NIGHTCAFE AI, BELLMER AND THE RADIATOR LADIES

“What is at stake here is a totally new unity of form, meaning and feeling: language-images that cannot simply be thought up or written up … They constitute new, multifaceted objects, resembling polyplanes made of mirrors … As if the illogical was relaxation, as if laughter was permitted while thinking, as if error was a way and chance, a proof of eternity.”
― Hans BellmerNow if Bellmer would have known AI - what would he have had to say about it? I think the above quote may have been his (the visual artists) answer to Andre Breton, the poet, who proclaimed that only poets could be true surrealists.

---

Here are some of the images I filed into my INSANE ASYLUM and the DOLLHOUSE FREAK SHOW collection. They go back to prompts I wrote some 8 months ago, most of them done with the SD 1.5 and 1.4 models but also with the now defunct SD 2.0 and 2.1 models.
The gist of this, and what it is all about, you will find on the bottom of my post.

^{Here we take a short trip to the Insane Asylum, a related post.... and back again to the DOLLHOUSE FREAK SHOW}

---

INSANE STUFF ABOUT AI

^{this above is the 'safe' version, after fiddling with the prompt.}

---

From a "bad" chain of images, the original one was flagged by moderation, even though I had labelled it 'unsafe'.

I now added a additional leading prompt, but the difference is all in my tweaking of the negative prompt.
My prompts are open, so you can check them out. What is missing however is the artistic gutsiness of the original: no trace of Bellmer, Lynch or Giger. Just overall 'Nightcafe cuteness' pablum fare.

While I published the original on Flickr, this one (perfectly safe for Kindergarten, which Nightcafe has become) will remain within Nightcafe, and possibly posted on Facebook and X:

as confirmation of the criticism
that AI is not being considered art by many of the artists community worldwide
About this original on Flickr, called BELLMER AND THE RADIATOR LADIES - the upscaled version of it is on there, and (for private use) may be downloaded (up to the original size 7999x4363 at 72dpi). That image was generated with DreamShaper v8 which is a upscaled model that I use instead of the old SD 1.4 and 1.5. It is a free model, and the rendering is by far not as good as you would get in any of the SDXL models for which you have to spend points. DreamShaper v8 is sort of 'hit and miss', giving you some nice images at times, but failing at others. I use it mostly for testing prompts, but sometimes you do get some good stuff with it. Also here is a direct link to my published Nightcafe post, where you could check the prompts I used: THREE AGES OF DOLL WOMEN.

---

Above I mentioned the various models, and Nightcafe has now a large selection of SDXL models, but did retain SD 1.4 and 1.5. Aside from that, you get 'Community Models' which were developed/trained by members of the community, and they run in conjunction with SDXL 1.0 as do models you can train yourself (I so far trained one model for my drawings, and another for my selfies).

How do SD 1.5 and 1.4 compare to SDXL models?
Understanding the Difference Between Stable Diffusion Models

---

SO I TALKED ABOUT TRAINED SELFIES?
after all the exitement, I went back to check on the Insane Asylum, which was housed in a abandoned Cryptid Taxidermy Museum building
^{btw, same place some of the Bellmer Dolls wound up also}

---

Visit my website
▼

^{PRINTS OF MY ARTWORK AVAILABLE HERE:}


Bonus Special: JUST ANOTHER DAY IN BOGOMIL'S BRAIN

MY INTRODUCTION - ABOUT MY WORK

---

should you not be on the blockchain, or have no upvote power, or this post is already older than 7 days, you can always just buy me a coffee.