in copying real objects (depth reconstruction - because you'll need to build the damn thing anyway) or stitching textures from 50 thousands of existing photos (because copyright), facial reconstruction (because model release form, model salary and makeup) but a general purpose solution applying to modeling any object from nothing. So far, two videos are related to the problem - smooth teddy demo I posted, and muscle routing system I posted, and that's it. Now what? In the context of this little chat, I'm only interested in that one specific application, and won't care about any other tech that uses deep learning/neural nets, until I need it (right now I don't). Problem at hand and tools for that, you know? I'm looking for a workflow. --- Basically, I don't get excited about technologies easily, don't care about "possibilities" (and consider imagining possibilities a waste of time, since a possibility is a thing you can't use right now), and I'm only interested in tools I can use right now for the problem i'm dealing with.