Three things ChatGPT needs which only you can provide
Discussions claiming ChatGPT will solve all our digital marketing problems are missing three key things that only humans can provide.
Many people writing about ChatGPT have used the gimmick of revealing at the end that it was written by ChatGPT. Not only is it now cliché, it’s a bad idea for more important reasons as well. These reasons underscore the missing elements in discussions claiming ChatGPT will either solve all our problems or bring on the zombie apocalypse.
Authenticity and its connection to the consumer — to our customers — has always been critical to messaging success, especially during and after the most intense days of COVID-19. Authenticity is a complex emotion because we’re talking about authenticity to a brand and authenticity to a person. At its core, authenticity helps us connect with people on an emotional level.
Machine learning, AI and ChatGPT can emulate authenticity, but they can’t replicate it as humans can. That’s the key point — at a human level.
Empathy is connected to authenticity. Both are essential for marketing success, another lesson we learned during the COVID disruptions. As a marketer, you have to have genuine empathy and authenticity. Sure, you have to know how to sell. But you can learn that.
Dig deeper: Does ChatGPT pose an existential threat to marketers?
I can’t teach someone to be empathetic and authentic. I can teach them how to break down the barriers to let empathy and authenticity come into their mindsets. The best marketers have these qualities and know how to act on them.
You know you have empathy and authenticity when you’re creating an email campaign, writing a blog post or a tweet, and you get a feeling in your stomach that something’s not right. That feeling comes from your innate understanding that your actions might exceed your brand or the customer relationship.
We can ask ChatGPT to write a first draft — something we can read through and adjust. But reading is different from writing. I have difficulty understanding how we can capture that innate emotion of empathy or authenticity.
2. Connection and extrapolation
I will never let ChatGPT or one of its sister or brother models write my articles or presentations. One of the reasons is that my brain “works goofy.” That’s a direct quote from my wife. She’s not wrong.
One of my writing powers is connecting something I did in financial services with something I will do in retail. Or, I can call on something I did over the last 25 years of my career and relate it to a challenge a marketer faces today.
One example: I can call on groundbreaking research that changed how we structure abandoned-cart reminders — that sending a first reminder email within an hour after abandonment was more effective than waiting 24 hours — to help a marketer design a more effective program.
Extrapolating from my personal experience informs how my brain works and how I communicate it in a venue like this article or a presentation to 200 people. That abstract connection doesn’t come from any database other than the one in my head.
It’s hard for me to understand how people’s reaction to language models like ChatGPT misses the point in authenticity and extrapolation. We’re all brilliant people making abstract connections like this to help the industry and lift all boats.
We can make it easy for one audience to understand a concept and then dive into the complexities for a more knowledgeable or experienced audience.
At a recent conference, where ChatGPT was either on the agenda of many presentations or brought up by audience members, it was clear several speakers didn’t read the room right. They seemed to aim their presentations at executives, not the operations people in the room. They over-complicated things.
Gauging your audience’s reactions and changing up if you detect a lack of engagement — that’s something a content database can’t do. In my case, that extrapolation of history combined with my opinions, plus chance and dumb luck, drives much of my content.
Among people in my industry, I’m known as the email preacher. Reading something I write, or attending my stage presentations, is like going to church and hearing a sermon. That comes naturally to me after a long history of this kind of speaking.
But the underlying current of my work is inspiration. How can I inspire people to do better? How can I call on empathy to understand their position and then thoughtfully push them in the right direction?
If you look back at my writing for MarTech, you’ll find I use my end-of-the-year article to inspire marketers for the next year and review the events of the passing year.
Inspiration can be captured and defined, but you must have lived the experience to give it the impact that informs and engages your audience. In other words, to communicate the authenticity that makes your words worth reading or hearing.
A language model or other technology can’t capture or replicate that lived experience. That inspiration is a personal desire many of us in the email industry share to teach, show or motivate people to do great things and rise above their own skill sets and abilities to improve and do more.
AI will never write my articles for me
Many reasons I say ChatGPT and models like it won’t write my articles for me are based on emotion. That’s why I fight against people who say content is dead. That’s a foolish notion — that one technology still in its infancy will kill off an entire realm.
It’s foolish to say that copywriting is dead now that we can call a bot to do it for us. How fast we go from living to dead in this world. Direct mail isn’t dead. Email isn’t dead. Let’s focus instead on what lives, changes and evolves. What expands our understanding?
ChatGPT has been around for less than a year, but some want to use it to condemn an entire profession? The baby was just born and now we’ve decided it will be President? Nonsense!
How about we look at it from this perspective: “What could it be?” Not “What will it be?” Let’s ask, “How can it enhance what we do?” Could ChatGPT and other natural language processing (NLP) models write copy to appeal to a specific persona or model data? Sure.
People are floating many ideas and use cases but not functional applications. Let’s be patient and have fun with the possibilities, not the doomsday predictions. Heck, we don’t even have full pricing yet.
NLP models like ChatGPT are still the latest shiny new toy in the digital playground. Let’s see whether they evolve into tools instead of gleefully predicting what they could kill.
Get MarTech! Daily. Free. In your inbox.
Opinions expressed in this article are those of the guest author and not necessarily MarTech. Staff authors are listed here.
New on MarTech