Can AI “Think Different”?

Back in ‘97, Apple’s market share was evaporating. 

Major investors and media companies had written off its products as “toys” (used only by dizzy dreamers like Steven Spielberg), revenue and stock price were in freefall, and the company was dangerously close to bankruptcy.

As part of a sweeping initiative to right the ship, Steve Jobs & Co. solicited pitches from a handful of the world’s best ad agencies. Among them was TBWA\Chiat\Day, helmed at the time by creative powerhouses Lee Clow and Rob Siltanen.

The task, in Rob’s words, was simple: In a time where public perception was shaky at best, this campaign needed to “show the world that Apple is as strong as a lion.”

With spirits high and the stakes higher, Rob’s team got crackin’. As he recollects:

The next week we gathered in a large conference room at the agency where everyone had their work tacked up on wallboards. The room was filled with photos, pencil sketches, rough ideas and taglines. You know that scene in the movie “A Beautiful Mind” where the room is plastered with paper on every inch of wall space? Well, during a new business pitch or preparation for a big project, our conference rooms typically looked like that.

This wasn’t clinical, linear, precise progress. It was chaotic creativity. Eye-twitching, earth-shaking, “By-Jove-I’ve-Got-It!” human madness.

It’s that very madness, paired with a masterstroke of genius from Art Director Craig Tanimoto, that gave birth to one of the most iconic taglines of all time: Think Different. 

After a few campaign revisions, strokes of luck, and gentle words of encouragement (ha-ha) from Jobs, the concept played a central role in Apple’s astonishing turnaround, going on to win the Emmy for Outstanding Commercial. Within years, Apple was indeed strong as a lion.

So what if ChatGPT was around back then, and the lovely folks at TBWA\Chiat\Day had started by feeding their problem to AI?

Explosion and Cooling

Before we give the robot’s gears a chance to whir, let’s start with a foundational question: How does human creativity work?

Essentially, it can be divided into two phases:

• The divergent phase, where dozens of ideas explode into existence without judgment; what a friend of ours at Magneti might call “Letting your feeler feel.”
• The convergent phase, where bad ideas are ruthlessly tossed aside as we isolate the best one; what our friend might call “Letting your thinker think.”

If we imagine human creativity as a river, it might look something like this:

Much like the conference rooms at TBWA\Chiat\Day, the brain begins with lightly structured chaos. A problem has been defined, and our neurons start to chart the limitless paths to solving it.

There’s an explosion, then there’s cooling. Explosion, as thousands of ideas (most of them rubbish) leap to life. Cooling, as 99% of those ideas dissipate and only the most energetic and promising survive.

Of course, higher quantity doesn’t always mean higher quality. But there’s a strong correlation between the amount of time you invest in the divergent phase, whipping up ideas without judgment, and the strength of your winning idea.

Richly textured ideas rarely emerge spontaneously. They come from conference rooms plastered with dozens of scribbles, pictures, and posters — prolific minds, hard at work.

The Temptation of AI

AI makes it easy to make things easy.

With very little critical thought, our strategies, creative campaigns, pitches, and jokes can write themselves. When we put our faith in GenAI, we’re giving it (and ourselves) permission to cut down on:

• 1. The amount of time spent in the divergent phase of creativity
  2. The number of ideas generated from this phase
  3. The range of those ideas, as they are limited by the inputs received by the AI and the universe of information it can access

In other words, AI curtails the divergent phase of thought by instantly providing the foundation, the anchor point, for your creative process. Clamped by the jaws of AI, your river of creativity now looks something like this:

The consequence?

A significant number of solutions — and quite possibly, the most elegant and efficient — are never given the chance to exist.

AI’s Elephant

Even if you take the time (like any marketing leader should) to assess your AI’s output, to prompt and re-prompt, to engage thoughtfully, you’re still unavoidably tethered to the initial output.

It’s like not thinking of a purple elephant. AI has saddled you with a foundation of ideas, and it’s impossible to break free from that foundation into the realm of truly original (or at least, 99% original) ideas.

The pull of AI is always toward homogeneity. And homogeneity is hardly ever great.

If we aren’t disciplined enough to endure the chaos early, we won’t be rewarded with the next revolutionary idea. We won’t, in a very literal sense, be able to think different.

A Not-So-Scientific Experiment That More or Less Proves My Point

A few weeks ago, I offered the following prompt to ChatGPT-4o and Claude 3 Opus:

Imagine you’re the CEO of a tech company. Your company is struggling. It has minuscule market share and many major investors and media companies don’t take it seriously. You need to create a slogan that captures the nation’s attention, reminds them of why your company is so cool and revolutionary, and primes your company to make some serious sales. Pitch me 10 options for the slogan.

No, it’s not perfect. I didn’t give it additional context, I didn’t tinker with the prompt to achieve a better output, I didn’t give the AI 100 attempts. But I gave it enough to get the conversation going — to begin the creative process.

Here’s what I got:

Outputs from ChatGPT-4o:

• “Unleashing Tomorrow, Today”
• “Revolutionizing the Digital Age”
• “Where Innovation Meets Reality”
• “Experience the Future Now”
• “Your Gateway to the New Millennium”
• “Pioneering the Next Wave of Technology”
• “Empowering Your Digital Journey”
• “Transforming Ideas into Innovation”
• “Join the Tech Revolution”
• “Tomorrow’s Technology, Today’s Solutions”

Outputs from Claude 3 Opus:

• “Igniting the Digital Revolution”
• “The Future is Now”
• “Think Different”
• “Empowering a New Generation”
• “Technology with Soul”
• “Imagine the Possibilities”
• “Redefining What’s Possible”
• “Small Company, Huge Vision”
• “Simplicity Reimagined”
• “Why Follow When You Can Lead?”

Unsurprisingly, the usual suspects reared their ugly heads: unleash, empower, ignite, transform. If you’re like me, you probably thought something like: “Eh, not great, but it’s a start. Let’s keep trying.”

Innocent enough, right? But the damage has already been done. The foundation has been laid. The elephant has entered the room.

The rest of your creative process will be irrevocably tethered to these uninspired offerings, no matter how hard you try to banish them or how diligently you evolve them.

But there’s a deeper irony at play. Option #3 from Claude rings a bell, doesn’t it? When I asked Claude how it came up with that idea, it responded in part:

“Think Different” was Apple’s iconic advertising slogan from 1997 to 2002, created when the company was in a similar situation to what you described. Apple had low market share, was struggling financially, and needed a bold new vision.

Really, the AI played its part perfectly. It was able to accurately synthesize my input, map it onto a prototypical example of that same situation, and offer an intriguing solution that was brim-full of creative potential.

There was just one problem: Apple, uh, already did it.

Think Different(ly)

There’s another piece where we can explore the many (and there are many!) advantages of using GenAI in marketing. Here, though, I leave you with a word of caution.

AI is a whip-smart thinker, but there’s one thing it’s truly terrible at (which happily seems to be one area where humans excel): creating chaos.

And chaos, lovingly corralled by the human mind, is the fountainhead of creativity.

To chaos —