One of the primary innovations in the software side of 3D ecosystem is the use of Artificial Intelligence (AI) to be able to create 3D media from 2D media. The core 3D fundamentals are already in place where 3D artists are already using the traditional 3D design tools to design 3D content using 3D software pipeline. 3D animation artists working on the 3D software pipeline falls in this legacy 3D software systems. Softwares such as 3d studio max, Maya, Cinema4D being some examples of the traditional 3D tools used by artists to create 3D models and animations.
The main issue with the legacy 3D software pipeline is, it is a customized process to create 3D content for each individual tasks. Consider a 3D game design as an example.
The AI brings the scale in the 3D software ecosystem to be able to create 3D assets and environments at mass level in a quick way which was never possible in the legacy 3D software design systems. 3D creation using AI is still in its early days though many companies claim from the rooftop how they create it using AI, because only a small part of the pipeline of 2D to 3D is still automated while a large part such as texturing etc. is still done manually when it comes to photorealistic 2d to 3D conversion (not approximate conversion like in 2D portrait to 3D cartoonish avatars using photos, as an example), as its a hard problem and will take at least a few years to advance to a level where it can be fully automated without any manual intervention.
However, its getting there, one step at a time.