AI know the way. As a tool for content creators and artists, AI systems will become ubiquitous. The visual appeal of images created by GAN is impressive when properly curated, and the ease with which the tools can be used to create fully worked out visualisations of even rough text concepts is a miraculous time saver for creatives in many sectors. But what part of the images created by these algorithms can still be seen as art? The artists using AI today can provide both the question, the discourse, and the answer to it, if we look at the conceptual and curatorial decisions they made. This exhibition of artworks consists mostly of pieces from my collection, but I also added a number of artworks that I couldn't leave out and that I don't own; these are mentioned accordingly in the texts. Let's go!
"Empty Room #01" by Nicola Villa combines an AI generated environment - a room in a museum or a gallery - with a small GIF of bright colours and flashing pixels. The image thus is a hybrid of images created by an artist and an algorithm. By the distribution of these parts and the subjects the artist chose for each part, he created a comment on the emergence and impact of AI-assisted art. The AI part of the image resembles classical paintings in style and lighting; the included image on the left reminds me of the works of Goya, while the flashing GIF appears digital and pixelated, reminiscent of the colours and glitch used by digital artists like XCOPY. Nicola seems to say that, since the data base for AI algorithm consists of human art, the tools will recreate this kind of art, while the innovation at the source of it was created by humans. But did you notice how tiny the man-made artwork is in size compared to the huge, mostly empty room of AI origin? Maybe this is a comment on the current AI craze with nearly all visual digital artists trying out the tools and releasing the results of their experiments, creating an enormous flood of creativity that, at times, can seem to push the efforts of traditional digital artists into the background.
Who is the creator? But can we dismiss AI-art as derivatives of the human artworks that the algorithms are based on? "pixelfusion" by @pixelfiller is a generative art project that creates images with a GAN that was trained on the outputs of another AI-system, adding another step to the process that puts further creative distance between the human and the machine contribution. But the code that does this was coded by the artist and released as a generative art piece, delivering 256 editions of pixelated transcripts of the communication between two AI systems, carefully moderated by the artist. For their artwork "In Warhol - does anyone buy canned soup anymore?", the artist RARE approaches this concept in a different, non-generative way that emphasises the artist's contribution. They had a GAN recreate one of the most iconic pieces of mass produced pop art and manually reconstructed and repainted it. I'm sure Andy Warhol would approve!
Do robots dream of electric sheep? The algorithms' lack of comprehension often creates surreal and dream-like outputs that many artists went on to explore - and who wouldn't want to set foot in these endless worlds of incomprehensible and unintentional images? Artist DefNoiz creates a whole universe of purple haze, and, after looking at the barren, surreal landscape of "The Howl", or the pale, almost translucent beauty of "The Tenderizer", I don't feel any wiser about the background of the world I'm seeing, but I have a whole lot of stories on my mind. The lack of context and the distortion of fine details in many AI generated artworks is comparable to how impressionism changed painting. It shifts perception towards light and atmosphere; AI can't construct meaningful situations (yet), so our brains will put it into these images in their place. In a way, the images feel like visual prompts to tell a story. This is kind of the reverse action of what the artist did to get the outputs from the software. A bit like a broken telephone game - can we still make out the meaning after the to and fro?
Building the (un)imaginable. Architect and artist Michael Hoehn envisions edifices that seem to have an ominous life of their own. "Delirium" sets the stage, and the dark urban environment the image shows is reminiscent of a futuristic Mordor. The non-congruent reflection of something else in a puddle reminds us of the imaginary nature of the dystopian urban landscape and of the fact that laws of nature don't exist in the AI's algorithm. "Urban Imaginaries_001" seems to show the close-up of the buildings seen in the far distance of "Delirium". The concrete facade of a building is morphed in torturous, organic ways that scream "nightmare" at our subconsciousnesses.
Exploring space. When Michael combines AI with the traditional aesthetics of his profession, he creates artworks that resemble architectural drawings and sketches. And still, the strange intersections and an uncanny feeling of mismatched space and dimensions let us halt and stumble.
