Demand writing that’s intelligent, not artificial.
Whether through widespread media coverage or your own exploration of ChatGPT and other large-language model (LLM) AI platforms, you’ve likely seen how chatbots can now create social media posts, blog posts, and even longer-form essays in seconds.
It’s copy. But is it writing?
If we define writing solely as placing grammatically correct words on a page, then the answer is yes.
But as humans, we know writing is so much more.
The power of original content
Think back to a recent article that left you thinking or a book plot that’s stayed with you for years. There’s a reason why it stuck: strong writing often conveys a writer’s personality or a captivating point of view — both uniquely human traits.
Many recent reviews of AI-generated copy have noted its formulaic blandness. Yes, the sentences and paragraphs are there, but they’re lacking a human voice.
Here at Dragonfly, we believe writing with a clear human touch will rise above the coming wave of AI-produced copy. In this new landscape, writing by real humans will have the power to truly resonate, perhaps even more powerfully than before.
Human writers, after all, have certain skills AI simply cannot replicate.
Our Dragonfly writers create copy that is truly original. Our writing is fresh and forward-thinking — and distinctly unique from the product chatbots create by sourcing millions of other outlets’ previously published words.
Backed by interviews with experts, our own skilled research, and a knack for wordsmithery, our writing encapsulates new ideas customized for a specific client and audience. As a result, our messages are targeted, nuanced, and deeply captivating in ways AI-produced copy simply are not.
Current limits of AI
While the definition of strong writing is somewhat subjective, most everyone would agree on a few basic criteria. For starters, effective copywriting should be fact-based, up-to-date, and error-free.
But current AI models struggle to consistently deliver on these basic expectations. A quick survey of media coverage around chatbot-produced text notes that:
- AI writing models sometimes output information that is demonstrably inaccurate.
- At this point in their development, chatbots cannot distinguish between fact or fiction.
- At times, to back up their claims, current AI models fabricate sources that simply do not exist.
- The information sourced by LLMs can be glaringly out-of-date. For the latest iteration of ChatGPT, this data cutoff came in 2021. That’s why, if you ask ChatGPT to name the British monarch, it still confidently (but incorrectly) offers the name of Queen Elizabeth II.
Because of these inaccuracies, certain media outlets, including Wired, have banned the use of AI-generated copy on their sites entirely.
For the reasons outlined here — and many others — Dragonfly clients can rest assured the copy we deliver will always be created through the innate talents of our team of (very) human writers.
Going forward, our staff plans to research ways new AI tools might enhance our teams’ productivity — think initial idea brainstorming, SEO-keyword suggestions, or article outline development — but the copy we create for our clients will always be written the “old-fashioned” way, by hand, one keystroke at a time.
And by that method, we’ll continue to specialize in writing that reads as clearly intelligent and assuredly not artificial.