This community is an experiment using a combination of a few technologies to make emergent ‘Art’ that is ‘Fine’ (as in ‘good enough’).

The script runs once a day and produces two pieces (always have a backup) using an AI generator with a random prompt. Technical details are below, if you are interested.

If you don’t want to see this stuff, just block it and move on. The bot only posts these images to !fine_art@nom.mom. If you see them elsewhere, it was done by others without our permission. Speaking of permission, we claim no copyright over any of these images, and release them to the public domain.

If you have questions or comments about these images, this post is as good a place as any to leave them.

Technical Information:

The generator is Stable Diffusion, using the popular automatic1111 build. The model used has changed a few times (currently set to JuggernautX). The api is fed a jinja2 template that generates a sentence which becomes the prompt. The bot uses this prompt as the post title.

The sentence is built as a string of variables that are chosen randomly at each generation time. A small amount of effort was spent with regexes to make the sentence parse grammatically, however this was not the focus of the project. The awkward wording can be amusing at times, and doesn’t have much impact on the output.

The builder starts with the string:

{{adjective}} {{noun}} {{extra}} {{verb}} {{adjective2}} {{noun2}} {{adverb}} - {{lighting}}, {{artist}}, {{type}}, {{style}}.

It then fills in the variables with randomly generated phrases for each slot.

Examples:

Nouns - ( [ "narwhal","the ultimate warrior","king kong","godzilla","mothra","superman","batman","wonderwoman","spider-man","trout","red snapper","bluegil","tuna","shark","sharknado","baby shark","baby sharknado","glass","brick","tool chest","screwdriver","gravel","logging truck","lumberjack","bandit","commercial logger","senator","mattress","all the tea in china","magic apple","sleeping pill","glass slipper","ruby slipper","golden brick road","elves","trolls","winged monkeys","good witch","evil witch","little dog toto","dorothy","wizard","rogue","barbarian","warrior","thief","bard","gym teacher","history teacher","english teacher","rock band","operator","waffle","cupcake","bacon","dump truck","puppies","bowl of fruit","tinnitus","llama","gorilla","hippo","pizza","man","woman","girl","boy","cheesecake","cheese","duck","doctor","actor","singer","gamer","president","funeral director","mt everest","table","chair","kitchen sink","horse", "bacon", "duck", "telephone", "telegram", "velociraptor", "hulk hogan", "andre the giant", "mr. t", "doctor who", "lady gaga", "col sanders", "megatron","princess", "painting","witch","ninja","irritable bowel","outer space","cthulu" (..) ] | random )

Adjectives - ( [ "the","if you think that was bad,","for the second time today","after the dumpster fire went out,","while cutting off his own ear to impress a girl,","without even packing a suitcase,","hungry hungry","helpful","smiling","laughing","psychotic","bright","colorful","shirtless","armored","bachelor","compulsive","a heavy","a noisy","a nosey","chocolate","anime","anti gravity","translucent","upside down","slime covered","soaking wet","indestructible", "melting", "growing", "glowing","evil", "overpowered", "grape", "cherry", "banana", "orange", "vanilla","vain", "repulsive", "smoked up", "jr", "sr", "master", "king of the","cooking","duck hunting","a pack of wild","untold millions of", "solar powered", "mix and match", "underground", "weak", "patient", "an agitated","covered in bacon", "happy","salty","truck sized","rabbit hunting", "hundreds of","gigantic","upset","thirty","dripping","unsettled","imploding","exploding","obstinant" (...) ] | random)

Verbs - ( [ "bacons","plants a beautiful garden with","powders their nose while avoiding an awkward conversation with","powders their wig to prepare for court with","courts","leans down and kisses the lips of","finds themselves thrown off a 20' steel cage onto a folding table by","fries a pound of bacon with","puts up duck season sign with","hangs rabbit season sign with","blows down the house of","cries wee wee wee to""writes a ballad to","kisses","offers a kiss to","eats the same spaghetti noodle as","encounters","listens patiently to","forms a strong opinion of","scratches ","reaches behind ","looks confused at","glares daggers at","enjoys making a voodoo doll in the likeness of","mutates into a ","spends time mutilating","very carefully picks the meat from","savagely devours","gives a big hug to","shows how much it misses","sits down to talk to","falls asleep while driving a","repairs","ends","turns over a new leaf with","wishes it could life without","is emotionally distraut over ","offers condolences to","makes a backup of","takes a picture of ","draws a charactiture of ","paints ","licks","eats","stands on","burries","digs up","cheers up","offers a lollipop to " (...) ] | random)

Adverbs - ( [ "in a flash","for the fourth time","while waving to their mother","while knitting a sweater","while counting to ten","all the way home","instead of the opposite","without a net","in slow motion","three ways to sunday","encoded on an encrypted cypher","while wrapping bacon","while solving a rubik's cube","while parralel parking","without a license ","for luck","while everyone laughed","in the knick of time","while soaking wet","the other day","across the floor","without looking","at the zoo","at the park","in the parking lot","over the rainbow","in the woods","over the river","at grandmother's house","at danny's house","down the hill","over the hill","over the moon","and then ran away with the spoon","and then it started snowing","and then it rained","and it kept hailing","while driving","in the diving pool","in the bathtub","while counting snails","without saying please","while eating three hams" (...) ] | random)

The actual list is much, much, longer, but you get the idea. We experimented with having an LLM generate the prompts, but found that it didn’t generate as entertaining of results. I can post the full code, if there is interest.

After this, an LLM generates a few more words to pad out the prompt and bob’s your uncle.

Why?

What is the point of all of this? Honestly, just fun. These images display in the authors home on an LCD as framed art (part of a slideshow with other information). I thought they might be amusing to others.