

I’ve been impressed by how quickly gen-AI 3D models have become somewhat useful. I expected that with the significantly lower training data than text and images it would take longer, but maybe 3D files have a lot more tagging and metadata by virtue of being “new” compared to text and images, making them more useful for training AI?
As much as I enjoy modeling, I must admit I’m somewhat excited about a future where I can think up something to 3D print and not have to spend hours in Blender.
There was a “don’t use Amazon for a week” boycott thing at the beginning of March. It shocked me how quickly I lost the urge to just go order stuff the moment I thought of it. After the week was over I just kept not ordering stuff (with the exception of a subscription that came through).
Definitely at least helped me cut back.