I’m bullish on AI-generative tools and their ability to elevate and enhance creativity.
I’m not all that certain that’s where we’re headed, however.
Since the dawn of the Internet we have rushed headlong towards an approach to marketing and creative output that rewards commoditization and familiarity and certainty. This is why the Internet is a direct marketing extravaganza and less so a brand-building environment.
We are addicted to certainty. We yearn to know “it”—the idea, its craft, will work—especially when it means paying for Art to motivate behavior. Alas, Art is often too, well, too something that's not predictable, controllable and scalable.
So technology (i.e. clicks, attribution, measurement) and user experience (i.e. fewer, repeatable templates) seek to deliver assurance. And for the past 30 years that technical assurance has focused on content delivery—its formats, specs, frequencies. If we could automate resizing, we did. If we could separate component pieces (background color, headlines, feature image, button sizes, button copy, etc.) and let the robots test, optimize and assemble best-performing combinations, we did.
None of this is bad, by the way. A better Photoshop is a better tool. We’ve learned a lot about how the delivery of ideas creates impact. Or not.
Now one of the questions posed by the latest blend of artificial intelligence is: Can technology automate content expression? Can AI generate ideas and generate craft which contains greater degrees of certainty than those generated by humans?
The Ad Turing Test
The march towards certainty has begun.
I’m pretty sure the image below came from this week’s 2023 BrXnd AI Conference, hosted by
. This looks like the event’s Ad Turing Test—which appears horrifying, mischievous, and brilliant. The premise: Could you create a print/static ad using only AI tools (nothing else!) that a human judge might mistake as human-made?Ad Turing Test results aren’t yet posted, but
and Rafa Jiménez (of Seenapse fame) documented their only-using-AI-to-make-an-ad process in this post aptly quoting David Ogilvy. (And BrXnd posted some AI-generative work this morning.) The task appears complex, frustrating, and the end result—I mean, I suppose this looks ad-like?I take cold comfort in the complexity and relative poor creativity of these AI-generative ads—because the coding will only get better, and, one assumes, the end result will, too. After all, we mocked the fingers in Midjourney back in January; not so much anymore.
As legendary graphic designer Neville Brody put it recently, “AI is so sophisticated: it's all about virtuosity and showing how clever it is. It can do super sophisticated deep fake content but can't do the subtle touch, so I'm interested in embracing something very reductive, minimalist, and abstract. It also can't do lateral thinking very well – it will inevitably at some point.”
It will inevitably.
Can we automate authenticity?
This all sort of feels like a Kurt Vonnegut novel or Black Mirror episode unfolding. Or maybe it’s just the news.
As
put it yesterday, bolding mine:…The jaw-dropping speed of generative AI’s embrace is essentially a large-scale acknowledgement that modern life is sort of miserable and that most people don’t actually care if anything works anymore. I mean, at this point, pretty much everyone I know in an office job that isn’t in media is using ChatGPT at work basically all of the time. But as more companies push to integrate themselves into AI platforms, it’s also revealing that they don’t really care either. And we know this because there is simply no way you can say you care about something if you replace it with AI. What you care about is speed, scale, and, if this stuff works correctly, money.
When I say I’m bullish on AI-generative tools to elevate and enhance creativity I mean I care to create something authentic in a new media. Something you couldn’t generate otherwise. I’m a drummer, and I think electronic drums are amazing—for producing sounds that aren’t just samples of acoustic drums.
In mere months we’ve shattered the myth of Art-making which has been around for eons. Now you don’t need to understand how a paint brush works, or why film stocks function, or have experience in graphic design. You just need to know how to type words to create what look like paintings, street photography and packaging. This is stunning. Marvelous. And scary.
On the one hand this is incredibly freeing. More people can now express themselves more broadly. On the other, there was that whole desktop publishing revolution and people still use the Papyrus font without irony.
I have no doubt the Ad Maker 3000™ will emerge soon enough, guaranteeing creative certainty in concept and craft. And it might actually work.
Until then, I think we need to keep pushing the boundaries of AI creativity past the boring and inevitable low bar of certainty. Humans deserve better.