To me it looks very analogous to AI-generated "art", it's very easy to generate some generally esthetically pleasing visuals, but the depth of the art stays in proportion with the input effort... Which is often not much.
All of this shouldn't be very surprising really, and there's still a lot of usefulness to it, if only for depreciating the low-quality copy-paste productions and making the really unique and novel ones even more valuable.
Yeah, I couldn't say, it's just "vibes" I guess, just like the filler text produced by business consultants it's just not something that I feel would be missed if lost. It all looks the same at some level even when it's superficially different. Midjourney especially is very uniform in this regard, everything looks great, but it's kind of flat at the same time.
