I’m impressed by this; it seems to lift the image generation to the level where it might actually be useful - in much of the same way Chat-GPT has for text LLMs.
This has real promise in the rapid generation of personalised infographics for patients, and for me personally, icons and branding for online forums.
I can also see it starting to make life really uncomfortable for those making a living from graphic design and the like.
I’m going to register for early access, and will let you all know how I get on.