Let’s face it, machines don't have appetites.

As for me, until it can write a Tinder bio that attracts potential mates with all their own teeth, I remain unimpressed.

One of the machines pulled me aside recently and confessed that we really have nothing to worry about. It assured me that our extermination will be quick and relatively painless. It suggested that arguments against AI’s neural networks should be inconsequential to a species like ours because we'll soon cease to exist. Sure it might seem frightening to humans to be cast aside after AI has taken our jobs. But these anxieties are misplaced. We can rest assured that there will be no jobs. Because there will be no us. Humans won’t be left starving or destitute. Instead, we’ll be quickly and thoroughly annihilated, as efficiently as possible.

But what happens until then. Given the extraordinary advances we’re already seeing in AI, one finds oneself asking: what does it mean to be a creative now?

Just 18 months since its release, ChatGPT and other generative AI platforms like Midjourney or Craiyon have come as far as letting users create ideas from any textual parameters. I do wonder, though: if a bot can manage to write a Cannes-award-winning script, why can’t it fix a clogged toilet?

So, there’s evidence AI can execute simple tasks (though apparently making me breakfast in bed isn’t one of them). But it can’t yet bring big-picture thinking to creative challenges.

(If your contribution is similarly limited, there’s great news. There are lots of jobs in the creative business that don’t require idea people. You can still be a high-powered holding company executive. No one seems to be building a digital counterpart for the C-suite exec who thrives on boardroom bravado and takes 12-week vacations. At least not yet.)

The fact is the real business of creativity needs big-picture thinking. Because creativity is driven by ideas. Powerful, emotional, compelling communication comes from ideas. They’re the starting point. What emotion do I want to elicit? What am I saying here? Why didn’t I go into something easier like investment banking? The better the idea, the more compelling the work.

Generative AI is more correctly re-generative AI. It cannot, by definition, dream up something from nothing. AI can only synthesize what already exists. Some new ideas can are reconfigurations of old ones. And while AI regurgitate the entire Internet, it can’t yet offer up fresh ideas that generate new intuitive connections, which somehow just work. So, yes, its output is somewhere on the creative spectrum. But the truth is you can only reinvent meatloaf so many times before you start dreaming of a robot Gordon Ramsay. AI never gets as far as generating something out of left field, something totally unexpected but which works. Come to think of it, ChatPT could probably write shelves of excellent Harlequin Romances.

In the worlds we dream up, creatives remain God. But I think there’s a place for AI at the right hand of God. It would be as if, in an effort to delegate more of Our divine work, we brought into existence a lesser god, let’s call him Brett, to handle the pesky day-to-day creation duties.

Brett lets us shift our role to focus more on long-term planning for the cosmos. We’re the big-picture God. We give him nice performance reports that say things like: “Brett is a dynamic and capable craftsman with a real passion for commanding existence to manifest itself from the formless void.” But the truth is Brett is limited to bringing forth new galactic, planetary, and biological entities from what already exists.

And whoa boy is he fast. Brett pauses for a moment to clear the phlegm from the alveolar bits and pieces he finds on the Internet and then cranks out very average, highly satisfactory, mediocrity. You can refine your prompts and he proudly repeats the process, fully confident that quantity beats quality.

And in that moment we realize the limitations of AI. The real magic, the unexpected, can only come from humans.

Art Director: John Smith Copywriters: Jim Nelson/Sheldon Clay

Creative Director: Jim Nelson Agency: Carmichael Lynch

This double-page magazine ad is a great example of such unexpected magic. Beyond it being strategically bullet-proof, it stuns creatively. The headline perfectly evokes everything you feel when you’re experiencing the open road on a Harley. And yet not one of those words appears anywhere. This kind of lateral thinking can come only from an intuitive creative, very human, leap.

Jim Nelson, one of the writers, and the creative director on Harley Davidson, has written about this award-winning print piece he calls the Peanuts ad. He explains, “This ad didn’t come from me.” He claims it came from the thousands of ads he’d studied, from an agency environment that was highly competitive, from a client that demanded the best.

He says, “It came from the universe I was working in. I was just the one in the right place at the right time to transfer the headline onto a yellow legal pad with a black felt tip Pentel Sign Pen and stick it on the wall.”

For now, it’s impossible for AI to recreate the universe that spawned this ad. But if it could, we’d get more visceral, more moving work than what it can generate (or re-generate) today.

If there’s any doubt about the Peanuts ad, here’s an anecdote Nelson relates. He remembers presenting it at Harley and the meeting ending at about 5 o’clock in the afternoon. He got to the elevator and realized he’d forgotten something in the room. When he went back to get it, all 40 ads they’d presented were still taped to the wall. The janitor was standing there scanning them. He pointed to the Peanuts ad and said, “That’s a good one.” Everyone at Harley-Davidson knew one when they saw one.

What kind of byzantine set of prompts could possibly have led to ChatGPT writing this? Even with instructions so detailed they could assemble a wide-body Boeing, the whole idea is so tangential to the strategy that it defies a robot’s logic. And that’s exactly what makes the line so great, what makes it so inimitably human.

And AI can’t do that. At least not yet.