POV: AI-generated album covers prioritise virality over creativity

POV is a new column written by It’s Nice That’s in-house Insights department. Published fortnightly, it shares perspectives currently stirring conversation across the creative industry. POV digs deeper into industry discussions and visual trends, informed and inspired by creatives we write about. To uncover visual trends and insights from within the global creative community through our Insights department, get in touch here.

There is a certain kind of lore that surrounds visual artists who work in the music world. It’s the iconic pairing every creative yearns for: an opportunity to translate the world of another artist and give it new meaning. There are countless examples throughout design history. Reid Miles’ era-defining typographic work for Blue Note Records. Stanley Donwood famously paints alongside Radiohead as they write and record. Peter Saville even once told me, while standing in front of a body of his own work, that its influence “has spread through everything”.

It’s understandable, then, that record labels’ decision to opt for AI-generated artworks has divided opinion amongst artists and listeners. Jon Rafman’s cover for Lil Yachty’s Let’s Start Here left fans debating whether it was genius or detrimental. While Nicki Minaj’s AI-generated imagery, in promotion of her 2023 album Pink Friday 2, led to a series of bandwagon-jumping memes. The latter has seemingly piqued the interest of virality-hunting music execs prioritising strategy over creativity, leaving the role of the designer, photographer, art director and so on, in limbo.

Although artificial intelligence’s capabilities divide opinion no matter the industry in question, scrutiny appears to intensify when it comes to visuals related to music. Perhaps there is no home for “artificial” visuals in an artform which audiences form deep emotional attachments to. Or are we focusing on the drawbacks of this technology due to its increasingly unimaginative widespread use?

It’s a question that sparked discussion amongst designers early last week, led by the Kentucky-based artist and musician Robert Beatty. Recognised for his psychedelic-leaning work for the likes of Oneohtrix Point Never, Bedouine and Tame Impala, Robert explained how seeing musicians choose AI over human collaborators is not only disheartening, but disappointing in its execution. He isn’t wholly against working with such tools, though. As his original Tweet succinctly put it: “I’m not anti-AI, I’m anti-shitty artwork.” However, in the landscape of generated works, “it’s hard not to see music videos or album covers that are made with AI and feel like it’s artists or labels trying to save money and cut corners,” Robert tells us. “So much of the work I’ve seen in the music world that uses generative AI feels tossed off or unfinished. It feels like a quick shortcut to generate content and nothing more than that.”

At Beggars Group in London – a group of independent labels including 4AD, XL, Rough Trade Records, Matador and Young – global head of creative Alison Fielding spends her days building out aesthetic worlds for album campaigns. The look and feel of records is “always led by conversations with the artist,” she says. At times this can lead to Dry Cleaning’s Grammy-award-winning pubes by Rottingdean Bazaar and Annie Collinge, through to David Rudnick’s landscapes for Black Midi. “It’s very important an artist is represented in a way they’re happy with,” Alison tells us. “It’s their music, after all.” With this ethos in mind, Alison and her team’s creative commissioning expands to AI too, but with understandable hesitation.

For Alison, there are potential positives to incorporating machine learning into the artistic process behind an album’s campaign. When ideating proposals, it can visualise tests to show an artist, and there are, of course, financial benefits to not hiring expensive location shoots. There are also numerous artists creating boundary-pushing work in collaboration with machine learning. Beggars’ own Holly Herndon incorporates AI into her compositions. There are also many inspiring artists “in the immersive, interactive and installation worlds,” notes Alison, highlighting Sougwen Chung’s painting robots as a personal favourite she’d love to collaborate with. But such experimental cases are rare. “Often, the process creates these soulless and highly polished images which, to me, look very flat and lifeless,” she says. “The audience is not as connected or engaged. It’s formulaic in a way, and not as fresh nor exciting as it was.”

Like any visual field, certain methods of communicating have their moments in the spotlight, causing the zeitgeist to form around them with varying degrees of success. Easy-to-access AI tools also lead to the assumption that this technology can make anyone an “artist”. But without narrative, or the understanding an art director or designer brings to a project, a record sleeve where generative visuals are the entire concept just appears inanimate. As Robert puts it: “Part of what interests me about art in the first place is seeing people work within a set of limitations,” he says. “How those limitations (whether being forced to learn a new skill or doing something in a short period of time) inform the work is a huge part of what makes things ‘good’ to me. It would be way more interesting, to me, to see what a human could do with a set of constraints, than trying to evade the constraints by using a shortcut.”

This sentiment is mirrored in the reaction to several AI-generated releases in recent months. In fact, even when this technology is incorporated into an artist’s process, it can have a divisive effect. For example, one reference Robert mentions as “especially revealing” is Tony Oursler’s music video for Beth Gibbons’ recent release, Floating On A Moment. “If one of the most famous video installation artists alive can’t get away with using this stuff, who can?” he asks.

Ultimately, for many audiences, the use of AI dispels the magic that listeners are used to feeling towards a new release. “It’s the application of it that’s key,” adds Alison. “You need to step back from it. Ask if it’s actually good, or am I just excited by a new tool? Because when it’s done well, with thought and integrity, it can be incredibly effective and engaging.” By and large, true resonance lies in sincere intention and audience connection. “In general, people will always be drawn to work because of the people who made it,” says Robert. “If an artist isn’t putting a piece of themselves into the work, why should anyone care?”