The AI Noise Machine: More Content, Less Communication
As AI takes over more of the communication process, the real question isn’t what it can produce, it’s whether anyone’s actually behind it.
By Ryan Newfell | Guest Contributor
When I set out to research AI’s impact on communications for my graduate coursework at Clark University, I expected to find a story about efficiency. What I found instead was a story about noise.
I interviewed Mike Matchett, an IT industry analyst at Small World Big Data whose career spans military intelligence, engineering, and decades of technology analysis. I also drew on research from Kate Niederhoffer, a social psychologist whose work on AI-generated workplace content has appeared in the Harvard Business Review, and Aviv Ovadya, an AI and media integrity expert focused on synthetic media. Together, their perspectives painted a picture that challenges the dominant narrative around AI and content creation: faster is not the same as better, and more is not the same as meaningful.
The “Net Middle” Problem
Ask a large language model (LLM) to write something and it will almost always produce something that looks finished. The grammar is clean, the structure is logical, and the bullet points are neatly bolded. What it won’t produce, at least not reliably, is something original.
Matchett described this during our interview with a term I had not heard before: the “net middle.” LLMs, he explained, are built by averaging vast amounts of existing data. They don’t generate new ideas, they synthesize existing ones. “It’s basically averaging the data that’s out there,” he told me, “giving you the net middle. not the best, not the worst, not the outliers.”
For communicators, this is a fundamental problem because the whole job is differentiation. Press releases, campaign messages, and unique brand stories are supposed to stand out and make an impression on an individual. AI-generated content, by its nature, tends toward the consensus. It checks the boxes, but it rarely challenges anything.
This doesn’t mean AI has no place in the communications process. Matchett was quick to acknowledge its genuine strengths: it’s a powerful research tool and a useful first-draft engine for routine material. “It’s like running a good search,” he said. The problem is when organizations and communicators mistake the AI draft for the deliverable.
Workslop: When “Faster!” Creates More Work
That mistake of slinging AI content forward as a finished product has a name. Niederhoffer and her colleagues at Harvard Business Review coined the term “workslop” to describe AI-generated workplace content that masquerades as good work but lacks the substance to actually move anything forward. The sender saves time, but the receiver loses it by having to wade through and often redo AI output that looked complete but wasn’t.
Marketplace Tech reported that 40 percent of surveyed workers had already encountered workslop in their organizations. And as Niederhoffer noted, the costs compound fast: “If it takes someone just under two hours to deal with every single episode of workslop, then there’s an enormous cost to the organization at scale.”
This dynamic plays out in communications constantly. AI summaries that capture the topics but miss the argument or nuance. A media pitch with the right structure but no hook. An event abstract that lists the themes without connecting them into a reason for the reader to care. Readers need something genuine and authentic.
The temptation to accept this work is real, especially when speed in an AI-driven corporate environment is a growing pressure. But the hidden cost of workslop – especially in credibility and in work relationships, is higher than the time saved in producing it.
The Authenticity Crisis Happening Right Now
Beyond quality, the deeper problem is the erosion of human presence behind it, the part of communication that makes it feel genuine and authentic – or truly human.
Matchett put it plainly during our conversation. “If what I’m presenting came from an AI,” he said, “and then the audience is not going to read those notes and will have another AI read those notes – then there’s nobody in the room anymore. So what’s the point?”
He’s describing something that’s already happening. Zoom meetings attended by AI note-takers on behalf of people who aren’t present. LinkedIn feeds where marketing automation talks to other marketing automation. Reports written, sent, and summarized by successive layers of AI without a human meaningfully engaging at any point. The process becomes efficient yet empty at the same time.
What’s striking is that audiences are already starting to notice and respond. Matchett observed that people are already developing a kind of AI radar, an instinctive recognition of the “net middle” voice that lacks a clear agenda or a real point of view. “People are already at the point where they can look at a page and go, that’s AI slop,” he told me. “There’s no thought put into it.”
I noticed this in myself during this research. My ability to detect AI-generated writing has sharpened considerably over the past two years, and with it, my skepticism of it. Matchett made the same observation: “Right now we’re trying to actively avoid looking like an AI. We still want a human creative voice.”
The Verification Problem Coming Next
If authenticity is the current crisis, verification is the one on the horizon.
Aviv Ovadya, whose work focuses on synthetic media and AI-generated content, has warned that as detection practices become more accessible, they also become more easily circumvented – a kind of arms race with no obvious finish line. The deeper challenge, as he frames it, is societal: “If we can’t distinguish fact from fiction, or reality from fakery, we can’t make effective decisions as society.”
Matchett arrived at a similar conclusion from a different angle. In a world where voices, faces, and written personas can be convincingly synthesized, the question of provenance becomes central to all communications work. Source, authorship, image authenticity and intent will need to be demonstrated, not assumed. “What we’re really going to need,” he said, “is a way to determine authenticity in digital communications.”
That’s not a problem technology alone will solve. It’s a challenge that puts human judgment back at the center.
The Race Nobody Wins
There’s a temptation when a powerful new tool becomes available for companies to believe that adopting it first or using it the most will create a lasting competitive edge. Matchett is skeptical of that logic when it comes to AI, and he has a name for why: the Red Queen race.
The term originates in Lewis Carroll’s Through the Looking-Glass, where the Red Queen tells Alice that in her world, you must keep running just to stay in the same place. Evolutionary biologist Leigh Van Valen borrowed that image in 1973 to describe a principle he observed in nature: species must keep adapting indefinitely not to advance, but simply to survive relative to other evolving species. Matchett sees the same dynamic playing out across industries racing to integrate AI. “Everyone thinks they’re getting ahead,” he told me, “but they’re really just keeping pace.” Because AI tools are easily accessible, any advantage one company gains is quickly matched by the next. The race intensifies and the output multiplies but the net result is more content and more noise.
For communicators, this is the trap hiding inside the efficiency argument. If every team is using the same tools to produce faster, speed stops being an advantage when everyone has it. What remains as a differentiator is the one thing AI can’t replicate at scale: genuine human judgment about what’s worth saying, and why.
What This Means for Communicators
None of this means AI should be avoided. That would be both impractical and shortsighted. What it means is that the role of the communicator is changing in a specific direction.
Matchett used an analogy that stuck with me: the spreadsheet. When spreadsheets arrived, the fear was that they would put accountants out of work. Instead, they eliminated the tedious parts of the job and created more demand for the parts that required judgement and thinking. AI, he argued, is likely to follow a similar pattern. “The blue collar part of the white collar job – the part that was a lot of grunt work – a lot of that grunt work is going to be collapsed by AI.” What remains, and what becomes more valuable, is everything that requires real judgment: strategy, interpretation, originality, and earned trust.
The communications professionals who will thrive are the ones who understand that clearly. AI can help generate ideas and produce starting points. The job of the communicator is to take those starting points somewhere worth going: to bring a point of view, make a real argument, and create something that a human on the other end actually wants to engage with.
The future challenge isn’t learning to produce more content. It’s learning to produce content that’s actually worth producing.
Ryan Newfell is a marketing and communications professional with more than a decade of experience helping organizations turn complex ideas into clear, useful content. He is currently pursuing graduate study in communications at Clark University. This article draws on original research conducted for a graduate course on AI and communications, including an interview with Mike Matchett of Small World Big Data.