Artists and Algorithms Friday Feb 23 @ 7pm @Kaleid DTSJ

Neither Artificial nor Intelligent: Artists Working with Algorithms

Panel discussion: Friday, February 23rd 7pm–9pm. Free admission

Featuring Artists: Julie Meridian, James Morgan, and Steve Cooley

Moderator: Cherri Lakey

KALEID Gallery is proud to host this panel with three talented artists who are using various AI tools in their artistic practice while navigating all the ethical and creative dilemmas that arise with it. With all the controversy around AI collaborative / generated art, we’re looking forward to hearing from these avant-garde artists that are exploring the possibilities of a positive outcome for artists and creatives in this as-of-yet undefined new territory.

Julie Meridian:

Julie Meridian captures the essence of people and places through representational art with impressionist details. She has extensive experience humanizing technology through UX design and has been active in the Silicon Valley arts community since 2014. “Terraforming the Uncanny Valley” exhibit featuring paintings by Julie Meridian in collaboration with Berlin-based tech artist Lucidbeaming is currently on view at KALEID Gallery. You can visit her website at www.juliemeridian.com/

James Morgan: 

James Morgan is an artist, educator and narcissist. As an artist he is interested in how theory and meaning intersect with the work, his interest in the way that people communicate and work with and around the artwork has led him into the game space and virtual environments. As a collaborative artist he finds Artificial Intelligence to be a way to expand his work into new areas that he has little or no expertise. You can hear his album at https://bit.ly/ujm-frank

Steve Cooley: 

Steve Cooley has been active in the south bay arts community for about 20 years. He is a former president and current advisory board member of Works/San José. He’s also in a cool live techno music band called Haptic Synapses. Cooley has worked in tech jobs since 1997. You can find his art and music here:  http://sc-fa.com

Moderator Cherri Lakey: 

Cherri Lakey and partner Brian Eder are the owners and curators of ANNO DOMINI // the second coming of Art & Design, KALEID Gallery, and a bunch of other passion projects that give artists opportunities to be their authentic selves and share their extraordinary art with a wider public audience. She is intensely interested in the history and power of Art and Imagination juxtaposed with the challenges of the human condition in modern times. Additionally, Lakey is a contributing writer at KQED Arts & Culture. You can visit gallery sites here: www.galleryAD.com & www.kaleidgallery.com

Panel discussion: Friday, February 23rd 7pm–9pm. Free admission

Reserve your spot now by RSVP to our event page: https://fb.me/e/50XblB06r

KALEID Gallery

320 S. 1st St.

Downtown San Jose

info@kaleidgallery.com

Post taken from announcement and designed by Cherri Lakey of Two Fish Design

We all know that Art is not the truth.

“We all know that Art is not truth. Art is a lie that makes us realize truth at least the truth that is given us to understand. The artist must know the manner whereby to convince others of the truthfulness of his lies.”

― Pablo Picasso

And CEO’s may be catching on, but certainly AI is not the truth either.

https://www.theverge.com/2024/2/2/24059955/samsung-no-such-thing-as-real-photo-ai

It is funny to think about the number of times we’ve been through this, with photography, with drum machines, with photoshop.

Give me more Sizzle: http://bryanfpeterson.blogspot.com/2010/06/every-photograph-is-lie-yet-within-that.html

You Love the Game (the story of Uncle James Misty)

This is how it begins (Sept 21, 2023):
“I have this intriguing idea for a project centered around Taylor Swift’s 1989 album. I believe she lost the rights to it at some point and subsequently re-recorded it. In homage to this situation, I plan to re-record 1989 using an AI tool that I’ll train with my own voice. Afterward, I’ll replace the lyrics in the original version, not Taylor’s rendition, and release it on YouTube, possibly elsewhere, though it may raise some legal concerns.

One question that arises is whose copyrights am I potentially infringing upon? I am willing to compensate for the usage, much like Ryan Adams, who was presumably did when he recorded Taylor Swift’s 1989 album, and certainly, for the father John Misty recordings of Ryan Adams’ 1989 album. This project involves numerous layers, making it beautifully complex.”

