What does AI have to do with storytelling?
This was the question that came to mind when I saw the speaker lineup for the Virginia Professional Communicators Spring 2023 Conference, which I had the privilege of attending with my Version A colleagues. (Side note – we won two awards!)
At first glance, the three lectures were an interesting fit. The first and last discussions were rooted deep in the history of Fredericksburg, where the conference was held, while the other focused on a very current topic – ChatGPT.
First, Dr. Gaila Sims presented “A Monumental Responsibility: Curating A Monumental Weight at the Fredericksburg Area Museum,” which centered around the cultural impact of displaying a slave auction block that had been a controversial cornerstone of downtown Fredericksburg.
The final talk, “Mary Lumpkin: How One Enslaved Woman Freed Her Children and Herself and Also Ensured that Generations of Black Men and Women Were Educated,” by Kristen Green, dug deep into the remarkable true story of a local woman, her descendants, and her key role in establishing one of the first historically Black colleges in the United States.
The middle presentation, at least on paper, stood out like a shining beacon of modernity between the two pillars of history. In his discussion, “ChatGPT and Communication: How Generative AI Will Transform Our Field,” Dr. Anand Rao explored how generative AI is revolutionizing the way the world thinks about writing.
As someone who writes about technology for a living but started out as a material historian, each talk was interesting to me for a different reason. It was the intersection of the three that turned out to be the biggest takeaway of the day.
The Difference Between Generative and Creative
Ever since ChatGPT went live in late 2022, writers have been concerned about job security. If you can simply plug a prompt into a chatbot and receive a full-blown article in return, why would you spend time or money hiring a writer?
The same worries have arisen for artists, whose careers seem threatened by AI art generators like DALL-E 2 and Midjourney. The images created by these bots can be visually stunning. Imagine portraits of modern figures in the style of classical artists or detailed landscapes in fantastic settings.
On closer inspection, the art doesn’t quite hold up. The perspective is a little off. The wording on signs is meaningless. The subject has six fingers.
The same can be true for the text generated by ChatGPT. At first read-through, it seems like the perfect solution to spending hours on research, outlines, and writing. Get in a little deeper, though, and there are telltale signs that it was created without a human touch.
Dr. Rao’s discussion delved into one of the main limitations of ChatGPT – namely, that it doesn’t really produce facts. Instead, AI is very good at identifying patterns. The generative programming strings words together based on contextual proximity and frequency, not actual knowledge.
ChatGPT’s database is limited to the data it was trained on, which was scraped from thousands of web pages and text on the internet. This content was written by people, which lends to the human-like quality of AI-generated writing. But it doesn’t have access to anything that isn’t already online.
In short, ChatGPT can create compelling content, but it’s not rooted in any kind of deeper understanding of a given topic.
Maintaining the Human Element
In contrast, the work done by Dr. Sims and Ms. Green was fundamentally human, centering around lived experiences and personal connections.
Dr. Sims collaborated with the Fredericksburg community to develop a living exhibition that responds to people’s reactions to a sensitive topic. While researching Mary Lumpkin’s life, Ms. Green deciphered handwritten records and talked to her living descendants.
This is the kind of research that can’t be reproduced by ChatGPT, and it underscores the importance of keeping the human element in writing. It happens offline, in person, and it can’t be replicated by a computer, no matter how sophisticated the technology. These are the kinds of insights that make our work meaningful.
We often use the term “insights” when discussing the way AI processes data, making it easier for users to interpret. But while AI can analyze patterns and produce insights, it’s not insightful. It doesn’t make connections in the way people do.
Many of our clients use AI in their businesses or products to optimize company resources. AI can help representatives better serve customers by summarizing interaction histories and suggesting resolutions. It can rapidly analyze thousands of sources of data, distilling it down into actionable information. It can fill out forms, automate communications, and identify trends.
These are all activities that help organizations run more efficiently. More importantly, it frees up employees so they can focus on the things only real people can achieve: making personal connections, creating new ideas, and keeping the human element intact.
Will ChatGPT Be Our New Cowriter?
Generative AI will certainly have a role to play in reshaping the way we write. Not even a year after its public debut, ChatGPT has revolutionized everything from school essays to resumes. You’ve almost certainly read AI-generated content without even realizing it.
At Version A, we’re experimenting with using it in different ways to enhance our work. When we’re stuck with writer’s block, we’re testing if it can be the first step in helping us move past the dreaded blank page. We’re also wondering if it can be a starting point when brainstorming topics for content planning. While we’re not sure the final role AI will play in our process, we’re excited to see where its potential takes us.
That said, even the most advanced AI is no replacement for truly understanding our clients’ needs and capturing their voice. Although we write about tech (which often includes elements of AI), our pieces need to be engaging for our readers. Making those kinds of connections can only come from a shared human experience.
So, ChatGPT may not be taking over our writing duties anytime soon. It’s a useful tool for things like figuring out the direction for a post or summarizing an article, but it’s no substitute for the depth of understanding that can only come from a real, live writer. That’s where we come in.
2 thoughts on “Real Writing in the Age of AI: Why Human Writers Aren’t Going Anywhere”