In 2019, Michael Shermer of Skeptic magazine came forward with an embarrassing but honourable admission. A freelance contributor to Skeptic, John Anthony Glynn, had faked his CV. Glynn, Shermer alleged, had made up his Ph.D:
Further research revealed that Mr. Glynn represented himself as a Ph.D. psychologist to several academic institutions (academic fraud), and under those credentials he published over 40 articles in 15 different publication outlets in 2019 alone (journalistic fraud).
Subsequently, another of Glynn’s publishers, Areo, removed his articles after determining that he:
… had so closely paraphrased Michael Gurian that Gurian’s words should really have been indented as a quote rather than a simple “As Michael Gurian points out…”
“John Anthony Glynn” disappeared from the internet. Soon afterwards, though, a prolific freelance writer named “John Mac Ghlionn” began writing online. An examination of the profile pictures of John Anthony Glynn and John Mac Ghlionn heavily suggests that they are the same person:
Additionally, John Glynn and John Mac Ghlionn are both Irish, have both spent time in Thailand, have both laid claim to an academic background in psychological fields, and have both advertised an interest in or admiration for Irish Republicanism (with Glynn using the handle “Irishdawg1916” and Mac Ghlionn using a photograph of Michael Collins in his Twitter banner). How many academically-minded Irish Republicanism enthusiasts named John writing prolifically from Thailand can there be?
We at The Critic noticed this resemblance and chose not to publish Mac Ghlionn when he pitched to us. Mac Ghlionn has, however, been published extensively in the Spectator, Newsweek, the New York Post, the Catholic Herald, Courage Media and a dizzying array of other publications. His biographies suggest that he is a “psychosocial researcher”, though the nature and substance of that research is unclear.
There is nothing wrong with being prolific. (This author takes pride in the scale of his own output.) But one does have to put quality above quantity. It seems very possible that Mr Mac Ghlionn has an uncredited co-author: AI.
AI detection tools are imperfect. Still, a pattern might be telling. According to “GPTZero”, several of Mac Ghlionn’s recent articles have been AI generated. GPTZero was “highly confident” that “RFK vs Big Pharma is the battle that will define America’s future”, for Human Events, and “Peleton’s debt crisis is spinning out of control”, for the website UnHerd, and “The Philosophy of Jerry Seinfeld”, for The Philosophical Salon, were created using AI (what’s the deal with large language models?). Ironically, GPTZero also held that Mac Ghlionn’s latest Catholic Herald article “Spiritual starvation in the age of AI” was also AI generated. “Spiritual starvation” indeed.
Mr Mac Ghlionn’s most controversial piece was a Newsweek article making the case that Taylor Swift was a bad role model for young women because she was unmarried and childless. According to GPTZero, this article was — yes, you’ve guessed it — created using AI. (Say what you like about Ms Swift as a role model, but I don’t think that she uses AI to generate her songs.)
Using some of my own pieces for the sake of comparison, I could find no examples of GPTZero being anything but “highly confident” that they were “entirely human”.
We asked Mac Ghlionn if he was indeed “John Anthony Glynn”, and if he was using AI to generate his articles. Mac Ghlionn denied that he was “John Anthony Glynn”, suggesting that he could not see a resemblance between their photographs (the reader may be the judge) and claiming that he arrived in Thailand 18 months ago (he wrote in an article from January 2023 of visiting Thailand “frequently”). Additionally, Mac Ghlionn made the point that “AI plagiarism detectors are notoriously unreliable”. AI plagiarism detectors are unreliable. But how often are they unreliable? And do we have cause to trust Mr Mac Ghlionn?
Freelance commentary depends on trust between contributors and editors. To misrepresent yourself and your work is to poison the well for other hopeful writers. But this case represents not just one man’s failings — or even just the problems of AI generation (which David Scullion and I warned contributors not to use for The Critic this month). It represents the failings of a media culture that prefers volume to substance when it comes to commentary. None of us are perfect — but when there is such a market for formulaic arguments and robotic prose, editors have to ask themselves what they are encouraging.
