on March 24, 2026, 5:56 pm
It's a lot easier to create content if you use genAI to write your stuff for you, and easier still if you don't even bother fact checking the results, but I won't be doing that
Another Angry Voice
Mar 24, 2026
When the journalist Andy Twelves took a look at MattGPT Goodwin’s new book "Suicide of a Nation" he found misused statistics; made up quotes; deceptively framed quotations; and several tell tale signs that portions of the book have been generated by ChatGPT.
Aside from the made-up quotes attributed to writers like Cicero, Friedrich Hayek, and Roger Scruton, one of the most obvious signs that Goodwin relied on ChatGPT to compose his book was that three of the book’s twelve citations still included the query string ?utm_source=chatgpt.com in the url!
Goodwin isn’t just any old right-wing agitator either, he’s a top GB News presenter, and the figure that Reform chose to try to parachute into the Gorton and Denton constituency, which they thankfully lost to the Green Party.
It’s obviously a lot easier for lazy opportunists to create loads of content if they use genAI to write their stuff for them, and easier still if they don’t even bother fact checking the results to ensure they’re not full of fabricated quotes; hallucinated sources; and misrepresented statistics, before publishing it as if it’s their own work.
There’s no way I’m going to use genAI to create content for AAV, and it goes further than simply not trusting the technology to avoid stuff like all the fake quotes, misleading statistics. and hallucinated sources.
If I was intellectually lazy, I could prompt genAI into creating my articles for me, then go through the outputs to check that all of the quotes, statistics, and sources are actually real. However that appears to be a lot of boring work just to evade composing articles myself, right?
From a personal point of view, it just seems much less tedious to just write what I actually want to say, and then find the necessary quotes, statistics, and sources for myself where required.
Presumably the people who follow and support AAV continue to do so because they feel I’m adding something beyond what they could easily generate for themselves by typing a few prompts into the genAI model of their choice.
If I were to go down the lazy route of getting ChatGPT to write my articles for me, I can’t see how it would be possible to maintain a distinctive writing style, or avoid trivial and derivative analyses.
I guess there’s a big difference between my intended audience, and the target audience of a far-right rabble rouser like MattGPT Goodwin.
Some people want real analyses, written by real people, containing real information, and some people are apparently content to waste their time reading lazy and misleading AI-generated slop, as long as it confirms their pre-existing biases.
Who cares if it’s contributing to the destruction of our shared objective reality, if it tells them what they want to hear, right?
I’m not trying to claim that all uses of genAI are inherently bad. I know people who work with AI, and I’m not a luddite. What I’m saying is that I’ll never simply generate AI content and present it as my own work. Corrected, or uncorrected, it’d be a fraud against my readers and supporters.
GenAI can be useful in certain circumstances, however it represents a threat to writers in a lot of ways, not least the AI industry’s absolutely brazen copyright theft in feeding vast amounts of people’s work into their language models without compensation, and the fact real writers (many of whom had their work stolen to create this tech) are all having to compete in online spaces that are increasingly filled with so much lazily-generated AI slop.
However I’m hopeful that there are enough people out there who still want to read real analyses, written by real people, and containing real information, that they’ll continue to visit and support independent media sources like AAV, which don’t lazily publish misleading AI-generated slop as if it’s their own work.
Responses « Index