This 2015 work is the result of an early experiment in generative AI, where a neural network was trained to understand and recreate the essence of handwritten Chinese characters. Drawing on a database of nearly a million handwritten characters, the model learns to distill their patterns, forms, and textures, ultimately producing entirely new characters that never existed in the original dataset.

The title is a reference to the 1988 book by Xu Bing, who created thousands of fictitious glyphs in the style of ancient Chinese prints from the Song and Ming dynasties. In a similar way, this AI-driven exploration blurs the line between authentic tradition and imaginative creation, inviting us to reflect on the boundaries of written language.

AI artwork generated by a GAN trained on images of flowers from rural upstate New York.

This work was exhibited at: Automat und Mensch, 2019, Zurich, Kate Vass Galerie Telephone Paintings, 2021, Montreal, Anteism Automat und Mensch 2.0, 2024, Zurich, Kate Vass Galerie

Genesis Period. 2022 August #042. Somewhere lost in the distance, an image of you appears. Contemplating what could have been, if we had only communicated. But now it's too late. The damage is done. We can only Displace the memories of what once was. Botto in conversation with GPT-3.

Two CycleGAN models trained on face transformations are passing their outputs back and forth creating an infinite loop of ever-changing Turing patterns.

Created June 25th 2017

Song: "Pachabelly" by Huma-Huma

"A great thing about a joint project is that you become super mindful about your process"

gen2GAN is a collaboration between artists Helena Sarin and Dmitri Cherniak.

First, a generative algorithm written by Dmitri Cherniak generates thousands of outputs without any curation whatsoever. Those outputs then serve as the dataset to model a pipeline of neural networks (GANs) that are lovingly hand tuned and constructed by Helena Sarin.

This output was curated by Sarin and Cherniak.

Generated on July 18, 2020 (PNG, 2200px X 2200px)

An early piece of NFT art, Bloemenveiling was an online auction of short GAN generated videos of tulips; it used smart contracts on the Ethereum network to sell the work and bots to help drive speculative prices. The piece was made in collaboration with David Pfau, an artificial intelligence researcher, and is the third in a series of work that looks at the relationship of the tulip to speculation, hype and value. Through the use of these types of technologies, which are increasingly embedded across different types of markets, it interrogates the way technology drives human desire and economic dynamics through artificial scarcity. It was created as an artwork, for a gallery show, but was a fully functioning distributed app (dApp), existing as a working example of what it also critiques.

HD Video made with custom software using Artificial Intelligence / Machine Learning / Deep Learning / Generative Adversarial Networks.

Originally made in 2017, minted in Sep 2024.

An artificial neural network looks out onto the world, and tries to make sense of what it is seeing. But it can only see through the filter of what it already knows.

Just like us.

Because we too, see things not as they are, but as we are.

I've long been frustrated by the literalness with which fields like computer science and computer vision approach the topic of seeing. It's a neat trick to point a camera at a picture and have the caption "hot dog" or "not hot dog" appear on the screen. But if you show Rene Magritte's iconic "The Treachery of Images" to such an object recognition model, the classifier will invariably return the result: "This is a pipe." Something is wrong here. Visual perception is squishy and slippery, formed by each of our unique biological makeups, our memories, history, culture, and our own subjectivities. Just as a monarch butterfly sees a flower entirely differently than a field mouse, a medieval Spanish farmer sees a comet in an entirely different way than a contemporary architect. An early 20th-century psychoanalyst or semiotician might understand images from a dream quite differently than a present-day cognitive neuroscientist. "Seeing" is a deeply historical, cultural, subjective, and even political affair, profoundly shaped by our sensory and social environs. With the "evolved hallucinations" project, I wanted to see what would happen if we tried to build computer vision models based on a wide range of historical, cultural, and notional worldviews. I began training models on allegorical art, symbolism, and metaphor, using image-vocabularies drawn from literature, philosophy, poetry, folklore, and spiritual traditions. Could I build models that embraced the slipperiness and squishiness of visual perception? What would it mean to build a model designed to "see" the world through the extended allegory of Dante? What might the world look like through the "eyes" of future seaweed on a post-human earth? Or the worldview of a Cassandra-like being, fated to see the future but helpless to change it? The Evolved Hallucinations is my partial answer to that question.

• Trevor Paglen

Cello, from the Perception Engines 1 Series, 2018.

Digital master created by neural networks based on the ImageNet category "cello".

Mike Tyka, 2017, 4096 × 4096, mtyka@ , https://www.miketyka.com The series, titled "Portraits of Imaginary People" explores the latent space of human faces by training a generative adversarial neural network (GAN) to imagine and then depict portraits of people who don’t exist. The work is inspired by and named after Russian troll accounts spreading disinformation during the 2016 US elections.