History:
Several years ago, I became enamored with the parody covers performed by Josh Tillman, also known as Father John Misty, singing renditions of Taylor Swift’s 1989 album. This fascination coincided with my growing appreciation for FJM and the Hollywood Forever Cemetery. It didn’t hurt either that the video featured Aubrey Plaza in a striking role.

I even managed to incorporate a discussion of Taylor Swift into one of my lectures for my Intro to Digital Media Arts class. I admire the sense of appropriation displayed by Ryan Adams when he covered the entire album 1989. While it may not be entirely fair to Ms. Swift, who has faced numerous challenges in her career, she always seems to handle them with grace. I often jest about Ms. Swift, but I hold her in high regard as a performer deeply dedicated to her audience and her craft.

What do I mean by “mishaps?” Well, there was that infamous incident with Kanye West, where she was interrupted, and the entire conversation about her work got derailed by ego. Then, there were the master recordings of her first six albums, which ended up in the possession of her record company and were regrettably sold to hedge funds, even as she attempted to buy them back herself. The silver lining in these two incidents is that she is now worth more than Kanye West, and she has re-recorded her early albums “Taylor’s Way.” It’s a tale of poor judgment on Scooter Braun’s part, betting against her for a quick profit on her workf. The re-release of her albums not only made her a substantial sum of money but also granted her complete ownership of the new master recordings.

Now, let’s talk about the album 1989, released in 2014 (https://en.wikipedia.org/wiki/1989_(album). It’s far from an unremarkable album, featuring several chart-topping hits and selling millions of copies. This album played a pivotal role in Taylor Swift’s transition from a country singer to a pop sensation, spawning numerous big hits.
The song Blank Space has a delightful lilt to it and a terrific story laden video:


In 2015, Ryan Adams released his own version of the album 1989 (https://en.wikipedia.org/wiki/1989_(Ryan_Adams_album)), providing a track-by-track cover of Swift’s album of the same name. This reinterpretation was well-received, particularly by Swift herself, who regarded it as a fresh perspective on her work by a respected musician. Adams considered the album to exist in its own alternate universe, filled with joy and inspiration. I view this as a form of artistic appropriation and reinterpretation, akin to the work of Marcel Duchamp.
Adams song was apparently a hit with Swift who was pleased to have her work covered by someone she respected and grew up with. I find the cover and the change of lyrics (regendering it) to be a little off but the work of the whole album gains a certain amount of respect from me.


Josh Tillman, known as Father John Misty (https://en.wikipedia.org/wiki/Father_John_Misty), initially embarked on re-recording (in a parody fashion) the Ryan Adams tracks, taking them in the style of the Velvet Underground. However, he had a dream visitation from the spirit of Lou Reed, who urged him to cease the project. Consequently, he abandoned it. What Tillman did manage to record though is haunting and beautiful. The obvious parody of it, and in the style of the Velvet Underground sells me completely on the artistry. Tillman recorded Blank Space, and Welcome to New York by the time he received the revelation that derailed the effort to complete the 1989 album. We are poorer for this.


This leads me to 2023 when I am teaching a course on AI tools for artists and the above idea hits me. I spent an entire weekend working over and listening to the album 1989. I can’t say that the music has really stuck to me in an appealing way, but many of the tunes are quite catchy. I recorded and trained an AI algorithm on my voice (I read The Raven by E.A. Poe to stay on brand). The plan was to cover all of 1989 in the weekend and release an entire album, but guess what? AI is not that sophisticated, yet. And vocal effects throw it off, I hadn’t realized that there are quite a lot of vocal effects and background singing going on in much of the album.

But I am left with Uncle James Misty, my parody of Father John Misty because I am actually an Uncle James (drive me nuts, but it is what it is). So this is my parody of Josh Tillmans parody of Ryan Adams cover of Ms Swifts song Blank Space. Please enjoy and if I am able to continue this project, I’ll continue to post here

a non apology

[Your Name] [Your Address] [City, State, ZIP] [Email Address] [Date]

[Recipient’s Name] [Recipient’s Position/Title] [University/Organization Name] [Address] [City, State, ZIP]

Subject: Apology for Using ChatGPT and Committing Academic Dishonesty

Dear [Recipient’s Name],

I hope this letter finds you in good health and high spirits. I am writing to express my deepest regrets and offer my sincere apologies for my recent actions, specifically in relation to my usage of ChatGPT and the subsequent plagiarism and academic dishonesty it has caused.

