What is strange is that last year I spent a lot of time creating the exact same 3D gallery experience with Claude. I picked up the style of the lights, the benches, the walls, the cards position. I also added music, lights on top of the images and the exact tour experience.

A year later, I see a very similar implementation!...I find that very odd.

But I do like the experience and site, there are few things I can pick from it myself for the project I've.

But it seems to me anything you use Claude for will be used aggressively to train the next model and some else in same prompting space can recreate it.

Not take the anything from the author's work, mine was in a slightly different context.