KRЯRL DЯAWINGS: DECEMBER SUMMARY -- 2025

I had to try a lot of the new AI ART programs that were rushed out before Christmas, while it was a slow month for printmaking in December, but a decent month for figure drawing:

December 20, 2025

My drawing pushed through the latest AI programs

Quick QWEN Training -- AI ART Models

I generated this hyper-Baroque anthropomorphic figure

with a LoRA model based in part on the sculpture of Bernini

REVERSED -- Drawing from AI Figures

I drew free hand from some of

the FAKE figures I generated with AI --

and Nano Banana Pro added clothes to them

Laser Cutting -- into Metal Silk Screen

We successfully used the laser cutter for silk screen printing,

by using a metal mesh

DRAWINGS

KRRRL 2025 Playlist

KRRRL Drawings 2025

My YouTube Playlists

Best KRRRL Drawings -- December 2025

December 2, 2025

December 3, 2025

December 9, 2025

December 10, 2025

December 15, 2025

December 16, 2025

December 17, 2025

December 20, 2025

The final drawing session of the year

happened at the SECRET GALLERY in Albuquerque,

put on by Ben Harrison

I copied figures that I had previously generated with KREA (Flux) AI, thinking that I might absorb some of the aspects that I liked from the FAKE drawings:

REVERSED -- Drawing from AI Figures

I drew free hand from some of

the FAKE figures I generated with AI

PRINTMAKING

While I did not do much printmaking in December -- we did manage to laser cut into metal mesh at Quelab, to make silk screen prints. That was HUGE:

Laser Cutting -- into Metal Silk Screen

The laser did not cut the metal mesh screen,

but it did burn off the emulsion

to make a silk screen for printing

Remarque Print Workshop had their 13th Annual International Juried Print exhibition, which opened on December 12, 2025:

December Exhibitions and Events

The opening at Remarque for the International print exhibition

December Exhibitions and Events

Aniol Barris standing in front of his print

HELLO PRINT FRIEND pointed to the 20 minute documentary they made about printmaking in Oaxaca this year:

Grabando Oaxaca

There was a great exhibition of Oaxaca prints this year in Albuquerque in October and November 2025
Last year 2024 there was a great exhibition of Oaxaca prints at UNM when Pavel Acevedo came to Albuquerque

Toru Sugita at Taller Grafica Libre of Oaxaca

The best gallery is in the street!

AI ART

An avalanche of new AI programs came out again before Christmas -- in particular GEMINI 3 (including Nano Banana Pro) and ChatGPT 5.2 (including Imagen 1.5):

Nano Banana Pro

let me put CLOTHES on my naked figures

December 17, 2025

Nano Banana Pro clothed my AI figure

before Midjourney animated her

December 2, 2025

Nano Banana Pro transformed my drawing

into a 1968 psychedelic poster

AI Revisited -- AI Animations from AI Stills

From my house to Tucson --

one of several AI Animations in a compilation YouTube

I wondered if I could make longer coherant Midjourney AI videos. Midjourney allows one to make videos with a starting frame and and ending frame. So repeated the process to generate a second video, but now using the prior ending frame as a starting frame, and adding a new AI image as the second frame. I repeated this process several times in order to make a longer and coherent video, and posted the result on YouTube:

Midjourney Continuity -- Rope Desert Revisited

Midjourney Continuity

I used the new Image 1.5 model from ChatGPT to composite some of the better AI still images that I had generated in the past. While the result was decent, the image was rather cluttered. I then animated the composite image in Midjourney and posted it on YouTube -- and got two negative comments, "Why?" and "Booooo." I can't say that I disagree with those viewers. This was all I could generate before ChatGPT cut me off (on the free plan) -- so I posted the videos to show how far I could go:

GPT Image 1.5

Cluttered AI image.

by a fantastic AI ART program

I uploaded a 2D image to GEMINI 3 which turned it into a 3D object, by writing code for OpenScad:

3D Revisited -- from Gemini 3

The far right grey 3D object generated in Gemini 3,

is the best representation of the far left 2D image

(better than the middle object generated by Gemini 2.5)

We even uploaded a freehand sketch, and got GEMINI 3 to generate a 3D file -- by writing code for OpenScad -- which we later 3D printed:

3D Revisited -- from Gemini 3

Hand drawn to 3D printed

GEMINI 3 even wrote code for AI ART programs, such as this VOXEL object generator below:

VOXELS -- coding

Javelina created in the

VOXEL object generator

However I could not always get GEMINI 3 to write the program exactly the way I wanted. I did get something mildly satisfactory with a program I wrote and named BURST :

HTML 3D Animated Lines Image Animator

BURST

December 20, 2025

BURST animates an image that one uploads

And GEMINI 3 did not do a great job at writing a simple Art program:

"Make a simple drawing program in the browser

where all the straight lines go to infinity (the edge of a box),

and all the curved lines make a circle."

Gemini Coding Revisited

The circles I drew

weren't even visible until I downloaded the image

AI WORK STATION

The NVIDIA DGX Spark - 4TB costs $4000, but it is plug-n-play to run many AI programs (like perhaps Docker files). Wouldn't it be nice to have this at Quelab to run AI ART programs?