When a user enters a textual query into Midjourney’s text-to-image generator, the AI searches its “training images” for a certain set of results. Then, these photos can be expanded or regenerated. It is a very helpful tool for exploring architectural concepts, particularly early on in a project or later on when examining particular aspects of it. However, creating a solid prompt can be a laborious process that calls for numerous iterations. You may learn some tips for using AI to locate relevant concepts an
Source: Midjourney for Architects: from prompt to visualization – Evermotion