How to write better than ChatGPT

My quest toward fully human writing

Friends,

When a colleague offered to edit an essay for our policy institute, I told him: Don’t neuter the personality! I don’t want it to come across like ChatGPT wrote it!

I paused, then added: It can’t sound like the Brookings Institution either! 

For many of us, having our writing mistaken for ChatGPT output is the nightmare insult of the current year. If my writing is indistinguishable from the bland but thorough output of a machine, what does that say about me? I might as well bow down, yield my keyboard, and surrender to the machines now.

How do we distinguish our writing from the output of a technology that gets smarter and more talented by the day?

I have scratched my head on this question since ChatGPT was released exactly a year ago this week. My conclusion, after lots of scatching, is simple: We distinguish our writing from ChatGPT by being more human.

The hard part is figuring out what this means in practical terms. And so, I am on a quest to explore: What does fully human writing mean?

One clue came from my friend

CansaFis Foote

, who reviewed an early draft of this essay. In his note to me, he wrote:

“I feel like a broken record man because we all fail this inspiration, but the more of yourself you drop on this the better it reads. US will find YOU the more YOU are US. So cheezy, but for real make this about you, to and through, and I think it will sing louder.”

He was encouraging me to put more of ME in this essay. Although I didn’t follow his advice too well on a literal level, I read a meta message in his feedback: Putting more of ourselves in our writing is central to fully human writing.

I recently completed a fun, 5-week writing program called Write of Passage. One of the frameworks they use is called POP, which stands for Personal, Observational, Playful. Participants would ask each other, “How can I POP this essay?” By this we meant, how can we punch up the playfulness, personal elements, and unique insights and observations in a draft essay. 

When I started the program, I was decent at observations, but personal and playful were out of my comfort zone. I thought that if I was too personal, it would be navel-gazing, vulnerable, and people might not like me. I thought that if I was too playful, it would detract from the way I wanted to come across — smart, professional, and kind of elite. 

Writing this Substack has allowed me to challenge these hang-ups and experiment with POPing my writing. It’s been fun! In the course of doing this, it occurred to me that the qualities of POP — the personal details, unique observations, playfulness that reflect me — not only make for better writing but are the elements that differentiate my output from ChatGPT’s. They are core elements of fully human writing.

Stories are core, too. ChatGPT did not share a laugh with the quirky girl in Ugg boots at Starbucks this morning, when I joked about arriving early and stealing the table where she sits every morning. ChatGPT did not lay on the grass and marvel at the full moon with my little one the other night. ChatGPT didn’t travel with a conman through Australia for a month a decade ago either! Sure, ChatGPT can spit out stories, but it cannot tell MY stories unless I feed it my specific data.

Some of the earliest evidence of human stories comes from the cave drawings in Lascaux and Chavaux, France. The drawings date as far back as 30,000 years ago.

When I first started thinking about how to write better than ChatGPT, I thought of it as an extreme example of the college admissions essay problem: How do you stand out to an admissions officer who’s reading thousands of essays from high achievers just like you? Or, to use a less nerdy analogy, how do you stand out in the New York City dating market when there’s always someone hotter, younger, funnier, more successful, who also by the way has a book deal or TV gig?

In either scenario, I would need to stand out by being MORE ME and also, frankly, showing the equivalent of cleavage. What I mean by cleavage are the unique assets and attributes that lure others in and give them value. In the context of writing, this could be unique experiences, weird areas of interest or expertise, or even mistakes. For example, I am gay and have a son. If I wanted to provide value to would-be gay parents, I could write an essay on parenting as a gay dude, and ChatGPT could not compete with my unique story and insights. My “cleavage” in this case is the fact that I’m a gay parent and have had this experience. To get noticed by the target audience, I would need to hang it out there in order to get discovered. Fully human writing means showing off these unique attributes and experiences.

Human connection is another thing ChatGPT cannot replicate. Cam Houser, an expert on visual storytelling, made an interesting point in one of his newsletters. He noted how many of us roll our eyes when we see mediocre ideas from an influencer spread like wildfire. But the reason they spread, he suggested, is because those people have built a human connection with their followers. Think of someone like Oprah. An AI-generated influencer may be able to attract a following, but “it” will never be able to replicate Oprah’s ability to hold the hand of a woman and cry with her.

Fully human writing means building human connection, and relatability is central to this. While I am not a fan of revealing my flaws and mistakes, I am learning that my inadequacies and imperfections are what make me relatable as a fellow human. They are what allow me to connect emotionally to others traveling through life. To err is human — literally.

And erring leads to wisdom, which is another advantage humans like me have over ChatGPT. While AI can provide answers based on data and machine learning, and there’s value in that, it cannot provide wisdom and insights based on LIFE. Wisdom, like mistakes, remains in realm of fully human writing.

To circle back to my policy paper, I am not sure how far to lean into BRINGING MORE ME into an essay for a professional audience, but I do believe personality and voice matter more than ever in the professional arena. For personal essays, fully human writing seems more straightforward. 

As I explore these issues, I am discovering that there’s a secret prize on the other side of the existential dread many people are feeling. It is permission to be fully human. Fully human writing is the perspective I will hold as I go about my writing. So far, I find it invigorating. Liberating even.

So thanks, ChatGPT, and happy birthday.