Great questions Christopher!

I don't think anyone behind these models (Stable Diffusion, DALLE, Midjourney, Imagen, etc.) believes we're "surrendering ... creativity." Nor do they want we do.

The reason is that current AI models--all based on deep learning architectures--, don't have human-like creativity. Current AI can interpolate from the training set but can't create something new if it falls too far from what it has learned.

You may argue that humans are the same, but we don't simply learn from a highly constrained set of images. Our understanding of the world is multidimensional and so are our sources for creativity.

Could future AI develop these same features? We don't know, but in any case it feels very far in the future.

Re: how these systems can help us, I think that, like it happens with other technologies, most use cases will be uncovered with time. The low hanging fruit is: embedding AI models into existing tools to enhance the possibilities, replace stock imagery, and serve as inspiration for creative workers.

Will these systems affect the demand for creative jobs? Certainly. Will they replace all artists? No. Not now, not in the future.

Hope that helkps!