Excessive wordiness can alienate readers. I often find myself accused—yes, accused!—of employing a thesaurus to inflate my writing with complex words, but such claims are only partially accurate. While I do utilize various resources in my writing, it’s primarily to swap out my usual words for simpler alternatives. The rise of artificial intelligence has introduced writers to a valuable tool designed not to supplant human effort but to enhance it.
The literary community was startled by Washington Post columnist Megan McArdle’s recent admission regarding her use of AI tools. She described how these tools now perform tasks that traditionally required a team of magazine editors, including fact-checking her work, transcribing interviews, and scrutinizing her assumptions to uncover flaws in her reasoning. Likewise, I leverage AI to simplify complex expressions, verify grammar, and track down infrequently referenced posts from years past that seem lost in the depths of Google’s search results. Essentially, it serves as my copyeditor and research aide.
Such practices contravene no ethical standards that I’m familiar with. Nevertheless, for those who hold a strict view of AI’s role, McArdle’s candidness has been regarded as a serious misstep. This overreaction hints at a broader anxiety surrounding AI, which often does not differentiate between appropriate use and misuse. Yet, there have been instances of clear abuse.
Critics of AI may bristle at the description of their position as panic-driven, but they have valid concerns. Predictions about AI’s impact on white-collar jobs are indeed concerning. However, economic worth is derived from scarcity; the more unique a human skill is, the greater its value.
Any large language model can mimic a crowd, regurgitating conventional wisdom with a veneer of authority. However, such models fall short in delivering writing that conveys vulnerability, profound insights, or grapples with controversial ideas that challenge contemporary norms or contradict elite consensus.
As long as writers like Megan McArdle exist—those who are both provocative and notably successful—critics will always try to undermine them. In this context, AI also acts as a weapon for detractors. Their broad interpretation of the ethical restrictions surrounding AI use serves their own agendas. Ultimately, history will remember McArdle’s contributions to the collective understanding more than it will the voices of her critics.
If AI becomes pervasive, journalism is doomed
The rapid integration of AI into the professional landscape has been striking, especially considering the lack of regulation surrounding tools like ChatGPT, Claude, or Gemini, many of which frequently misconstrue facts. A global study by the European Broadcasting Union and the BBC revealed that these AI tools misrepresented news content approximately 45 percent of the time.
So, when Megan McArdle, a columnist for the Washington Post, shared on social media just how extensively she relies on AI in her writing, it caught me off guard. Not only did she openly discuss her dependence on AI tools, but she did so in a decidedly casual manner. Here is a prominent columnist from one of the nation’s leading newspapers, boasting over 100,000 followers on X, who assured her audience that her expert analysis is partially grounded in tools that can be unreliable—tools that diminish the human aspect crucial to valuable writing and journalism.
Credit where it’s due: McArdle uses AI not to completely produce her columns, but to enhance her arguments. As she stated, journalists should view their chatbot as a blend of an intern, an initial editor, and a fact-checker.
However, if this approach becomes widespread among journalists, the future of the industry looks bleak. The tasks McArdle identified for AI tools—those typically carried out by an intern, editor, and fact-checker—are roles that should remain human-driven. As more journalists embrace AI as a replacement for human collaboration, media companies will feel less hesitation in reducing staff to cut costs. This shift will disproportionately impact entry-level journalists, driving them to seek employment elsewhere.
Essentially, AI threatens to dismantle the traditional pathways through which many accomplished writers enter the field. Where will we cultivate the next generation of skilled journalists if they cannot secure jobs within the industry or receive training in our newsrooms? Rejecting AI isn’t merely technological skepticism; it is a commitment to journalism’s future, which relies on human contribution.
That said, there’s potential for these tools to be beneficial. They can assist in research much like Google or Wikipedia have done for years, serving as a gateway to deeper inquiry. Yet, as AI starts overshadowing human roles in writing, reporting, and editing, we risk edging closer to plagiarism. Cases like Nota, an AI initiative meant to address local news shortages, demonstrate how lines can easily be crossed.
I know I’ve got to figure out AI. That doesn’t mean I like it.
Despite knowing that I should learn how to disable the AI tool in my email application, my lack of technical skill prevents me from doing so. I chuckled recently when an email I sent to my editor returned with an AI-generated summary: “You complained to Marjorie.” It perfectly captured both my complaint and Marjorie’s spirited response.
For some time now, I’ve joked that I’ll never delve into cryptocurrency or artificial intelligence. At a certain age, one can only absorb so much new information.
However, I’ve come to realize that as a journalist, I need to engage with AI. It has permeated our industry. Yet, the messages we receive about its appropriate use are often contradictory. While aware that it’s employed in creating headlines and for research, we also receive cautionary tales regarding its misuse.
In my writing, I rely on the journalism 101 tenant: always verify sources. Not too long ago, this meant sifting through an envelope stuffed with aging press clippings from the Globe library; these days, it involves searching for verifiable information online. Recently, I’ve noticed how tempting it is to accept an AI summary that appears in a Google search. It requires more effort to scroll down to legitimate news articles and click on them.
And yet, what about the challenge of finding the precise word or phrase? In the past, I would flip through a physical thesaurus. Now, a quick Google search yields a plethora of suggestions. Recently, I sought a quote concerning contempt, related to President Trump and his dealings with NATO allies, considering its implications in the ongoing conflict with Iran. In a flash, I found this from Lord Chesterfield: “Wrongs are often forgiven, but contempt never is. Our pride remembers it forever.”
I didn’t specify my intent for the quote when I found it, but it felt just right. Naturally, I then needed to gather information on Lord Chesterfield, whose name rang a bell but whose biography I could not readily recall. I discovered he was a British politician and diplomat, adding even more relevance to the quote.
While I haven’t written that column yet, I find myself questioning whether AI shaped my thought process. The initial idea was genuinely mine, but I wouldn’t have uncovered the quote without that Google search. As long as I attribute it accurately, I believe it’s appropriate to use. Yet, in the interest of journalistic integrity, do I need to disclose how I came across it? The answer remains unclear.
If another person had introduced me to Lord Chesterfield’s quote about contempt, I would have verified its accuracy without worrying too much about citing the source.
This is an excerpt from Globe Opinion’s weekly politics newsletter Right, Left, and Center.
Noah Rothman is a senior writer at National Review and a regular contributor to Globe Opinion. Abdallah Fayyad can be reached at abdallah.fayyad@globe.com. Follow him @abdallah_fayyad. Joan Vennochi is a Globe columnist. She can be reached at joan.vennochi@globe.com. Follow her @joan_vennochi.
