There’s no doubt that AI is changing marketing, data science, and pretty much any other industry you can think of. One AI model, ChatGPT, has been getting a lot of attention thanks to its ease of use and seemingly high-quality output. ChatGPT can debug code. It can answer exam questions and even help pass the bar exam.
Most recently, there’s been a huge hubbub about how ChatGPT could replace human writers. All over the country at this very moment, thrifty marketing VPs are asking themselves, “Why to pay for a human copywriter when you can press a button and get what seems like the exact same result in thirty seconds?”
While I agree that ChatGPT will definitely change how we think about and value human writers, that change will happen in a direction many AI fans won’t expect.
Throughout human history, technology has helped us produce at scale. We’ve witnessed the development of mass production of furniture, books, cars, art, and now text. But at each point in the past, technology has not rendered the original creator obsolete.
In fact, it has often made the work of the original creator even more valuable. Look at music, for example. Today anyone can stream their favorite band on repeat using a service like Spotify.
Technology has enabled the rapid creation of something at scale and has allowed the original creator to reach more people and impact more people.
So far, text produced by ChatGPT feels a lot like any other example of where tech has to create anything at scale. It can generate copy, landing pages, and other content very cheaply and quickly. But far from rendering writers obsolete, AI places a premium on the originality and deep thought that great writers bring to their craft, which remains a necessary input.
So, what does all this mean for the value of human-generated text?
How Will Readers Value ChatGPT-Written Content?
AI-generated text is not error-free, at least for now. For areas in which expert knowledge is needed, such as data science or finance, AI can’t be trusted to generate text without producing mistakes or plagiarizing.
Regardless of the quality level, which may improve in the future, my question is this: In a world where AI-written text can be bought and sold for a fraction of the price of human creations, what happens to the value of the human output?
Let’s go back to our art example. Original art is a byproduct of human consciousness and intention. The very knowledge that it is the byproduct of human thought, and experience, is inextricably tied to its value.
AI text generators, meanwhile, are predictive in design. Intentionally, they use algorithms to predict likely outputs based on what’s been created before. So while they’re creating work that at best hasn’t been created before, they are by definition repetitive imitations of that which has been done before. This puts a premium on originality. That originality is the spark of input that can be given shape and form and amplified by AI technology.
Can Human-Authored and AI-Written Content Be Distinguished?
It’s only possible to value human-created text more highly if you can tell the difference between AI-written and human-generated content.
Today, bad AI text is easy to spot, but it’s getting harder and harder to detect the difference between mediocre human text and good AI-generated text.
However, even in the near future when AI models will be better, I expect it will still be possible to tell the difference between the two because content platforms and the laws that govern them will force users to either avoid or disclose AI-generated text. There may even be technology similar to watermarks that make it possible to spot AI-generated text.
For one, many content gatekeepers like Google will be on the lookout for ways to spot and penalize AI-generated content, since it goes against their webmaster guidelines. On other platforms like Medium, where AI-generated content is not outright banned, users are already required to say when text has been AI-generated.
Additionally, while it may be a bit longer time in coming, I suspect we’ll see more legal restrictions in this industry, too. Right now, it’s illegal for me to paint something and sell it as a Van Gogh original, no matter how good the painting itself is. AI-generated text may face a similar future.
So even if AI text reaches the level at which it is impossible to distinguish between AI text generation and human-written content, there will still be incentives for disclosure and transparency. This means that human-written content will be identifiable.
Incorporating both ChatGPT and Human-Written Content
As I’ve established, ChatGPT forced us to question the value of human content. And yet, the most successful companies won’t go just with humans or AI alone. The smartest business owners will go for a mix of the two.
I’m the Co-Founder of a company that wields an AI-powered platform, so it shouldn’t be too surprising to hear I’m bullish on the future of AI-generated text. I’ve seen firsthand how building an AI platform makes me appreciate my human writers all the more, but I’ve also seen how to use AI to scale my business.
For example, we use AI to analyze an incredible amount of content produced in nearly real-time to identify media opportunities that are relevant to our users. Based on the detected trends, our AI can write draft pitches to journalists based on the established positions of our clients. But without the human input – the actual opinion or perspective of the client – you don’t have much of a story. It’s that human input that brings the story idea to life.
AI can handle explainers or basic how-tos. But we also need human writers to create meaningful insights.
The Value of Human-Generated Content
It’s likely that we will see AI continue to change industries across the board. But one overlooked area of change is that technology like ChatGPT will highlight the value that human-generated content provides. The smartest use of AI will continue to be in combination with the human touch: AI can help scale the business while human writers ensure quality and impact.