On that topic there's so many things that are so simple and should have been discovered 50 years. Diffusion feels like it should have been discovered 30 years ago (adding noise to an image and solving for the reverse to generate new images of the same kind).
> adding noise to an image and solving for the reverse to generate new images of the same kind [should have been discovered 30 years ago]
Diffusion networks sre new, but denoising autoencoders work on the principle you outlined and were invented in the eighties.
The reason neural nets didn't become popular sooner is because achieving the sorts of results they produce nowadays requires a ton of computing power and a ton of data to train from.
