I came across three separate articles on writing and AI in the same week, each from a completely different angle and all describing the same thing.
A novelist turned writing professor at MIT confronts students who outsource their essays to AI. A new Graphite study showing AI-generated articles now make up about half of all new content on the Web and have remained stable there. And new data from The Accountancy Partnership shows that half of freelance creatives say increasing stress is affecting their work, as client budgets for human creative services shrink.
A data point is a fact. Two is a coincidence. Three is a trend.
When read together, these articles formed an argument that every SEO professional, content marketer, and creative freelancer should take seriously, recognizing the content divide that is occurring and asking, “Whose side are you on?”
The First Story: What Happens When Students Outsource the Struggle
On May 10, Micah Nathan, a novelist and professor of fiction and nonfiction writing at MIT, published an article in The guardian about confronting his creative writing students on their use of AI. The confession session that followed, he wrote, became one of the most productive teaching moments of his eight years at MIT.
His main idea was not about academic honesty. It was about what writing actually does. “Writing is not just the production of sentences,” he explained to his students. “It’s training in endurance through sustained attention. It’s a way of learning what you think by trying to say it. An LLM can reproduce the appearance of this activity, but it cannot replace it, because the value lies not only in the object produced but in the transformation that takes place during its manufacture.”
He described the AI’s prose as “impeccably flawless, glacially regular, magnificently rubbish”, borrowing Tennyson’s description of a beautiful but blank face, producing what he called “simulacra of thought, generated by the recognition of patterns learned from millions of words written by humans, rooted in no particular experience by no particular person”.
Insightful readers, he argued, feel this emptiness even if they cannot express it.
For SEO professionals, this is not an abstract literary concern. This is an accurate description of the content quality issue that Google’s Useful Content Systems Are Trying to Solve Since 2022. The signal Google is looking for is exactly what Nathan identifies as the thing AI can’t produce – evidence of a mind actively grappling with a specific problem from a specific experience. Pattern recognition learns from what humans have written. It can’t replicate the reason they wrote it.
→ Read more: Why quality content is no longer enough and what beats it in AI search
The Second Story: The Feared Takeover Has Not Yet Happened
On May 15, Megan Morrone showed up for Axios on new data from digital marketing agency Graphite, which analyzed 55,400 online articles and lists published between January 2020 and March 2026, each via three AI detection tools. The result is more nuanced than most coverage of AI content is when it comes to the share of content primarily generated by AI, which has stood at nearly 50% for over a year and appears to have plateaued.
The much-feared takeover did not materialize. AI content briefly surpassed human-created content in late 2024, but the two have remained roughly equal since.
The important caveat included by Morrone is that many articles are no longer written solely by humans or AI. A human can use AI to describe, draft, rewrite or edit, making the line truly blurred. Dan Klein, a professor at UC Berkeley and CTO of the AI model, flagged the risk of a feedback loop. Once models train extensively on AI-generated content, the Internet could become a machine that produces low-quality content that forms patterns that produce more low-quality content.
For SEO professionals, the board is both reassuring and warning. The volume panic was exaggerated. But the the problem of quality dilution is real and growingand that creates the same opportunity Nathan identified in the other direction. In a web that is half AI-generated content, content that carries real human experience and specific expertise becomes more differentiating, not less.
→ Read more: AI Platform founder explains why we need to focus on human behavior, not LLMs
The third story: the people producing this content are under significant stress
On May 13, Emma Hull at The accounting partnership directly emailed me data from a new report on creative freelancers in public relations, marketing, performing arts, graphic design, photography and adjacent industries. Half of freelance creatives (50.7%) say increased stress levels are affecting their work. Half (50.2%) say customer budget cuts are the biggest challenge they will face in 2025. More than two in five (43.3%) believe AI will have a negative impact on their industry. Nearly half regularly work unpaid hours each week.
Lee Murphy, managing director of The Accountancy Partnership, put it clearly: “Creative work is often closely linked to marketing budgets and discretionary spending. When businesses start to cut costs, creative departments can sometimes be one of the first areas to see investment reduced.”
The irony of these three numbers is worth thinking about. Clients are cutting budgets for human creative work even as AI generates about half the content on the web, while an MIT professor documents the specific cognitive cost that outsourcing the writing process imposes on anyone who does it, whether student or professional.
The freelancers most under pressure are the most tempted to use AI to produce more content faster to compensate for lower rates. The content they produce in this way is part of the 50% that is indistinguishable from the machine’s output. And content that is indistinguishable from machine output is exactly what Graphite data and Google’s quality systems train users and algorithms to ignore.
→ Read more: Too much reliance on AI backfires for businesses
What the model actually means
The three stories, read together, describe a market in the process of bifurcating. On one side is high-volume, poorly differentiated content, produced quickly, cheaply, and increasingly difficult to distinguish from AI production, regardless of who generated it. On the other is content that carries the specific expertise, direct experience, and editorial judgment that Nathan’s students were trying to circumvent. Content that takes longer, costs more and is increasingly the only one that pays off Significant search visibility and reader trust.
This is not a new argument in SEO. What’s new is the empirical clarity with which three independent sources from three entirely different disciplines – literary education, web content analysis, and freelance gig economics – all point to the same conclusion in the same week.
Shelley Walsh made this point in her recent Search Engine Journal article on AI content scaling that the real strategic question lies in the divide between raw materials and non-raw materials. The three stories above prove that the gap is already there, already measurable and already affecting people’s livelihoods.
Writers who understand this and produce accordingly are the ones who will still have work to do when budget cycles resume.
More resources:
Featured image: SvetaZi/Shutterstock
