Artificial Intelligence (AI) is here and more accessible than ever, and while some are criticizing the material realities of this new technology and its impact on the lives of artists, it does feel somewhat inevitable that AI will be a tool that professionals and novices will be using in the years ahead. Is AI really that different than the innovation brought on by the printing press, photography, or even computers? Rather than thinking of AI as a way to sideline artists, I think it’s more likely that it helps creatives realize their visions with greater ease. Because, sure anyone can take a photo, but that doesn’t mean everyone can take a good photograph.
In contemporary art, art observers have long noticed how artists when they enter the land of blue-chip galleries start churning out increasingly similar and mundane pieces, often turning aesthetic innovations into branded objects for the very wealthy. So, I thought I’d try using press releases from five “blue-chip” (aka really expensive) art galleries in New York City — or at least as much of the press release that the AI system would allow me to input — and see how DALL-E (one of the most popular AI tools nowadays) reimagined the exhibitions using only those words.
So, without further delay, here are the results. The human-made works on are on the left, and DALL-E’s reimagined versions on the right, with some commentary. Is the future already here?
Rick Lowe’s Meditations on Social Sculpture at Gagosian gallery (541 West 24th Street, Chelsea, Manhattan)
I have to say, I think AI might win this one, though the DALL-E rendering definitely feels more like a concept drawing rather than a finished work. Lowe’s work has a digital feel to it, so maybe it’s a more natural fit for AI. The original work also has thick textures and patterns on its surface but I imagine texture is something AI will only get better at rendering.
Thomas Ruff: d.o.pe. at David Zwirner gallery (533 West 19th Street, Chelsea, Manhattan)
Artist Thomas Ruff seems perfectly suited for AI, and DALL-E’s version is quite convincing. If you told me the image on the right was a little-known series by Ruff, I’d believe you.
Jenny Holzer’s Demented Words at Hauser & Wirth gallery (542 West 22nd Street, Chelsea, Manhattan)
This example reveals a weakness in the current DALL-E system, which is text. The AI system renders all text as gibberish, which is curious, but then again Jenny Holzer’s text nowadays is becoming harder and harder to read. The image on the right could easily be mistaken for the work of Holzer, maybe one that “questions legibility and our ability to read without understanding …” Sorry, I couldn’t resist.
Jason Martin’s Vortex at Lisson Gallery (508 W 24th Street, Chelsea, Manhattan)
It’s curious that this was the only exhibition text that I cut and pasted into DALL-E that rendered people along with art on the wall. Perhaps it was the phrase “the relationship between painting and sculpture” from the press release that triggered it, but all four of the renderings did indeed have human figures, even if they were blurred or distorted. I think AI might also win this one, but maybe I’m just not a big Martin fan, though the smell of oil paint that permeated Lisson Gallery on West 24th Street is not something AI can reproduce, yet.
Jill Mulleady’s Bend Towards the Sun at Gladstone Gallery (530 West 21st Street, Chelsea, Manhattan)
Another creepy rendering that could very well be by the artist in question.
Ennui ✔️ Moody scene ✔️ Brushy rendering ✔️ Strong foreground figures ✔️ Backgrounds that create a sense of shallow space ✔️ Strange light ✔️
Pretty unnerving how much DALL-E gleaned from the text to make this image.
Should artists be worried? Nah, but I can’t wait to see how many will use this tool to transform the way we see and experience the world.