MattGPT

Right, where Ben Sixsmith writes:

Here is an exceptionally easy argument to make:

1. Mass migration is ensuring that the historical majority in Britain is becoming a minority.
2. This is the result of policies that have been pursued regardless of popular opinion.
3. This has had many kinds of destructive consequences.

The first claim is so obviously true that one might as well deny the greenness of the grass. The second is proven by decades of broken promises (see Anthony Bowles’s article “Immigration and Consent” for more). The third requires argumentation, but I think that it is clear if one considers hideous incidences of terrorism, grooming gangs and violent censoriousness, as well as broader trends of economic dependency and electoral sectarianism.

Again, this is not a difficult argument to make. So why is it made so badly?

Matt Goodwin’s Suicide of a Nation is a very bad book. It reads like the book of a political operator extending his CV. The left-wing commentator Andy Twelves caused a stir on social media by pointing out various factual mistakes and what appear to be non-existent quotes. Twelves speculates that these “quotes” are the result of AI hallucinations, which is plausible, if not proven, in the light of the fact that two of Mr Goodwin’s sparse footnotes contain source information from ChatGPT.

Inasmuch as Suicide of a Nation makes a form of the argument sketched out the beginning of this article, there is truth to it. But it contains a fundamental problem — it assumes that this argument is so true that there is no requirement to make it well.

“Slop” is an overused term but it feels painfully appropriate for a book that is spoon fed to its audience. Goodwin, who had a long academic career before becoming a successful commentator, is not a man who lacks intelligence. But he writes as if he thinks his audience lacks it. “I did not write this book for the ruling class,” writes Goodwin, “I wrote it for the forgotten majority.” Alas, he seems to think that the average member of the “forgotten majority” has the reading level of a dimwitted 12-year-old. As well as being stylistically simple, the book is full of annoying paternal asides. “In the pages ahead I shall walk you through what is happening to the country …” “In the next chapter we will begin our journey …” Thank you, Mr Goodwin. Can we stop for ice cream?

The book is terribly derivative, with a title that reflects Pat Buchanan’s Suicide of a Superpower and a subtitle — “Immigration, Islam, Identity” — that all but repeats that of Douglas Murray’s The Strange Death of Europe — “Immigration, Identity, Islam”. It is written in the humourless and colourless rhetorical style of AI. I’m not saying it was AI-generated. (Indeed, a brief assessment using AI checkers suggests that it was not.) I’m just saying that it might as well have been.

There is a lot of highly dubious sourcing. Mr Twelves did not exhaust the number of questionable quotes in the book. “A nation that cannot defend its borders,” Goodwin claims the Roman historian Livy wrote, “Will soon cease to be a nation.” I can find no record of this quote. “Language is the tie that connects past with future, and binds together the citizens of the same nation,” Goodwin claims the lexicographer Noah Webster wrote. Again, I can find no record of this quote. To be fair, perhaps Goodwin has one — but he should have included more footnotes. He’s an academic! How did this not occur to him?

The writing is lazy. Goodwin, for example, twice refers to “the Canadian psychologist Gad Saad” within forty pages. (Thank God — I would have thought it was a different Gad Saad.) But the arguments are also lazy. “Would countries such as China, Saudi Arabia or Japan tolerate this?” Goodwin asks, after listing a series of crimes committed by asylum seekers, “No. Only Britain’s elites are this deranged.” Germany’s? Sweden’s?

What makes all this so strange is that it is so unnecessary. This should not have been a book that was difficult to write thoroughly and effectively. Again, I can only assume that Goodwin does not feel as if he has to make the argument well. His apparently abundant sales suggest that in cold economic terms this might be true.

Why? One explanation that occurs to me is that the right is unusually post-literate. Its success, in marketing terms, has been more prominent in the realms of social media and, especially, podcasts. The right has developed a formidable podcasting ecosphere, in which Goodwin is a major player, where the failure to systematically articulate one’s perspective is less obvious. A book, with this in mind, could almost be an afterthought — an excuse for generating more podcasts.

Secondly, the spectre of political correctness provides a built-in excuse for poor argumentation. “The elites will attack me because I wrote this for you,” writes Goodwin, “They will call me every name under the sun because I dare to tell you the truth.” On Twitter, Goodwin is already claiming that the left is “having a meltdown” — “cherry-picking” and “misrepresenting”. “Ignore the losers on the Left,” writes Goodwin, “Read it for yourself.” We should read books for ourselves. But we should also read critiques for alternative assessments. That we should ignore censoriousness does not excuse ignoring criticism.

