Recent Thoughts on 3D AI
Given how quickly generative models are improving, I sometimes wonder how long standalone 3D AI products can remain relevant.
Still, problems in production workflows don't disappear as easily as we might expect.
Low-poly generation is still frustrating.
I'm increasingly interested in the combination of agents and geometry tools.
The process of users generating again and again is becoming a more interesting problem to me than model quality itself.
3D printing might actually offer a clearer value proposition for AI-generated 3D.
Aligning the research roadmap with the product roadmap is still difficult.