Hi there – have we met? If not, I’m Chelsi, marketing manager and content specialist at ODEA. I have a hand in writing content for ODEA and our clients. On any given day, you can find me with my headphones on with music blasting while pecking away at my keyboard. Most likely, filling a Microsoft Word document with thoughts, insights, and facts about adhesives, steel, technology, new tax laws, or something in between.
About a year ago, I created an account for ChatGPT 3.5 to check out my competition. ChatGPT is an artificial intelligence chatbot that can compose responses that mimic a human when prompted. It can produce content for just about anything you can think of – blogs, essays, emails, social media posts, and more. I read articles and watched video after video about how this chatbot would change the game for content writers. A few even mentioned how this powerful tool might push content writers out completely. Who needs a human to spend hours writing content when this intelligent program can digest your input and spit out a well-written output that meets your subject and length requirements in minutes? A content-generating machine that never has to eat, sleep, or take a break.
It walks like a human, but does it talk like a human?
But I forgot to mention my superpower – and if you can read this, your superpower, too – is the gift of the human brain. When writing for ODEA or a client, I can easily slip out of my normal voice and into their brand voice, like a chameleon blending into its surroundings. This is because of my knowledge of these brands, whether from the years of working with them, the in-depth research conducted, or the time spent with their subject matter experts. Even if I haven’t personally experienced what I’m writing about, I can connect with similar real-life experiences to evoke the thoughts or emotions needed for the piece. As a fellow human, I can understand and empathize with my audience, enabling me to create content meant to evoke a certain feeling or response. And no matter how powerful or smart ChatGPT or other language models become, they’ll never be able to replicate that.
While you can prompt ChatGPT to write content using a specific voice or tone, the intelligent tool can only mimic a human-like tone to a degree. It can’t grasp emotion or use its unique experiences to draw inspiration or insight. It’s solely dependent on the data the model was trained with. This limitation underscores the unique ability of human writers to infuse their work with genuine emotion and understanding.
Let’s compare, shall we?
One of our clients, Chicago Glue & Machine, has an incredibly distinct voice and has become one of my favorites to write for (don’t tell the others). Their voice is knowledgeable, caring, and positive, with a bit of humor, one of the reasons I enjoy writing for them. But there’s a fine line between giving your audience a chuckle or two while educating them on your adhesive services and turning your blog into a cheese fest with one too many glue puns.
So, we decided to test ChatGPT using Chicago Glue & Machine as our guinea pig client. They recently received a message from a client about an alleged glue failure and wanted to educate their audience on why it might have occurred. The blog needed to be caring, positive, and humorous while explaining the few reasons the glue failure happened.
After I finished writing my blog, I gave ChatGPT the following prompt: Imagine you are a seasoned Detective investigating a mysterious glue failure case involving Chicago Glue and Machine. Your task is to write a law and order-themed blog uncovering the reasons behind a customer’s sealing failure on a box stored in the freezer. Dive deep into the investigation, gather evidence, and piece together the clues to solve this sticky situation. The blog should be no more than 500 words.
So, how does ChatGPT measure up? I’ll let you be the judge. But I feel pretty comfortable that I’ll be jamming out content for at least the foreseeable future.