Finally, and most generously, perhaps people have a sense that urgent truths should be communicated plainly and directly. Of course, there is something to this. Arguments, in many cases, should be clear and accessible. But this should never come at the expense of accuracy (what is Goodwin claiming to provide if not the truth?). Nor should it entail dumbing down complex issues (and what should be done about the state of Britain is complex, as Goodwin surely realises since he has spent much of the past months arguing with members of the nationalist party “Restore”).

Besides, call it idealistic but I just don’t think a book that claims to defend British culture should be so short on eloquence, wit, scholarship, poetry et cetera. British culture is pretty meaningless if it has the literary standards of ChatGPT and the argumentative standards of a telemarketer. Goodwin is talking about major issues here — tremendous demographic and cultural change, democratic dysfunction, and societal pathologies — but to do it so carelessly diminishes their status rather than inspiring or persuading people. The narrative around Suicide of a Nation has nothing to do with immigration, Islam and identity but with its stylistic and argumentative deficiencies — and the sad truth is that the author can hardly complain when there are so many of them. The lesson for his colleagues in Reform is not to let short-term popularity make them complacent when it comes to political and intellectual seriousness. 

Left, where Sophia Sheera writes:

Defeated Reform parliamentary candidate Matt Goodwin has been accused of using ChatGPT to research his latest book.

Goodwin, the GB News presenter who contested the Gorton and Denton byelection last month, published Suicide of a Nation: Immigration, Islam, Identity on 16 March. 

Early readers of the book say that it includes fake quotes, cites non-existent articles and relies on “AI hallucinations” to make statistical claims.

“I’m only 5 chapters in and have found a huge amount of what appears to be false quotes and basic misinterpretations of data that appear to be AI hallucinations,” wrote Andy Twelves, a New Statesman contributor, in a long thread on X.

Twelves accused Goodwin of “statistical illiteracy” regarding the claim that most primary schoolchildren in Leicester, Luton and Slough do not speak English as a “main language”. Children who grow up speaking an additional language at home are more than capable of speaking English fluently or as a “main language” at school, Twelves pointed out.

Goodwin’s book also includes quotations from Roman statesman Cicero, economist Friedrich Hayek and political theorist James Burnham – but Twelves claims that none of the quotes exist in the public domain.

John Merrick, deputy editor of The Break-Down, pointed out that several of the book’s footnotes include ChatGPT in the URL – meaning the source of information was not found by Goodwin, but generated by the AI chatbot.

“If the research of Goodwin’s book was conducted mainly by AI, then he’s done a terrible job of hiding it,” Merrick wrote on X.

He added that there are only 12 footnotes in the book, with five referring to Goodwin’s own Substack and two generated by ChatGPT.

Multiple X users who have criticised Goodwin’s methods claim that he has since blocked them.

Goodwin responded: “Everything in this book is based on official UK census data and the very same projections that are used by the Office for National Statistics and expert demographers.”

He added: “The Left don’t want you to read it, they don’t want you to know what is in this book, because they do not want you to know what is happening around you.”

Goodwin was a professor of politics at the University of Kent for nine years until 2024. In February he ran for MP against Green candidate and plumber Hannah Spencer, who won the vote by a margin of over 4,000.

And centre, where Mic Wright writes:

When the journalist Andy Twelves took a look through Matt Goodwin’s latest book – Suicide of a Nation – he uncovered a litany of false quotes, misinterpreted data, and a heavy reliance on the deceptive output of ChatGPT in the first five chapters alone. There are 12 references in the book; five are to Goodwin’s own Substack; three indicate they were sourced via ChatGPT; plus one Telegraph article, which quotes… Matt Goodwin.

At first, Goodwin’s response was to crow on X that the Left were “driving the algorithm” and simply contributing to the sales of the book. After 24 hours of consistent pressure, he returned with a long attempt at rebutting the objections, titled with his characteristic pompousness, ‘A response to my critics’. But while he dismisses criticism of his use of AI – claiming he simply used it to “[obtain] datasets… [which were] cross-checked with the original source” – defends his use of statistics and argues that the book’s paucity of references is not a problem, he does not address the issue of fabricated quotes.

