The Replicant Effect
The decline of trust in writing, and how the human spirit will solve the problem
In the race to hit numbers, it’s easy to lose sight that communicating with audiences (thought leadership, marketing, sales, internal communications, etc.) is, at its core, a human-to-human activity.
When communication is wholly owned by technology — whether through thoughtless automation or fully AI-generated content — the result is what researchers call the "replicant effect."
The replicant effect was defined by Hancock et al. (2020) as follows: In a system that mixes human- and AI-generated writing, the trustworthiness of writing perceived to be written by AI will decline.
Let’s pause on the word “perceived” for a moment.
It doesn’t matter whether the writing was actually written by a bot or not. If the audience thinks there’s a chance it wasn’t produced by a human, their trust erodes.
And when people don’t trust your writing, they won’t trust you or your message.
This doesn’t mean AI is inherently bad. AI has its place and its legitimate uses (grammar checking, brainstorming, literal [not nuanced] translation, data analysis, and data transformation come to mind). But there’s a very real cost when it comes to using AI for writing: It undermines the audience’s trust.
What you might be surprised to know is that this decline in trust due to technology’s increasing role in writing isn’t new. The Hancock et al. study I referenced above, which first introduced the concept of the replicant effect, came out in 2020. That’s years before generative AI like ChatGPT became part of our everyday reality.
People act like generative AI has completely upended the world of communication, but the truth is, technology has been slowly affecting how we communicate for years. We aren’t in the middle of a sudden crisis in communication; rather, we’re nearing the end of a long, slow, downhill slide.
Now, I’m not someone who claims to know what the future holds. I’m not a futurist, and I don’t have any predictions about where AI will take us next. But I can tell you what I see today:
People are actively seeking clear, creative, interesting, and, most importantly, trustworthy communication. They’re unsubscribing from traditional news sources and flocking to platforms like Substack, where they feel more connected to real human voices.
They’re tuning out the replicants.
This shift should tell us something. As AI-generated content becomes more common, there will be a growing premium on human authenticity in writing. Audiences will continue to gravitate toward content that feels genuine, not mass-produced by machines.
Audiences will seek content with soul.
As we move forward, the challenge for writers and communicators of all stripes will be to embrace the efficiencies of AI where it makes sense, but to never lose sight of the human element that fosters trust.
Because ultimately, it’s the human touch that sets us apart from the replicants.
Bibliography:
Hancock, J. T., Naaman, M., & Levy, K. (2020). AI-Mediated Communication: Definition, Research Agenda, and Ethical Considerations. Journal of Computer-Mediated Communication, 25(1), 89–100. https://doi.org/10.1093/jcmc/zmz022