At the end of November, new technology was introduced to the public that is poised to revolutionize the field of public relations.
And it had absolutely nothing to do with Elon Musk’s Twitter.
While much digital ink was spilled trying to figure out what the chaos at Twitter means for clients, far less attention has been paid to ChatGPT, the new digital chatbot from OpenAI.
In simple terms, it is a large language model which uses artificial intelligence to analyze and learn from massive amounts of text scraped from books, articles, and websites. The result is a tool that provokes equal parts fun and fear.
Users can ask questions or make requests and get clear, human-like answers. As the New York Times noted: “It can write jokes (some of which are actually funny), working computer code and college-level essays. It can also guess at medical diagnoses, create text-based Harry Potter games and explain scientific concepts at multiple levels of difficulty.” Microsoft Clippy this is not.
Researchers are still exploring all of ChatGPT’s potential, but there are some immediate applications for our industry. ChatGPT's ability to quickly generate written content can be a valuable asset for public relations professionals. By providing suggestions and ideas for stories and campaigns, ChatGPT can help PR professionals to generate fresh and engaging content. Additionally, ChatGPT can assist with fact-checking and editing, ensuring that all written materials are accurate and free of errors. This can save time and allow PR professionals to focus on other important tasks, such as media outreach and building relationships with journalists.
Let’s test it out.
Recently, advocates have pushed policymakers at the federal, state and local level to adopt universal free school meals. A go-to PR tactic for this kind of advocacy campaign is a letter to the editor of the local paper. So I asked ChatGPT to do just that.
Me: Write a letter to the editor arguing that school meals should be free for all students.
ChatGPT: Dear editor,
I am writing to argue that school meals should be free for all students. Providing free, healthy meals at school is not only a moral imperative, but it is also essential for the academic success and overall well-being of our youth. Studies have shown that when students have access to healthy meals at school, they are more likely to perform better academically and have better overall health. Additionally, providing free school meals can help alleviate the financial burden on families and reduce food insecurity among children.
Furthermore, offering free school meals can help combat the persistent issue of childhood obesity, which is a growing concern in our country. By providing healthy meals at school, we can help educate students on the importance of good nutrition and encourage them to make healthier food choices.
In conclusion, providing free school meals for all students is an important step towards ensuring that all children have access to the healthy, nutritious food they need to succeed in school and in life.
Not bad. If an intern on my team produced this as a rough draft, I’d be pretty pleased. We’d want to make it specific, authentic and infused with pathos—but this is more than a good start.
Here’s another example. Let’s say a beverage company is rolling out a new kind of soda; they’ll need some tweets to promote the product. I asked ChatGPT to do the thankless task of drafting social content.
Me: Write a tweet from a beverage company promoting the launch of a new soda.
ChatGPT: We're excited to announce the launch of our new, refreshing soda! Made with natural ingredients and no artificial sweeteners, our soda is the perfect choice for any occasion. Try it today and experience the delicious taste for yourself! #NewSoda #RefreshingSoda #NaturalIngredients
Again, it’s a place to start!
But while both examples hit the major points, they don't do it in any inspired way that would truly excite a reader or move them to action. As with most of chatGPT’s prose, both are rather bland and what WIRED’s Amit Katlaw calls “superficially impressive but largely lacking in substance.”
Then again, so is much of human writing. Nonetheless, the ability to produce a serviceable first draft is incredibly valuable. In a profession as hectic as PR, anything that can save a little time is greatly appreciated.
Where things get tricky is in the potential scale of this technology. Software that can write near limitless letters to a member of Congress will supercharge already dubious “AstroTurf” campaigns. As Tyler Cowen warns in Bloomberg, “Over time, interest groups will employ ChatGPT, and they will flood the political system with artificial but intelligent content.” Like any new tool, it is up to practitioners to use ChatGPT wisely.
Does all this mean that us PR pros should worry about our jobs?
Not yet, and maybe not ever. While ChatGPT can do the grunt work of a first draft, it takes a real human to elevate the writing to a level of sophistication that will inspire engagement. AI cannot capture all the complexities of the world. It pays no attention to the nuance and sensitivity required when messaging to diverse stakeholders. And it certainly plays no role in the day-to-day client management and relationship development that is essential to PR success. It is a blunt tool in a discipline that most often requires a scalpel.
Benji Englander is a Director at Seven Letter, a leading full-service strategic communications firm with offices in Washington, D.C. and Boston.