As is often the case with AI-generated art, there is confusion whether attribution should be given to the person who created the AI challenge, the AI itself, the creators of the AI, or the artwork that the machine learning model “borrowed” from. In the case of the Museum of Modern Art’s latest exhibition on display this Saturday, you could have dedicated it to the first three. The last part is a bit grim.
MoMA is set to unveil a colossal 24-by-24-foot project titled Refik Anadol: Unsupervised this weekend, the exhibition will run over March 5, 2023. The artwork uses an artificial intelligence model designed by Anadolu himself, using 380,000 images of 180,000 artworks from MoMA’s collection to create a bubbling and turbulent stream of moving images. According to the museum releasethe new work on display is a “unique and unprecedented meditation on technology, creativity and modern art” that aims to “reshape the trajectory of modern art, pay tribute to its history and dream of its future”.
Anadol originally digitally published Unattended back in 2021 on the platform Feral Fileas part of his Machine Hallucinations series, which he started in 2016. According to the project description on Feral File, Anadol used several different kinds of generative adversarial network (GAN) algorithms to create generative artwork from “publicly available sources”. These images were then sold as (sigh) NFTs (which is also something MoMA is investigating recently). According to Fast companywho spoke with Anadolu, the team working on the AI that powers Machine Hallucination, also trained using an Nvidia DGX Station A100, a desktop-sized box used for “AI Supercomputing.”
Unlike the first time the art was displayed online, MoMA said the in-person exhibition also monitors changes in light, movement, ambient volume and even weather that affect the moving image.
In an email with Gizmodo, a MoMA spokesperson confirmed that the artist has been working with MoMA since early 2021, before the original Unattended art has spread to the internet. The work draws directly from MoMA’s publicly available resources GitHub metadata.
This kind of image generation GAN is different from diffusion models used to power state-of-the-art AI art generators, including OpenAI’s Stable Diffusion and DALL-E 2. While both use a kind of Gaussian blur to distort and then reconstruct images, Diffusion is an iterative process which works by using machine learning systems to attempt to generate “realistic” visual images. GANs on the other hand are more based an “adverse” system that produces “credible data” then uses a “discriminator” to decide whether that part of the image belongs. While Diffusion models often cluster images based on similar tags, the Anadol GAN model allows AI to generate its own images “unsupervised”.
Michelle Kuo, a curator at MoMA who was a leader in organizing the exhibition, he said in a report that “AI is often used to classify, process and create realistic representations of the world. Anadol’s work, on the other hand, is visionary. It explores dreams, hallucinations and irrationality and presents an alternative understanding of modern art – and art making itself.”
In an interview with Fast Company, Anadol said that essentially this kind of AI image generator was becoming “its own entity,” adding, “We don’t know what forms it can create.
This new artificial intelligence art installation may benefit from the fact that it doesn’t use the most notorious kind of image generation. While in 2021 far fewer people knew about the art of AI, in 2022 it was the biggest names in AI art generators have been rife with controversy, with much of the backlash coming from mainstream artists. These systems have scraped millions upon millions of images from the internet without permission, which are used to train AI models. Artists often find that their art styles have been co-opted by AI image generators, leading to concerns that their work could become indistinguishable from humans simply entering prompts such as: “Pablo Picasso style art”.
It’s there too burning issue of copyrightand whether an individual could ever claim to own art that was created using an artificial intelligence system that itself generated images from another copyrighted work.
MoMA Chief Curator Paola Antonelli said the new exhibition “underscores her support for artists experimenting with new technologies as tools to expand their vocabulary, their impact andd their ability to help society understand and manage change.’
#MoMAs #latest #exhibition #swirling #hallucination #museum #art #generated #artificial #intelligence