Anyone who does creative work knows the Muse can be elusive. The Greeks conceived of her as a goddess (nine goddesses, actually) for good reason: Inspiration feels like a magical visitation from the spiritual realm.
While scientists study how we achieve breakthroughs and develop new ideas, they’ve yet to develop a panacea for writer’s block... or a bad day at the office.
The human brain needs help, and there’s no shame in admitting that. And while many creatives have rituals to jumpstart their processes—a walk around the block, a conversation with a trusted friend or mentor, or a new experience, for example—technology can also offer a boost.
The advent of AI offers artists and marketers alike another resource for thinking through their ideas, testing them on audiences, and executing their concepts.
And different creatives are, of course, using AI in infinitely creative ways. As with any tool, AI can conform to its user’s vision and needs.
We spoke to a digital agency CEO, a fine artist, and a Senior Art Director who all use AI to enhance their creativity and streamline their work. Each of their projects retains indelible impressions of the people who developed them, and the unique language with which they’ve prompted this new intelligence: their own songs to the spirit, one might say.
Michael Treff, CEO of technology-first creative agency Code and Theory, noted that he and his team—comprised of roughly 50% creatives, and 50% engineers—have used some form of artificial intelligence for years. Machine learning and algorithms have long contributed to company strategy, though the recent boom of LLMs marked a turning point.
Treff and his team are among the first to create “AI personas” that can approximate the responses of a target audience. “In the concepting phase, we run our creative against those personas to get good insights, and good feedback. It’s remarkable how accurate and helpful that can be,” he says.
When the agency worked with Tipico, a sports betting and casino platform, they wanted to appeal to specific demographics in one particular state. “Being a Cavaliers fan in Cleveland is different from being an Ohio State fan in Columbus, which is different from being a Galaxy baseball fan in Cincinnati,” Treff explains. “We fed these AI brains and created personas of each segment.”
Under Code and Theory’s direction, AI was trained to absorb Reddit forums, Ohio meme accounts, and social media platforms to create personas of amateur and experienced bettors. Treff’s team gained insights from the reactions that the resulting AI personas had, which were then used to re-optimize their creative before it went launched. The result? In the month post-campaign, Tipico experienced over 40,000 registrations, representing 57% of their annual target.
AI bolsters other stages of Code and Theory’s creative process as well. It helps the team better co-ideate; aids in versioning creative assets; and offers new, personalized possibilities.
For a recent partnership with RealClearPolitics, Code and Theory built ContextLens, which uses AI to instantly generate insights and visualizations from polling data. In this case, AI created a brand-new user experience—bespoke to each visitor.
Code and Theory is now re-building its own website using AI. Thanks to a variety of “personalization triggers,” the site will learn about individual users and respond to their specific needs in real time.
Where does Treff see AI moving in the coming year? “You’re going to see AI-driven interfaces,” he offers. “Sites, apps, and software will be really ephemeral, tailored to each person.” In other words, content will be oriented around users’ specific desires and behaviors. With the help of human input—from both the initial coding and concept side and the user’s side—a single interface will become endlessly generative.
Treff compared AI to Photoshop, which upended our relationship with photography. AI is an enabler of creativity, not a replacement for human ingenuity: He believes that the future of creativity is “humans plus AI.”
“In the arc of history, technology is undefeated,” Treff concludes. “Those who adapt [to AI] are going to flourish and blossom and do incredible things that none of us could even think of. And those that don’t will have a really hard time.”
“AI makes accessible the heretofore unimaginable,” pronounces Kenny Schachter, a writer and artist who's been quick to latch onto these tools to create new work. Employing AI feels natural to Schachter, who has also been a steady proponent of other bleeding-edge tech, like NFTs.
Schachter has experimented with LLMs and similar tools, energized by the audio, video, and written outputs he’s conjured. He created an artificial voice that emulates that of contemporary artist Jeff Koons. AI has allowed him to integrate special effects, which otherwise were not in his budget, into his digital films. And he’s even used AI in a conceptual way, to write an article about an art fair he did not attend.
This March, Schachter will open an interdisciplinary exhibition at Jupiter Gallery in New York. The title, The Painting in the Age of Robotic Reproduction, is a riff on Walter Benjamin’s famous 1935 essay “The Work of Art in the Age of Mechanical Reproduction.” The show will feature digital art including photographs (some manipulated and some not) and sculptures with paint applied by a robot to the surface of a cube, a sneaker, and a canvas. Schachter said he’s not able to paint himself: New technology, like AI and robotics, allows him to work in the medium.
“Artists and creatives are opportunistic in the best sense of the word,” Schachter says. “A community of contemporary artists is compelled to utilize what’s around them to make art. I’d be remiss not to grab every tool around me."
Schachter also noted how AI has helped those in adjacent creative fields, like architects. The late Zaha Hadid was a friend and collaborator of his. Hadid’s firm is now under the direction of Patrik Schumacher, who uses AI to generate design ideas and aid most projects. “He compels his architects to use it,” Schachter said. “Why? Because he has an open mind.” (Meanwhile, the firm’s revenues have notably increased.)
Like Michael Treff, Schachter views AI as one of many technological advances that can help him achieve his own singular ambitions.
“There are these new innovations and I look at them, I'm interested in them, and then I find ways to make them work towards the manifestation of my ideas,” he says.
Pegah Yazd’s entry to AI was personal. She began using DALL-E, an AI text-to-image model, for creative exploration and play. Now, she integrates AI into her day-to-day work at Stagwell Marketing Cloud.
Yazd is a Senior Art Director and the sole designer on her team; she needs help creating new illustrations for articles on the company blog (including the one that illustrates the article you’re reading right now). AI lends Yazd a hand. It cuts down the time to generate an illustration, offers her a place to create a “mood board,” and allows her to easily visualize how her ideas might be actualized.
"I’ll have a rough sketch in my head of what something should look like, and I need to see it drawn out, with some styling, to give the team an idea of what I’m envisioning,” Yazd says. “It would be really time consuming to create the asset myself.”
Instead of spending hours or days creating an illustration, Yazd can concentrate on the concepting phase. This process is nebulous, and very human: It features the associative, abstract thinking that AI still struggles with.
In her professional life, AI is now smoothly integrated into Yazd's instinctive creative process. She reads the draft of an upcoming piece of content, considers its themes, and tries to find a visual metaphor to represent them.
Once she’s landed on something fruitful, she feeds an AI platform like Midjourney with specific stylistic direction in order to generate a unique, high-quality image that avoids cliche, anything that might resemble stock imagery, and the “uncanny valley” ick that can result from poor prompt engineering. Yazd can then sift through outputs and refine both her ideas and the resulting images using more traditional digital-editing platforms..
Yazd predicts that in the future, AI will develop its technical abilities and work out its glitches (it’s still notoriously bad at depicting fingers, though better than it was 6 months ago).
Regardless, she says, a human-in-the-loop will remain vital. “You need a human doing really creative thinking, and who also knows how to use the AI tools to get good results,” she says. It's a reminder of a maxim that has now become a mantra: AI won't replace you and your job, but humans using AI might.
