The relentless pursuit of imperfection

As AI continues to seep into our everyday lives in increasingly innocuous and inconspicuous ways, I find myself welcoming it and resisting it in equal measure. Recently, when the Wendy’s drive-through robot got our order 100% right, Dawn and I simultaneously said, “Well, I don’t hate that.” The Wendy-bot didn’t try to trick us into thinking it was human. It wasn’t bored or excited to take our order. It just performed. Perfectly.

Yet when I’m cruising LinkedIn for thought leadership, deleting inbound sales emails or reviewing a designer’s cover letter, I find myself actively seeking signs of human intelligence. Or more specifically, human fallibility. And, of course, that got me cogitating on the nature of perfection and what it means to be human.

Buckle up.

Perfectly imperfect

Early in my career, the agency I worked for proudly presented a new client with the beautifully printed catalog we had all been working on for months. Thousands of copies had been printed, trimmed, bound, and boxed — ready to ship. The client flipped through the pages, thrilled with how well their products translated to print — the feel of the paper, even the smell of a freshly-printed marketing piece. Then they stopped suddenly.

“Huh,” she said, as we anxiously awaited what was to follow. “There’s a typo.” It wasn’t a misspelling of the company name or an incorrect product price, but it was a typo, nonetheless. A word that’s supposed to be spelled one way — spelled an entirely different and completely wrong way.  

Silently, poker-facedly, each of us thought, “Ahhhhhh, a typo!!! Did the client sign off? Whose fault is it? How did we all miss this? What’s are print going to cost? Who’s getting fired?”

Before the first drop of panic sweat could drip onto the conference room table, the client spoke: “Well, I guess that’s our Navajo mistake. All good.”

She went on to tell us about nearby Navajo weavers and their longstanding tradition of intentionally working mistakes into their patterns. The weavers consider their skills a gift from the Holy People and regard perfection as the exclusive purview of said Holy People. For a human to create something perfect (or to even try) is an affront. This conscious practice of deliberate imperfection keeps both artist and audience grounded in their humanness.

Uncanny

In 1970, robotics engineer Masahiro Mori discovered something unexpected in his study of human-robot interaction: humans felt increasingly favorable towards robots the more human they appeared, but only up to a point. Precisely when the robot-human likeness became uncanny, the human subjects’ feelings about the robots went from “Oh, cute.” to “Oh, hell no.”

We’ve been feeling that uncanny valley ick ever since, and not just with robots. Whether it’s the Polar Express, CGI Princess Leia, or creepy robo-callers, our aversion to things that seem slightly less human than human is baked into our very evolutionary makeup.

Human touch

Just today, I got a sales email from one of my favorite online retailers. It was brief, bold, and bulleted. Crafted for engagement.Just the right amount of emojis. (Unlike this article, amiright?) Then I read the sentence: You golf us what you love, and that’s what we’re delivering. Then I read it again. You golf us? What? Then I looked at my keyboard. Two simple slips of the left hand, and you’re typing golf when you meant to type told. You told us what you love … That’s a mistake AI would never make.

A year ago, I might have silently judged a fellow marketer for not getting a second set of eyes on their work and moved on. Instead, I had a warm, emotional reaction that surprised me. A fellow human — who loves socks even more than I do — wants to sell me socks. They crafted a message, didn’t overthink it (likely didn’t even proof it themselves), and hit send.

Diff’rent strokes

We judge robot output differently than we judge human output. The general public and the marketing community alike are still very much in the novelty stage with artificial intelligence. It’s fun to play with.It’s exciting and a little bit scary. It’s mind-boggling how quickly we can useAI to create content. And the level of polish is nuts. As creators, we’re right to want to explore where this new and super-powerful tool can take us. And we’re right to proceed with caution.

AI is unburdened by the self-doubt that keeps us mere mortals curious and questioning.  

As consumers, we’re open and curious about how other humans might employ AI in their efforts to connect with us. AND, like Masahiro Mori’s subjects, when the robots start to drift out of their lanes and into ours, we don’t like it. We’ll disengage before we run the risk of letting a robot trick us into thinking it’s a human. We’re increasingly sensitive to the little ways AI shows its hand:

Short, punchy statements.

One-sentence paragraphs.

Emojis.

And confidence by the bowlful.

AI writes with a level of confidence that is both inspiring and cringeworthy. It is unburdened by the self-doubt that keeps us mere mortals curious and questioning.  At its best, it’s teaching us how to be more straightforward, clear human communicators. (AI eats clarity and craps simplicity.) At its worst, AI comes off like that know-it-all blowhard we all know and loathe — cleverly regurgitating what we feed it without adding any real insight or wisdom.

Slop, slop

As we continue to weave robots into our everyday lives in cute and helpful and frightening ways, we’ll increasingly seek out signs of real human fallibility and get savvier at distinguishing human mistakes from robot mistakes. No human illustrator accidentally puts six fingers on one hand, but we’ll be accidentally spelling public as pubic for the rest of time.

I suspect that as we all explore this new frontier together, we’ll become increasingly intolerant of AI slop and increasingly forgiving of human slop.

No human illustrator accidentally puts six fingers on one hand, but we’ll be accidentally spelling public as pubic for the rest of time.

What’s wrong?

We learned early on, from presenting our work to engineers, not to be offended when they skip the compliments and go directly to identifying what’s not working. It’s their nature. It’s who they are. Innovation doesn’t happen without a robust system for identifying, fixing, and learning from bugs.Especially in the early iterative stages, it’s more valuable to fix what’s failing than celebrate what’s succeeding. Mistakes make us better.

Our increased openness to signs of human fallibility is not a license to get lazy. Mistakes will always be sloppy and frustrating and preventable and bad. Run spell check. Get a second set of eyes on it. But done is better than perfect. And perfect is a fundamentally flawed construct. Make something. Stumble. Fail. And next time, fail better. (Let’s see a robot make a Beckett reference!)

Humans will always be humans, and we will always be evolving and adapting and finding real, meaningful ways to connect with other humans. Make no mistake about it.

‍

{{OrangeQuote | ABOUT TOM CAMPBELL | Tom is Toolbox’s co-founder and creative director. When he’s not validating the merits of ordering pizza for lunch at the office, Tom can be found tending beer league hockey or playing drums for local bar bands. He’s also the keeper of TomLovesTheLibertyBell.com, a quirky repository of stories and stats on Liberty Bell replicas around the world.}}

{{OrangeQuote | ABOUT TOOLBOX CREATIVE | Toolbox Creative is a B2B Brand Engineering firm helping cool people in Additive Manufacturing, CleanTech and Emerging Technologies change the world. We distill complex technologies into powerful identity systems, websites and marketing tactics that create lasting impact and build brand love.}}