What will really be a game changer is AI that can generate the bulk structure having been trained on our best of the best hand built models and we can iterate on the emotional details.
I’m working on the “bulk structure” part, training models to generate random game worlds that roughly adhere to the rules, look and feel like everyone favorites right now (while avoiding copyright issues).
After that my goal is empowering consumers directly to nudge the styles in their preferred direction.
I’m mostly motivated by the MBA-ifying of everything. My goal now is to just have AI produce new content for me, even if such a thing puts game developers out of a job. I’m starting to experiment with cartoons as well. Optimizing for myself like we do.
I still want to comment because your comment doesn't do that demo justice.
It's the first time I expierienced that awkward feeling looking into reality while watching the video of that city.
We are Eye opening close to cross over uncanny valley after reaching it that soon.
The next step change will have to be some form of advanced kinematic rigging that creates advanced motion granularity akin to the graphical fidelity of nanite graphical tech.
UE5 is a step in that direction, but there is still lots of work to be had.