Without and within. Colour is something that AI tools seem to have a special approach to. "Spatial Complexity No. 001" by Michael Hoehn introduces a dash of rainbow colours and gives the billowing sculpture an animated vitality. The watercolour-like texture is a thing that to me is a special quality of GAN-generated images; this is contrary to the glitch surrounding the architecture of "glitch station alpha" by voxeljunk. It makes the building seem to float on a faraway planet's ocean, and though it uses a similar palette, it has a different, harsher visual quality. I wonder if it was manually added to the AI landscape by the artist. Despite being an artefact of data processing by computers, glitch seems to be something that AI algorithms usually don't recreate.
Vivid Emotions. When they are given the appropriate prompts, AI systems show an astounding mastery of colours to evoke feelings. In "Emotion in Minimalism" and "Emotion in Minimalism pt 2", Rikkar managed to achieve strikingly contrasting colour temperatures that support the interpretation of each of the artworks as a display of conflicting emotions.
Scanning for Symbolism. One of the key vectors that transport the symbolism of Kika Nicolela's "Octopus Dreams" is her use of colours in this surreal and fascinating series. Deep, thoughtful blue, and brigth, lascivous red support the notions of alien intelligence and strong and confusing eroticism ("Hold that thought" not part of my collection).
It's the end of the world as we know it. "Nice views from here" and "Ramen Bar 2077" by cybxrg.tez make use of contrasting colour temperatures, too. Especially in "Ramen Bar 2077", this contrast conveys an isolated homeliness of the old-fashioned establishment that is surrounded by a harsh, technical world. Ironically, the scenery of a bar and the colours that are used are reminiscent of Hopper's "Nightawks", but while that masterpiece of American realism speaks of loneliness, emotional alienation, and a growing fear of war, the artist chose a more straightforward symbolism here. The unnaturally intense saturation of the whole image feels fitting for a world of the future that is flooded with commercials that are screaming their messages into their targets' faces with the intensity cranked up to 11.
Finding solitude. AI-assisted art doesn't have to be glaring or disturbing, though. Artist makeitrad uses GAN and combines the outputs generatively into peaceful, yet subliminally surreal images. Their "Solitude" series shows scenes of crystaline buildings in naturalistic surroundings. The images have a photorealistic clarity to them: occasional campfire sparks fly through the air, and the lighting feels so real that you can actually think you can sense what time of day the images are supposed to show. But this is contrasted by the incomprehensible, mind-boggling details of the AI-generated structures and buildings, and the surreal, mirage-like mountains looming in the skies of some of the images. The confusing effect of seeing a mirage here is mirrored by our reactions to AI-generated images. A masterful combination of digital art techniques to achieve a unique atmosphere.
Symphonies and coffee. In his series "Smyphonic", Evan Greene uses AI to explore images of musical notations and instruments in various interpretations of early 20th century abstract art styles. All of the images seem to use Piet Mondrian's bright primary colour palette and show an impressive stylistic cohesion and beauty. What makes this series especially noteworthy to me is the fact that AI understands neither music nor visual art, which makes my mind go on a small journey and lets me think of the saying that to write about music is about as adequate as to dance to architecture. I find it both mind-blowing and hilarious that AI tools deliver fascinating, Frankensteinian creations to any prompt, even to those that humans would struggle to combine meaningfully, without knowing about the contradiction of what they are doing ("Symphonic 15" not part of my collection).
Artificial ambiguity. Jerome Herr has mastered the creation of stylistically homogenous images and video loops with AI assistence. His most famous creations show heavy influences from American realism, something the artist self-reflectively approaches with "American Awkward". This image can either be seen as a parody of, or a hommage to "American Gothic" by Grant Wood, one of the most iconic artworks of that style and era. "Trippin' hard", one of Xeronimo's surreal, confusing video loops from AI images, shows both the artist's precise directions in creating these artworks, and the fractional variations AI tools iterate between individual frames. The shifting and morphing animation results in a wavering, unsteady style that, combined with the sober, muted colours and the explicit title, lets me think of the Beat Generation's desire to break free from a bourgeois post-war society depicted in the artworks of American realism ("American Awkward" not part of my collection).