In Twelve’s original thread of X, he identified quotes that Goodwin attributes to Cicero, Hayek and James Burnham that do not exist in the published works of the three men. I’ve also been through Suicide of a Nation and found other examples of what appear to be fabricated quotations. A particularly bad example can be found early in Chapter one when Goodwin claims that the American lexicographer Noah Webster wrote: “Language is the tie that connects past with future, and binds together the citizens of the same nation.”

Webster, whose central project was the evolution of a distinctly American English, did not compose that quote. Instead, he wrote: “A national language is a band of national union.” Goodwin attempts to bolster his false claim that “millions no longer speak the national tongue” with a quote that appears to be fabricated and is spun off from an entirely different argument made in a totally different context.

I don’t use ChatGPT or any other generative AI tools in my work, but it’s clear from the book references and his subsequent statements that Goodwin does, so I asked ChatGPT about Webster and the line “language is the tie that connects past with future…” At first, it claimed that the quote was taken directly from Webster’s work, but when I asked for the source, it said it is attributed to him but “not found verbatim in his major writings”.

It then went on to caution that “if you need to cite it academically, avoid using the quote as a direct citation”. That suggests that ChatGPT has a more stringent approach to referencing than the former professor, Matt ‘MattGPT’ Goodwin.

Let’s look at another of the seemingly fabricated quotes. Also in Chapter one, Goodwin writes: “As the American writer Christopher Lasch once warned: ‘A country without collective memory, shared identity, and emotional loyalty cannot withstand the centrifugal forces of individualism and group difference’.” Again, the quote doesn’t seem to appear in any of Lasch’s publicly available published work or interviews, and that clunky triad at the beginning has all the hallmarks of AI writing.

Where Goodwin cites statistics or research in the book, they’re almost always presented in the same way as they appeared in a newspaper article, usually from the Daily Telegraph, but occasionally from the Daily Mail or The Times. For example, Goodwin writes:

“In 2025, the NHS spent £64m million solely on translation services, enough to cover the annual salaries of nearly 2,000 nurses.”

In a Daily Telegraph report published on 13 September 2025, the line reads: “The data show that the total spending on translation and interpretation over the five financial years was £243m – equivalent to the cost of employing nearly 2,000 NHS nurses.”

Notice that Goodwin’s removes any reference to interpretation. As NHS consultant, David Oliver wrote for Byline Times back in 2024, in reference to similar stories published by the Daily Express, “£100 million on translation might sound like a big number, but it is a tiny fraction of expenditure and would make little dent in nurse staffing across all NHS organisations.”

Goodwin doesn’t just regurgitate lines from the newspapers. He’s also very keen on a partial quote and a tossing aside of context. One example of that is when he writes: “Andrew Neather, speechwriter to Tony Blair and his immigration minister Barbara Roche, admitted in 2009, [that] mass immigration had been a deliberate policy. The aim, he said, was to ‘rub the Right’s nose in diversity’.” Now, Nether did write those words, but in an Evening Standard column headlined Don’t listen to the whingers – London needs immigrants and in a paragraph that read in full:

“I remember coming away from some discussions with the clear sense that the policy was intended – even if this wasn’t its main purpose – to rub the Right’s nose in diversity and render their arguments out of date. That seemed to me to be a manoeuvre too far.”

With that context, the line becomes a lot less punchy than the six words dropped into the heart of Goodwin’s anti-immigration invective.

Putting aside the intellectual dishonesty that runs through Goodwin’s work, like the message in a peculiarly poisonous stick of Blackpool rock, there’s a wider danger from this kind of book. Those fabricated quotes will enter the information stream and, when generative AI gobbles them up, will end up being treated as real. Goodwin tells his readers, “The elites will attack me because I wrote this for you. They will call me every name under the sun because I dare to tell you the truth.” These are the words of a man making a preemptive strike, of someone who knows full well that his book is a carnival barker’s act that he’s trying to pass off as a statesman-like argument.

This is about more than a bad book by Reform’s notorious bad loser. Goodwin is explicit that he aims to further shift the Overton Window of acceptable political discourse. He assumes that he can wave his book around and use it to justify his position as a commentator and thinker, safe in the knowledge that most people – even those who buy it – won’t read it.

Not only is it a badly written, self-satisfied, and smug book, but it’s also filled with falsehoods and inventions. If they go unchallenged, they pile up over time and ossify into received wisdom, raised as proof purely because someone stuck them between covers. MattGPT is as bad for the intellectual ecosystem as ChatGPT is for the environment.