It wasn’t meant to happen like this. Yes, the robots were always going to come for everyone’s jobs, but it was the menial ones that were set to go first. Freed from the need to fill out spreadsheets and perform administrative duties, we were all supposed to have extra time to indulge in more creative, fulfilling pursuits. Yet Microsoft Excel still exists while AI algorithms are producing works of art that are both commercially viable and critically respected.
An AI artist, Jason Allen, recently caused outrage among old-school digital artists by winning a digital art competition. One of the writers of US publication The Atlantic, Charlie Warzel, provoked the ire of illustrators around the world by choosing to adorn an article about controversial radio host Alex Jones with an AI-generated caricature as opposed to using a stock photo or commissioning a portrait.
Three new products – OpenAI’s Dall-E 2, Midjourney and the truly open-source Stable Diffusion – have transformed the way we see creativity since their launches this year. Given a simple text prompt, they can create remarkably accurate visuals in an infinite number of styles. From superimposing Kermit the Frog on to some of cinema’s best-known images to imagining an expanded version of Johannes Vermeer’s Girl with a Pearl Earring, they are astounding visual playgrounds.
Nando Costa, director of Google’s design outreach team, is optimistic about the creative opportunities being opened up by these new technologies. He argues that tools such as Dall-E 2 and Midjourney will “change the way we do everything across creative production”.
But it’s unclear whether anyone truly grasps the significance of these advances at such an early stage. Discussing AI art on their podcast, My First Million, internet entrepreneurs Sam Parr and Shaan Puri compared the arrival of these technologies to passing a purple unicorn on the road and not stopping to make sense of what you have seen. ”Dream” or “generation” engines, as they called the likes of Dall-E 2, are indeed something we can recognise as special without necessarily being able to understand their implications.
If the output is good enough for both expert eyes and the mainstream media, their application for commercial purposes should be straightforward, right?
For some design, brand and PR agencies, the potential is clear. “This is the future of creativity,” argues Melanie Commey, creative director at Bound to Prosper. “It gives non-designers a new opportunity to create and blend concepts, offering the capacity to generate imagery in multiple styles.”
Could this effectively reduce time spent on design or reduce the manpower needed? “This is dependent on how the technology is integrated in agencies over time,” Commey says. “But it definitely has the ability to change the nature of creative work as we move to a more virtually focused realm.”
For other players in the space, AI can serve as a well of inspiration – at least, that’s how Diagram is attempting to use it. The firm’s AI-based systems aid automation via design tools such as Figma.
“AI augments and enhances human creativity,” says Diagram’s founder and CEO, Jordan Singer. “Our AI will extend designers’ existing creativity and embed itself in their workflows to allow them to work faster and iterate on ideas in tandem with the team.”
After all, the impressive outputs from these AI art generators are possible only because human artists and photographers originally perfected each of the styles and subjects depicted. In that sense, these image-generation toolkits are useful sieves, enabling us to explore recombined versions of what we’ve already seen. They can become an aid, just as libraries and Wikipedia help writers with their research. But these technologies are clearly not without controversy.
“Every tool is permitted and AI is happening one way or another,” observes David O’Reilly, an award-winning artist and animator. “But this species derives its entire value from the creative work of uncredited and unwilling participants.”
AI algorithms are trained on images and accompanying texts from across the internet. For many people, this root data belongs to our shared human heritage. And a lot of it would originally be the work of individual artists. Yet, when you use Midjourney or Stable Diffusion to illustrate an article, there is no credit awarded to the creatives who originally fuelled the training data. This makes many artists, even those working in the digital sphere, wary. There’s a difference between having your job taken away by a machine that’s simply better at making art and being usurped by one that uses material stolen from you and countless other artists.
The speed and relatively low cost of generating high-quality outputs is the most immediately troubling aspect for illustrators and other creatives. For editors and marketers, it’s a cheap option for bespoke content. For freelance illustrators, it might mean industrial-scale redundancy in real time.
Nonetheless, the opportunities for streamlining design briefs, trying out new ideas and replacing dull stock imagery will make life distinctly easier for many in the profession. But it is unclear whether the world of visual creativity will necessarily be richer for this new technology in the long term. After all, even the text prompts themselves can be automated.
“The question is no longer whether a computer can make art, but whether a human is needed at all,” O’Reilly says. “AI is likely to lead to unimaginably beautiful art and powerful affirmations of humanity, as well as widespread job disruption and the atrophying of fundamental human skills.”
The question remains: will that outcome be worth it?