If you’ve been following the hype about the developments in artificial intelligence (AI) image-generation over the past few months, you might be wondering whether it’s the end of the road for artists. That’s the question many journalists have been posing.
If this issue hasn’t been on your radar, here’s a quick rundown. For years machine learning systems have been scraping through all the images they can find online whilst recording the accompanying text descriptions to create datasets. This process trains the AI. It has become so sophisticated that the systems have learned to associate words and phrases with certain types of pictures. We can now type any words into software tools like Dall-E 2 and Stable Diffusion and generate weird and wonderful imagery, a composite of all the images in the dataset. Here’s a two minute easy explainer of how this type of software works:
The surrealists of old would have had a field day with the combinations. A dog in sunglasses on Mars eating a doughnut. The Beatles playing lego with a polar bear in drag. It’s fun to play with, and you never get the same image twice. The results are not perfect because AI image tools still struggle with rendering faces and body parts that actually look human. But the tech is getting better all the time.
So should artists be worried about these AI image-generating technologies? Just this week the European market editor at Artnet News said ‘AI is already capable of mimicking human creativity. Whether or not it makes artists obsolete will be down to how they use it’. Already I’m thinking what nonsense. She goes on: ‘The kind of artificial intelligence we might imagine replacing artists - an entirely autonomous creative robot capable of human-like imagination and expression - does not yet exist but it is coming’.
I appreciate how far and how quickly AI technologies have developed. And I can see how this technology as it gets more advanced might pose a problem for the livelihoods of illustrators, graphic designers, digital animators and photographers producing commercial work. As the robot Ai-Da said to a select committee of the British Parliament back in October, AI offers both opportunities and threats to artists. But there’s a difference between commercial creative output and fine art. Statements that suggest the possibility of artist obsolescence lack real understanding of the role of artists in our world.
Art is a representation of the artist’s ideas, thoughts, experiences, emotions. It prods and challenges us to search for meaning or involuntarily elicits feelings in us. Even art made entirely in the digital realm does this. Consider Beeple for example, who made his name in 2021 when one of his NFTs sold for $69 million. He produces digital images that are typically set in futuristic landscapes, the kind of imagery that could easily be generated by AI tools. Is it the end of the road for him? No way, because Beeple’s work is more than just the skill and imagination he brings to the production of his images.
As an artist Beeple invites us to contemplate meaning in his images. When I look at his work I think about the hyper-stylised fantasy digital imagery that is so dominant in our culture, the prevalence of online worlds and avatars. I think about the visual landscapes that my son occupies in the video games he plays every afternoon after school. And it makes me wonder about how we value art now. You might not be interested in Beeple’s work, you might argue that it is crude in meaning and utterly banal. But it’s still generating a response in you because you’re making judgements about what ‘good’ art is.
AI art is not art that functions in this capacity. It is imagery created by computers that are very good at recognising and replicating patterns. More relevant than the lazy question of whether artists will become obsolete because of AI, is how artists will make use of it, work with it, and interrogate the issues it raises. Machines are controlling our lives in fundamental ways and it’s the job of artists to comment on this and to use the technology to comment on the technology and its effects.
That’s already happening. Last month an art critic from The Guardian asked six leading contemporary artists to make art using AI, and the results were fascinating. One of the six, the 2012 Turner Prize winner Elizabeth Price, asked questions of the AI image-generating tool, and started to think about “how it was putting images together; how that process differed from the human mind; what it ‘knew’; what it ‘understood’; and how much we could think of its dataset and search modes as a kind of cultural memory.” Artworks that question the bias and test the limits of big data that is increasingly running our world? Now we’re really into the good stuff.
Other artists are developing their own AI datasets to think about how systems function in our world, and the way we interpret information. The German artist Mario Klingemann, for example, recently created a delightful artwork exploring how much meaning you could get from a random selection of words churned out by the machine. For Appropriate Response Klingemann developed a dataset of quotes and aphorisms that he had sourced from the internet to train the AI so that it would generate composite phrases in a similar style. Viewers knelt in front of a board and waited for the machine to speak to them, like some sort of religious experience. Even though they were generated randomly by the machine, the effect of these phrases on viewers was often quite powerful. Here’s Klingemann describing the project.
AI is a fascinating phenomenon. We might be aware of some of AI’s manifestations, but there is plenty going on behind the scenes, out of sight, entwined in global systems that are so complex that we can’t possibly understand them or the extent to which they reach.
And it raises a whole load of questions that we’re only really just starting to ask. Questions like how AI affects privacy, freedom and truth? To what extent our prejudices are amplified and reinforced through AI? Who is accountable for mistakes in systems run by AI?
Through their work artists make us confront and think about these issues. When we stop being curious and thoughtful about who we are, how we function and relate to each other in this world (not likely) we won’t need artists. Until then, there’s no danger of them becoming obsolete.
I’d love to know your thoughts on any of the ideas here, so please leave a comment below and tell me what you think.
Need Help Writing About Your Art?
Do you struggle to communicate your ideas in words? Does the thought of writing an artist statement make you want to pop your head in the sand? Would you like someone to help you tease out connections and think through the ideas you’re exploring? I can help you! As a writer and university lecturer I’ve been guiding artists on how to think through their ideas and put them into coherent words for over 15 years. Contact me for a focused and constructive 1-2-1 discussion about your art. Here’s a link to more info or reply to this email to discuss what you need and to see if we’re a good fit.
Minimal Wage: Do We Value the Artist's Process or Product?
I'm sharing this article by
, who writes a brilliant, thoughtful Substack about art, . Her article is about the time it takes to make art and how we assign value to it. Apart from perhaps the top 5% of artists, all other arts workers' time falls somewhere on the spectrum between unpaid and underpaid. When you work within a culture that tells you that your time holds precious little value, there is an unconscious instinct to offer more of it in order to provide greater value to the world. Read more here:
I don't think AI could ever replace or compete with human artists because it lacks the most important aspect of an artist - their vision - which requires a unique personality and soul.
Within the VFX film world AI is being used quite extensively in 'concept' art. Some companies have an entire department dedicated to it as it may point the way to a different approach to filmmaking - I'm talking the sci-fi , fantasy world more than yer basic drama obviously. Maybe even taking over the process at some juncture.It requires a different skill set -of course, need art direction but execution maybe a different story.