It took me half of January 2025 to crawl out of the previous year, as I was still summarizing everything that happened in 2024:
I pushed AI ART as far as I could, deep into 3D:
Midjourney let everyone generate AI images with lightening speed during their Relax-athon this month -- however the bigger, but more confusing AI news, was that China released their DEEPSEEK and JANUS PRO 7B AI models. We printed at Remarque Print Workshop twice, during their OPEN HOUSE, and to make EXQUISITIE LITTLE BIRD tessellation prints. This year I am trying to draw darker, and even started painting this month:
QUELAB 3.0 is stepping it up in their new building next to the University of New Mexico. While everything is not finished yet, people have come back in 2025 with a vigor to make things. Now it's Turtles All the Way UP!
James Black projected his drawing during
on a building at Central and 4th in downtown Albuquerque --
and I animated it with AI in Haiper
DRAWINGS
I have also been converting my gesture drawings into vector drawings, and figured out how to manipulate them online:
Further altered online
with Vector Paint
LITTLE/BIG PRINTS
PRINTMAKING
We pulled LITTLE/BIG Prints -- twice at Remarque Print Workshop in January:
OPEN HOUSE
Ellie Weadock and Virgil Velasco pulled EXQUISITE TESSELLATION Prints during the OPEN HOUSE at Remarque Print Workshop on January 11th:
EXQUISITE LITTLE BIRDS
GLASS DOORKNOB
Ellie printed Burke Gurney's linocut by rubbing a glass door knob over the paper:
- Also we have an EXQUISITE TESSELLATION Print exhibition coming up at the University of Juarez in Mexico, in August, at the Centro Cultural de las Fronteras (see more details)
AI ART
Midjourney allowed everyone to generate images with lightning speed with their Relax-athon, and in January they were no longer severely censoring me -- so I pushed the Midjourney Moodboards to generate images (--p m7269599726876491811):
I ran my drawings through Midjourney for some interesting results:
CHINESE AI PROGRAMS
MULTIMODAL
I uploaded my drawing into (Hugging Space) deepseek-ai/Janus-Pro-7B, and it accurately described my drawing in words:
What is this?
"This appears to be a sketch of a person sitting down. The lines are drawn in a loose, abstract style, and the figure seems to be composed of various connected shapes and lines."
(Hugging Space) deepseek-ai/Janus-Pro-7B also uses AI to generate images. So I generated the image below from the description of my drawing, and got the following image:
- On January 16th I uploaded my drawing to Google and asked questions about it
- On January 21st I asked DeepSeek directly "What kind of drawing style does Krrrl use?" and got a robust answer when I allowed it to go to the Internet
- On January 21st DeepSeek wrote a program for me -- ANIMATION
- In the BRAIN MAN post I asked (Hugging Space) deepseek-ai/Janus-1.3B to describe a Midjourney image, and then generate an AI image based on that description
- On January 22nd I used (Hugging Space) akhaliq/anychat to write a program for me using AI, animating images -- LINES
AI TERMINATIONS
HAPIER closed down on February 1st, so that we could no longer generate videos for free on their website (likewise those who paid for a subscription could no longer generate AI videos):
24850 hours to generate an AI video
I liked the AI videos that HAIPER generated for me. However I was a free user, so I cannot complain about their service terminating. However it does look like they transformed into VideoCraft, a paid service.
VIGGLE also seemed to go down on the last day of January. At least their prompt website went black, and my request on Discord failed to generate a VIGGLE AI video.
PhotoMosh is not working.
Midjourney rejected my payment on February 2nd, 2025, even though there was definitely money in my bank account.
I was having this problem paying Midjourney in January last year, but managed to do it and get back to generating AI ART by January 10th, 2024.
**Essentially I am going to have to generate AI ART on a local computer, untainted by the middlemen of the web services. We are already setting up to run AI ART programs on a computer at Quelab, using PINOKIO to install Stable Diffusion and downloading models from CIVITAI. Perhaps we need to install JANUS PRO 7B as well.
3D ART
I converted a few 2D images to 3D objects using a couple of programs on Hugging Face:
DOWNLOAD: Mushroom_lady.GLB (1.79 MB)
HOLOGRAMS
for Looking Glass Go
The links below are PLAYLISTS of 3D views -- which can also be seen as "holograms" on the Looking Glass Go:
.gif)
3D VIEWS
January 19, 2025 (blog post)
.jpg)
Virgil Velasco made a new print
for the Year of the Snake
- HORNED TOAD PRESS got some love from HyperAllergic
No comments:
Post a Comment
Note: only a member of this blog may post a comment.