Today I discovered a new ability of artificial intelligence

Today I discovered a new ability of artificial intelligence — to alter real photographs while preserving their authentic qualities, without turning them into cartoons or obviously generated images. Even for me, someone deeply experienced in AI imagery, this feels revolutionary: it allows changing the viewpoint of an existing photo, slightly shifting the subject or even seeing it from above, without altering its story, background, or depth of field. — It’s a resource-heavy process — generating one photo (from several angles) takes about 15–17 minutes if no other programs are running. I’m using relatively small images (1200×800 px), and my PC isn’t weak (RTX 4090, 64 GB RAM, i9-13 900, fast NVMe). Google Gemini can do something similar much faster, but the result is low-res (1280×800) and looks a bit like cloudy plastic. — I’m curious—do you think this changes how we perceive photography?