For my sins, I find observing human behaviour fascinating. It’s probably one of the reasons I’ve stayed on a platform like X for so long. Years of watching online interactions have given me a front-row seat to how debate has changed. While it’s impossible to prove that what we see online directly shapes behaviour offline, I suspect some of it inevitably does.
I’ve mentioned before that I’ve deliberately changed the way I write on X. I’ve tried to make my posts less confrontational and more evidence-based, bringing across the style I use here on my blog. Subscribing to X has made that easier because longer posts allow for more context and subtlety. It began as an experiment to see what the algorithm rewarded, but it has since become something more deliberate. Given the amount of misinformation and, in some cases, the complete absence of information, I think we all have an obligation to at least try to be accurate. That matters even more as political misinformation takes on a life of its own and the consequences can be alarming.
Lately, however, I’ve noticed another interesting trend.
Years ago, thoughtful or well-structured posts were often dismissed as “cut and paste”. Today, they’re increasingly dismissed as “AI-generated”.
I recently watched someone argue with both Grok and ChatGPT over one of my replies. He would feed my response into one AI, then take that answer to the other, hoping one would validate his position. When neither did, and Grok suggested he actually engage with the points being made, he accused Grok of, to quote, “polishing a turd”. It was both comical and utterly bizarre.
What interests me is that accusing someone of using AI has become a convenient way of dismissing what they’ve written without actually engaging with it. A thoughtful, well-reasoned argument is labelled “AI” simply because it stands out amid the stream of reactive, emotional and often inflammatory exchanges that dominate platforms like X.
I experienced that again yesterday. Someone repeatedly demanded evidence and, each time it was provided, simply dismissed it. When there was no substantive argument left, the conversation shifted to accusing me of using AI. The evidence was ignored; the focus became the supposed source of the words instead.
Ironically, people who have spent years writing professionally or academically can sound remarkably similar to AI because AI has been trained on well-written human text in the first place. Add to that the fact that AI detectors are notoriously unreliable, and confident accusations amount to little more than guesswork. Those of us who use AI for research know its limitations. It’s a useful research tool, but its output still needs to be checked, verified and challenged.
Of course, not every allegation is made in bad faith. People genuinely do use AI to write posts, and sometimes the signs are obvious. But I’ve increasingly seen the accusation deployed as a debating tactic rather than a genuine observation.
Once an argument becomes difficult to rebut, it’s easier to declare the author “fake” or “AI-generated” than to concede a point. The label becomes a signal to like-minded followers that the argument doesn’t deserve consideration, allowing the accuser to sidestep the substance without acknowledging they may have been wrong.
In the end, I don’t think these accusations reveal much about the person being accused. They reveal far more about the person making the accusation. More often than not, it’s simply a reaction to having a mirror held up to their own behaviour.
Those of us who have something, hopefully, worthwhile to say (whether readers ultimately find it worthwhile is, of course, for them to decide), and who are prepared to support it with evidence and reason shouldn’t be intimidated by this tactic. For many of us, our written words are the only means we have of expressing what we feel, what we believe and why we believe it. If we allow ourselves to be silenced by lazy accusations, we simply reward those who seek to discredit arguments they can’t or won’t engage with.
