I don't know if anyone else has this, when you have prompts that work and you like them you want to experiment but if you vary anything the results will be different than they would be otherwise, I wouldn't have expected there to be a kind of fear of missing out in the back of my mind in regard to Ai but there is.
I also noticed some of my earlier work is better than what I am doing now and I think this is because one can try too hard to get particular results rather than just enjoying the unexpected.
I'm not sure it's FOMO for me, but I will often repeat the same prompt over and over trying to get that "perfect" image because I've gotten as close as I'm willing to get to the filters. It's kinda funny... sometimes I'll be getting 1 or even 0 results from a "near-perfect" prompt, knowing I'm sitting right at the edge of a filter, but then I'll tweak the prompt to unintentionally use a word or phrase that has gotten me blocks before... and I'm suddenly getting all 4 images!
I do agree about how trying to control every aspect often lowers the output quality, when more generic prompts from early on were leading to amazing, unexpected results. I've started going back through my early attempts at AI and running them through Krea. The early stuff was much less filtered (especially on sites like Nightcafe), even if the image quality wasn't good at all.
While I'm not particularly impressed with the results when I run a current favorite image through Krea, I've been really impressed at how it "remasters" the nightmare faces and not-quite-right anatomy of the early stuff.
When I started using AI to make images I was using a generator called pixlr which I thought was great at first but I came to realize was pretty lousy with a high rate of deformities, like missing arms, extra heads or three legs. And I had to pay to use it, to boot. But I've been able to salvage of the pixlr images with Krea, and the results are pretty good.
Love you, too
