By Sheila Lothian, Content Strategist & Editorial Director Generative AI can’t infuse a piece of writing with passion, experience, authenticity or soul. One writer wonders: What if it doesn’t matter?
I’m a professional writer. For the better part of 25 years, I have made my living creating written content on demand. So when I heard there was now an available-to-the-masses online tool that could do exactly what I do—have always done, have forged my identity and supported my family doing—AND do it not in frenzied can-you-fix-this-real-quick minutes, stressful do-you-have-it-yet hours, painful it’s-due-in-two-weeks days or torturous we-go-live-next-quarter weeks... but in a matter of calm, robotic, unemotional seconds? For free?
I expressed my anxiety on a group call not long after the news of OpenAI’s bomb hit. The objections were immediate and absolute: “What?”, “No way!”, “NEVER.”, “No machine could replace you!” They were SO nice. They were so encouraging! I really wanted to believe them. But I heard the ring of panic in their voices, too. I realized we were all just sitting frozen in the headlights of this speeding AI train, not knowing how far away it was, when the moment of impact would come or what it would feel like, only that one way or another, it was careening straight toward our livelihoods.
But one of my colleagues on the call didn’t join the chorus. He simply gave me a calm, knowing smile from his little corner of the screen, and quietly popped a message into our one-on-one chat thread: “I’ll call you after this.”
Minutes later, he and I were engaged in a combination generative-AI crash course/talk therapy session. “So, you’re worried this thing is going to eliminate your job.” I was. “Have you tried it yet?” I hadn’t. He explained that ChatGPT, while an incredible technological milestone and enormously powerful tool, was only as good as the human-generated prompts it’s given and human-generated information it draws on. That it was prone to bias, inaccuracy and error. That it might function like a kind of intern or junior copywriter for me, even a brainstorming partner, something that could spark ideas or get a piece of content off the starting block. But it wasn’t going to replace me, with all my talent and training and experience. At least, not anytime soon.
1“Here’s how many U.S. workers ChatGPT says it could replace,” CBS News, April 2023.
From acute threat to existential dread
It didn’t take long to realize he was right. As I punched in prompts and absorbed paragraph after paragraph of perfectly coherent, articulate, digestible responses, the bone-chilling fear for my professional survival started to dissipate. This was not ready-for-prime-time material. This stuff needed help. Craft. Expertise. Starting point, maybe, but not a finished product fit for public consumption—not on my watch, anyway.
But as the fear faded, another feeling started to seep into its place. One that grew in the weeks ahead, as more people began to dip their toes in OpenAI’s glittering waters. Microsoft Teams chats pinged with personal examples: a website landing page written in the style of Eminem... a sales talk track rendered in iambic pentameter... a picture of someone’s labradoodle fused with Vegas-era Elvis. Links started flying around with real-world use cases: creating delicious recipes out of random refrigerator contents... planning detailed European vacation itineraries.
It was cool. It was AMAZING. And yet. Something about it made me feel... sad. Empty. Like my brain was being drained; my soul sucked out and suffocated.
The feeling soon crystallized into a new, more universally existential kind of panic. Was this new tool the beginning of the end of originality? Authenticity? Creativity? Thinking? The end, even, of storytelling as we know it?
Love in the time of AI
As a corporate writer, the subject of my writing has never been myself, nor has my foundational source material and motivation been my own ideas or experiences. My projects originate in strategic plans and product launches, downloads from managers and requests from colleagues. I scour resources to support claims. I interview subject matter experts for facts, details and insight.
I’ve always believed, however, that any piece of writing should strive to be genuinely enjoyable to read, within the parameters of its genre. Of course, someone does not come to a set of release notes with the same expectations of entertainment and enlightenment they bring to a bestselling novel. And a technical whitepaper packed with inside jokes and ironic asides would completely detract from the purpose. But people are people. We want and, frankly, deserve to have reading be a pleasant experience, even—maybe especially—when we’re reading for something other than pleasure.
Whenever someone prefaces critical feedback of something I’ve written with an apology, I interrupt them: “Don’t worry about hurting my feelings: this isn’t my memoir.” Yet there’s an element of my writing process that’s ineffable. An alchemy that ties all the disparate external data, facts and perspectives I’ve gathered into a story that’s not only clear and cohesive but also, hopefully, delightful. Whether from superstition, lack of necessity or both, I’d never interrogated that process before. But contemplating what was missing from the AI-generated content I was reading forced me to think about what was present in mine.
The right friend for the journey
A few months into the generative-AI revolution, and I don’t have the answer. Maybe machine synthesis and regurgitation will ultimately be a good-enough substitute for human thought and creativity across a critical mass of content types and my contributions won’t be needed. Maybe the love I’ve poured into my writing all these years, and joy I’ve derived from it, will prove, in the end, to have been for my personal gratification alone, objectively elevating the quality of the writing, but ultimately unnecessary for achieving its intended purpose.
I have seen it posited that, in a world where trustworthiness is increasingly suspect, authentically human-crafted content will become a premium product, like bespoke suits or handmade furniture. I’ll admit it’s a prediction that keeps me warm at night.
But I’ve also embraced the fact that the day-to-day job I love is changing, just as countless jobs over the course of human history have changed in the face of innovation. And like everyone else, then and now, I need to adapt. “Prompt engineering” doesn’t involve precisely the same creative and intellectual muscles as from-scratch writing, but they’re adjacent. If that’s what I need to get great at next, I’ve got the chops. It’s a new skill to learn, hone and master, not a mortal threat to hide from or conquer.
Zero generative AI was used in the writing of this piece. But through trial and error, I’ve confirmed that I can do the things with this incredible tool that have been promised—kick-start ideas, accelerate output, do more in less time—without losing my brain or soul to it. While I have no plans to hand over the whole of my work to ChatGPT, I do find myself tapping it most days for one purpose or another, and will continue to do so.
The advent of ChatGPT has predictably given rise to an ever-expanding army of people offering tips, tricks, formulas and strategies for harnessing its power. Of the seemingly thousands that have come across my LinkedIn feed, the one that has resonated most is: Talk to it like you would a friend. The best outputs I’ve gotten from ChatGPT have come from inputs crafted through that lens. “So, listen, Chat, I need to bounce something off you. Here’s what I’m trying to accomplish. Here are a bunch of details about it. Don’t solve it for me, don’t do it for me, but give me some feedback; tell me what you think.”
Like it or not, we’re all headed into a brave new world, with no choice but to move forward. An invisible, unfathomable, career-killing, soul-crushing machine isn’t the partner I want for that journey. A friend, I’ll take. I can’t predict our destination. But who knows? Maybe together, we’ll end up somewhere better than either of us could have imagined.