I acknowledge that my behavior was completely unacceptable and goes against the core values of integrity and honesty that are expected within the academic community. My actions not only undermined the trust and credibility that I should have upheld, but they also disrespected the efforts of my fellow students and the educational institution as a whole.

I take full responsibility for my actions and understand the gravity of the situation. Plagiarism and academic dishonesty not only tarnish my personal reputation but also compromise the reputation of the university and the academic integrity of the entire community. I deeply regret that my choices have caused distress and disappointment to those involved, especially to you, the faculty, and my peers.

It is essential for me to emphasize that my actions were not a reflection of my true character or my commitment to learning and personal growth. I have always valued education and the opportunity to acquire knowledge, and I am fully aware of the importance of maintaining the highest standards of academic integrity.

To rectify this situation, I am taking immediate steps to address the consequences of my actions and prevent any further damage. Firstly, I have ceased using ChatGPT for any academic purposes, as I now understand the negative implications it can have on my own development and the academic community as a whole. I have also taken it upon myself to revisit the courses in question and relearn the material, ensuring a comprehensive understanding of the subject matter without any reliance on external sources.

Additionally, I fully commit to making amends for the academic dishonesty I have engaged in. I am willing to accept any sanctions or penalties deemed appropriate by the university. Moreover, I am eager to participate in any educational programs or workshops available to help me rebuild my academic integrity and prevent such behavior in the future. I understand that regaining trust is a gradual process, and I am dedicated to proving myself through my future actions.

Please accept my heartfelt apology once again for my misguided choices. I genuinely regret the pain and inconvenience I have caused, and I am committed to making a positive change in both my personal and academic life. I hope that my remorse and willingness to take responsibility will be taken into consideration.

Thank you for your understanding and for giving me the opportunity to express my sincere apologies. I genuinely hope that, with time, I can regain your trust and the trust of the university community. I understand the significance of the lessons I have learned and the importance of upholding the principles of academic integrity.

Yours sincerely,

[Your Name]

(written by ChatGPT) (prompt: Write me an apology letter for using chatGPT and potentially plagairizing and committing academic dishonesty.)

Regarding Chat GPT

We live in an interesting time. I find myself both amazed by its foolishness and impressed by its cleverness. In my class, I have had students use GPT chat for research purposes, such as finding new media artists to study. One project had them search for a doppelganger in the new media art.

I am also hopefull for improve writing quality through GPT chat. I plan to conduct experiments, including with this blog post, to see if chat GPT can make text more readable, even if it is transcribed through voice-to-text methods and may be a bit rambling.

I find the level of absurdity in GPT chat to be a source of fascination. Despite being an impressive and convincing liar, it sometimes spews out false information among truths. While this could be seen as undermining, as a teacher I view it as a valuable feature. It means that all information from GPT chat must be verified and researched, thereby promoting critical thinking and mental processing.

Its flaws actually align with educational objectives, as students must engage in critical thinking and verification While some students may attempt to use it for unethical purposes, I am confident that the majority will utilize it to expand their knowledge and deepen their understanding. GPT chat presents both a challenge and an opportunity, and it will be interesting to observe its evolution in the years to come.

Posted in AI | Tagged

Some Boring Dall E2

I am legitimately having some issues with WordPress eating edits, so this is likely to be short and ill informed.

Yesterday I ran into Don Feasel https://donaldfeasel.com/ in the hall and we talked a little about the upcoming semester at SJSU and I started talking about the Machine Learning class I am prepping for. I think I am going to start a short series where I try to do variations using Dall E2 and painters I know, at least to start with. Here are a few of his images, and the Dall E2 variations, with his permission, of course:

I am interested in abstraction here only partially because there is a lot of subjectivity about what makes a good image. Of course we have line, form, composition and half a dozen other features that I just don’t have a giant amount of experience with. I asked if he felt that the variations were soulless. Dall-E seems to capture a sense of the image but details is lacking, I am also noticing that Dall-E really favors symmetry.