A world beyond the blinds. When AI-assisted art doesn't give us a context or recognisable references, interpreting the intentionless output of a ghost in the machine can be quite revealing of our own thoughts - almost like a digital Rorschach test. For "The Fichtner Protocol", artist Alex Dalia provides fragments of an explanation in the piece's description, and I think that they intentionally use black & white to make abstraction easier - like said ink blots. The image shows a clear crescendo of brightness towards a large, featureless window with blinds, hinting at enlightenment and pure beauty waiting there. A standing person appears to be passive, waiting. There's open water between them and the window, and the majestic dimensions of the mountains in the background hint at the vast distance they would have to cover to get to the window. Where is this place we are seeing? Some white speckles of static are reminiscent of watching TV back in the days. And what else is a screen but a window into another world? The problem seems to be that the noise is on our side of the window - so are we the ones locked into the device, waiting for someone to turn it on and look at us or enable us to catch a glimpse of the otherworld? Who is watching, and who is watched? In the case of seeing meaning in AI art, it is definitely us who are watching ourselves.
Is this thing on? "Faces of Dementia" by Kaspar Zimmerman is a series of eerie portraits. It visualizes dementia - the process of forgetting. Most of the subjects seem to have given in to their fate, although some appear to speak or even cry out through partly opened mouths. The faces are distorted and in process of dissolving. The fact that these images were created by AI tools underlines their subject: computers had to take over because human brains don't work anymore. The high level of detail of the images results in a palpability that, ironically, alienates us further from the people we see, because it is not only the face with all its wrinkles and features that is shown intricately, but also the mist that the persons dissolve into. Being a medical doctor, the artist has released fascinating art that deals with the interaction of humans and medicine before. Being a medical doctor myself, I find this a deeply touching topic. There is no cure for dementia. There probably never will be. Gradually, a person will cease to exist, even before their death. Yet until then, the people suffering from dementia are still there, and they're often helpless and scared. The feeling when what you think doesn't seem to match with what the world around you experiences must be terrifying. Yet the artist shows both the inevitability that the person will ultimately dissolve, and also their resigned, sometimes even peaceful expressions. In the way the artist created these portraits, I see at the same time a call for sympathy for the people suffering from dementia, as well as a comfort to those around them. "The person you called is unavailable" by Haze Long seems to pick up on that thought of a person being gone from a body. Underneath the young woman we see on the outside, there is just a placeholder image left - a reminiscence of the test screens that are a relic from the past when linear TV still was a thing. But the colours also remind us of rainbows, and unlike in "Faces of Dementia", the human on the outside seems to be very much alive, even mischeviously grinning. Given the likeness of the woman to the artist's profile picture, artist and model probably are the same, and this is intended as a self portrait. The description included a refusals to "BEHAVE LIKE A NORMAL PERSON", and where Kaspar's piece shows people devoid of the person they once were, Haze's image shows the opposite: a woman full of personality, but just not behaving the way the world expects her to.
Escalate art. After this extended tour, we are back where we started. Or aren't we? Actually, we get to see "Empty room #12" by Nicola Villa - a frightening, dystopian counterpart to the artist's initially displayed masterpiece "Empty Room #1". The later iteration of the Empty Rooms series does away with the warm light and classical ambience of the opening piece. The walls of the room are made of bare concrete, and there is blood - a lot of blood - running down from it. This is not a gallery anymore! The central artwork grew from the tiny, minimal pixel grid of the first work to a big display of digital lightning strikes - and not the kind that deliver 1.21 gigawatts of power to take you back to the future, but the menacing ones from horror movies and ghost stories. Despite the fact that both the AI-assisted environment and the manmade digital artwork have changed into dark and threatening versions, this coherence doesn't unify both elements, but rather shows an increasing opposition, as if they both have dropped the veil of diplomacy. Whose blood is on the wall? The one of those denying AI-assisted works their status as actual art? Why is the manmade digital artwork no longer displayed in a gallery, but hidden in some bunker or prison? Is that spotlight reserved for AI art in the world we're looking at? Does the nervously flashing, framed artwork remind you of some kind of sci-fi portal, too? Where might it lead us? To a better place, probably our real world, where AI-assisted and manmade art coexist happily and without conflict? It feels like this image has a hammer of a story to tell! This story unfolds through the juxtaposition of the piece's two major elements, and through the inclusion of the artwork in the "Empty Rooms" series. And the journey has just begun...