Looks really cool. Definitely nitpicking, but their assertion that we are in "the metaverse era" feels out of place and personally irritates me as it reinforces a farcical hype cycle from the software industry that I hadn't expected was alive and well in academia.

Having grown up reading Snow Crash at least 10 times, I kind of ignore the recent hype. When I hear "metaverse", I still think Stephenson, not Zuckerberg.

In principle you could expand the concept into a game engine and have an AI which imagines and learns from a simulation that it also builds.

It's easy to feel hyped when stuff is getting so promising of something cool.

they should add blockchain based NFT-asset management in it and complete the trifecta of SV hype-cycle.

Pass the "Wahoo" moment, I kind of see the problem with this 3D-modeling-from-a sentence : - You have to take the given 3D model without a possible raffinement on important things like : - the form/geometry, - the texture style, - ect...

> You have to take the given 3D model without a possible raffinement on important things like : - the form/geometry, - the texture style, - ect...

Why?

Have a look at the meshGPT paper