Don’s work is the upper left and has a distinct shape to the action of the work, the variations are much more gridlike, thought they capture some of the energy.

The second image is interesting in the forms that are created but Dall-E seems to be stuck with an X form. Don’s work has a much more dimensional quality and he uses the space of the image in a much more sophisticated way. The Dall-E variations seem to be variatons on parts of the image and not the greater composition.

The Artist says: “These collected bodies of work trace a journey from more to less control. A growing sense of the futility of manipulating materials to produce expressive content consciously set me on this path. I enjoy constructing compositional processes that give me surprising results when they run their course. While painting I try to sustain a state of tension between knowing what I’m doing and being lost. I’m free to reject the efforts that don’t sustain my attention as every attempt pushes the boundaries of a self-imposed system. It provides the incentive to try again—as any gambler would play one more card.”

The artists work is far more rich and complex than the variations. How much longer until that gap closes?

Art Fart

Using the variations on classic art, and I hope to see some aspects that go beyond mere style.

Did Duchamp steal this from the Baroness Elsa von Freytag-Loringhoven? Do I feel guilty using AI to create variations? This fulfills the prophecy. I love the porcelin but I love the background even more. Toilets for everyone!

Ad Reinhardt managed to take abstraction to its logical conclusion. Surprisingly it caught something from the unphotographable image. Unsurprisingly, I think it comes from the extended edge of the image.

All black becomes all white, and though I have had my challenges with Malevich, the algorithm finds the most boring options for the variations. Unsupremitist.

Rothko broke Dall-E for the first round, I felt super guilty about cropping the image and so I cannot really get into the results. Same colors, thought.

Have you heard about the painter Vincent Van Gogh,
Who loved color and who let it show.

This is the basest most boring level of using these multi billion parameter language models, but I cannot help but be a little fascinated by it. Maybe it is about walking some of art history through this to see how it responds. I still hear in my mind responses to Malevich and Reinhardt. Good times.

Comparing Dall-E and Disco Diffusion

I have managed to spend some quality time with Disco Diffusion 5.2 since the beginning of May I have paid Google $10 per month to host some Collab notebooks. I have run well over 100 prompts and for many for 100-200 images, I have about 6000 created images in my drive. The process is excruciating as I have to run a browser and not close the window during generation. If I were to choose to, I could pay $50 to be able to run it in the background. I have considered this, but will be tabling that idea as I have located some high performance computing on campus. The translation from Google Collab to Jupyter Notebooks is reportedly simple, but I am just not familiar enough with the structure to make it seamless. Anyhow, that is the cost of DD for me at present, and though it takes about 7-8 minutes to generate each image, it seems like a decent deal. I have started to create cards / tarot cards and am including an example:

“mystic surreal arcane tarot card”

Regardless 6000 images satisfies my need to make lots of images with these algorithms.

Dall E 2 https://openai.com/dall-e-2/ is another beast entirely, it only runs on Open AI’s cloud and costs about $15 for 115 “credits” that can generate four images from an image or text prompt. Output is 1024×1024 and there is no tweeking the algorithm.

Dall E2 variation on the top of the previous image

Dall – E’s images are much cleaner and much more concise, they come off as legitimate illustrations of the text concepts and creating variations off an existing image is just beautiful. As with Disco Diffusion, there is a certain amount of time and experimentation that is done learning how to talk to the Algorithm and thankfully Open AI provides a LOT of guidance on crafting prompts.

original and three variations by Dall E

Dall -E is quick and the visual variations are stunning, but as of this writing 8/22, it will not accept human faces as prompts, or any part of the image, and is fairly restrictive about what it will permit in terms of text prompts. Poking at a few of these, I have come to love the prompt of “erotic Cthulhu” as part of

(this post was eaten. Well half of it was and I need to go back and rewrite it. 8-8-22)