Machine Learning: AI art - Harper's BAZAAR April 2023
The still-life photographer Anna Pogossova talks to Alexandra English about venturing into the world of AI
Anna Pogossova is a human artist who creates surrealist pop sci-fi images using photography, 3D imaging, virtual reality (VR) sculpting and digital manipulation. Her latest collaborator, Dall-E 2, is an artificial intelligence (AI) image generator that creates strange scenes by responding to her text prompts.
Pogossova was born in Moscow and migrated to Sydney when she was 10, and Dall-E was created by its parent company, OpenAI, in 2021, with the updated version, Dall-E 2, launched in late 2022. Pogossova and Dall-E 2 are worlds and species apart but have more in common than you’d expect of a sentient being and a soulless piece of technology. For Pogossova, her artworks are all about drawing upon the references she’s gathered in her thirtysomething years on earth as a consumer of pop culture and a student of fine art. To assist her, Dall-E 2 draws upon hundreds of millions of images from the internet — the keeper of pop culture and fine art — that its programmers have trained it to recognise.
Pogossova hasn’t always worked this way. Her first foray into art was trying to recreate a drawing of an animal she’d seen in an exhibition in Moscow. “I remember going home and trying to recreate it and feeling a really frustrating disconnect between what was in my mind and what I drew,” she says. The frustration between what she imagined and what she could create using traditional artistic mediums followed her throughout her studies and career, ultimately leading to her exploring technology-based mediums. “I think [trying to draw that animal] was the initial spark — what I really needed as a child was the AI,” she says with a laugh. “It’s always just been about: there’s something in my mind’s eye, and I just want to show people what it is.”
She was introduced to film photography as an artistic medium in her second year of a Bachelor of Fine Arts at Sydney’s College of Fine Arts. “Growing up, not a lot of people in Russia had a camera — you would have a designated person in your family who documents the family, and ours was my grandfather,” she says. “I never really thought of photography as a means of making something visual; it was a means of [documenting] memories.”
From the start, Pogossova preferred still-life photography to portraiture. “There’s always some hidden story you can tell through objects,” she says. “They are highly symbolic. They can trigger certain associations and tell their own stories, but it’s not always super literal or prescriptive.” How we decode these objects, Pogossova says, is through an unspoken but “culturally agreed-upon visual language”, which is stored in what she refers to as the “Collective Brain” or a “Cultural Cauldron” — it’s where all our pop culture references are, our awareness of shapes, colours, signs and symbols that we’ve absorbed by osmosis as we’ve moved through the world. It’s that part of your brain that tells you, That movie is sci-fi, when you look at a poster.
Pogossova’s work is hard to define: she plays with colour and scale in ways that are at once purely delightful and yet also strange. She’s interested in taking elements from the Collective Brain and pushing them to near obscurity. “I try to push [these cultural codes], abstract them as far as I can … I take away some things, put some things back together and put it back in this Cultural Cauldron, this soup of references, and then create an image from there that doesn’t have any of the original parts to see what happens. The
amazing thing I’ve found so far is that it will still remind people of something — it’s this uncanny feeling of I know this, and I don’t know why. It’s not about finding an answer; it’s more like, I wonder if my inner world is the same for everybody else, and more and more, I find that it is.”
She is also incredibly physical in creating object-based work. For her photography, especially editorial work such as “Size It Up”, which she created for the June/July 2022 issue of Harper’s BAZAAR, she balances, glues, suspends things with invisible fishing lines and throws objects in the air to distort reality and scale. Still, some objects in her mind’s eye don’t exist in the real world, and so she’s had to come up with innovative ways to create them. Enter, VR sculpture. “That began in lockdown to escape my space,” she says. “I was using VR for anything that would entertain me, and then I found out it was an amazing tool for making art.” She describes it as like being able to walk around in Photoshop, with all the tools you need sitting on a shelf. You pull down what you need and use your hands (with sensor gloves) to pull and push shapes as if you were sculpting with clay.
And then along came Dall-E 2, the AI text-to-image generator and Pogossova’s “dream machine”. A portmanteau of Disney Pixar’s character Wall-E and Salvador Dalí, Dall-E 2 has been designed to mimic how the neurons in our human brains signal each other while learning to recognise patterns and images. As Mark Chen, the lead researcher on DALL-E, explained to The Atlantic, “You can imagine a very small child that you’re showing a lot of flashcards to, and each of these flashcards has an image and a caption on it. … So it forms these associations, and then kind of builds its own kind of language for basically representing language and images, and then is able to translate that into images.”
Pogossova worked with Dall-E 2 to create the artwork for “Love & Other Drugs” in the February 2023 issue of BAZAAR. “I wanted to try out AI because it essentially does the same thing as me,” she says. “It has access to the cultural reference soup, so me and the machine, we’re essentially drawing from the same world of collective visual language. I tested a few things based on my own work because I wanted to see if we would come up with the same thing, and there were a lot of similarities, so I was like, Great, the Collective Brain is a solid concept. It works. We’re all tapping into the references we have.”
She is aware that it’s a controversial artistic medium. Some artists fear AI image generators will render human artists surplus to requirements, while some critics question its validity as an artistic medium. It’s going to replace art, but it’s not art. “I think people get hung up on it feeling unfair, like Why did I learn this craft? And now there’s this thing where you can jump on and type a few words?” she says. “But it comes down to ideas. If you’re just relying on technical craft to set yourself apart from others, that’s not a sustainable practice, but if you have the talent and you’re just lacking some technical skills, why not? I think it’s great that it’s making art more democratic.
“I think to show up with the pitchforks straight away and run it out of town is not the way forward. We have to ask ourselves, What are we hanging onto? AI is a different journey, and I don’t think it’s less valid by any means. I think it just destabilises a lot of ideas of what we consider to be authentic and valid artwork.”
This article originally appeared in the April 2023 issue of Harper’s BAZAAR.
