Is AI Stealing the Soul of Storytelling?
- 17 hours ago
- 4 min read
AI Prompt: Should foundations be thinking of new ways to leverage live events and first-person storytelling? Does underproduced video take on a more important role? What are the AI tells that will lead to donor mistrust? Is brevity more important now than ever? Are there predictive strategies to consider? Is it already too late? Is AI stealing the soul of storytelling? That was a real prompt from last week. And yes, I see the irony. I don't know if you will have better luck asking your AI. The answers I got were thin. Anecdotal at best. Which, in a strange way was reassuring. I've built a career out of words. Teaching storytelling, writing case statements, building talking points for gift officers and board members and the occasional CEO who has forty seconds with a donor and needs all forty of them to count. The product of philanthropy is nuanced and deeply personal and genuinely beautiful when somebody presents it right. So this moment lands bittersweet for me. AI offers a lot. It also feels like a dangerous time to be in this line of work.
The product of philanthropy is nuanced and deeply personal and genuinely beautiful when somebody presents it right.
I wrote this as much for me as for anybody else. Let’s start with the good, because there is real good here. AI is a strong brainstorming partner, a fast editor, a tireless researcher, a story excavator, a content organizer. It is useful for signal-spotting, catching the redundancy that turns into message fatigue before your donors catch it for you. My personal favorite is regionalizing a piece of content, then having persona profiles react. That used to take a week and sometimes real world testing. But the price is real, and it is worth naming. AI took at least two things from us. Neither of them is soul.
But the price is real, and it is worth naming. AI took at least two things from us. Neither of them is soul.
It took the apprenticeship. Voice used to be earned. You ground out bad drafts, somebody red-lined them and somewhere in the bleeding you learned to think. Skip that and you don't get worse writers in 10 years. You get no storytellers at all. It took the reader's benefit of the doubt. Every donor now reads with a suspicion budget. That battle is already lost. And if your plan is to police em dashes and count how often AI delivers things in threes, well, I just wrote that sentence in threes on purpose, and you didn't catch it until I told you. It's a game you cannot win. So the question stops being how to avoid detection. It becomes how to be worth reading. Stop writing like a chart note. Most philanthropy copy reads clinical because it was built from clinical source material. Diagnoses, service lines, volumes, outcomes. All of that is what we want people to know. Very little of it is what we want people to feel. I have never once seen a donor tear up over a service line. Write toward the feeling and the facts still get in. They ride along. Let a quote be a quote. A nurse says "I don't know, she just needed somebody to sit there." Leave it. Don't fix the grammar, don't finish her thought, don't hand it back to her as a complete sentence with the verb in the right place. I've done that to people's words for years and I regret most of it. The unpolished quote is now the most credible thing on your page, and every pass you make over it takes a little of that away. Relive it instead of retelling it. Retelling sounds like: The family faced a difficult diagnosis and found comfort in the care team's support. Reliving sounds like: It was four in the morning and the vending machine was out of everything but Sprite. Somebody's uncle was asleep in the chair by the window with his coat still on. There is no outcome in the second one. It doesn't need one. It puts a reader in the room, and once somebody is in the room they will stay for whatever comes next. Be short when short is honest. Not short for its own sake. But when the point is made in four sentences, the fifth one is doing something else, usually reassuring the writer. Put a real person on it. A photograph somebody actually took, of a face you could pick out of a crowd. Not stock. Not the lit portrait against a gray gradient. And a byline with a name on it. The author is part of the evidence now. Anonymous institutional voice used to read as authority. Today it reads as automated. On disclosure, which everyone is going to ask about. Don't disclose the tool. Disclose the method. Something like: AI helped organize the research. The interviews, the words and the judgment are mine. A donor can evaluate that sentence. "Written with the assistance of AI" gives them nothing to evaluate, and it spends their whole suspicion budget in the first line. Put it at the bottom. A disclosure up top tells people how to read everything underneath it. Here’s mine for this piece: This piece was edited with AI. The thinking wasn't. And don't claim anything you can't defend. If the quotes came from a model, don't call them interviews. That's the one version of this that actually breaks trust, and it is the only one I'd call unforgivable. So, is AI stealing the soul of storytelling? No. But it has made the soul much harder to prove.
About the Author:
Michael Beall is a Principal Consultant and Communications Practice Leader with Accordant. You can reach him at Mike@AccordantHealth.com and connect with him on LinkedIn.
Comments