Here was the first story in yesterday’s Executive Communication Report—to which you should subscribe because unlike most things in this world it, is both useful and free.

Several readers asked what I think about this. I went back as I usually do to a Fortune piece I wrote three and a half years ago in which I concluded about AI-generated leadership communication: “A leader who understands communication knows that audiences will sense the absence of a living person from the process and turn away in vague disappointment.” How many leaders really care about that? “I’ve wondered that for a long time,” I wrote. “And it looks like we’re all about to find out.”
That’s going to work itself out case by case, over many years—in human writers’ favor, I believe. Or after I am dead, I hope.
But in this particular case, AI isn’t actually the main problem: A guarded author and a lazy editor are. See, Druckenmiller was Scott Bessent’s mentor, apparently, and so he had a potentially human point of view that the WSJ should have wanted and a reader would have appreciated—and that would have been impossible for AI to render properly in any case. Except I’ve read the turgid Druckenmiller piece three times and it’s either a massive missed opportunity or a gross disclosure violation or both: It doesn’t even mention Druckenmiller’s relationship with Bessent, which apparently was close when they both worked for George Soros years ago.
Any decent ghostwriter would have urged Druckenmiller to bring humanity and extra authority to the piece by at least acknowledging the personal connection with Bessent. And WSJ editor Paul Gigot was surely ethically and editorially negligent in letting this piece pass without forcing Druckenmiller to disclose their onetime bond. Which would have made the piece more compelling and honest both.
Authors who truly want to communicate for max impact and editors who want to publish the finest and most credible writing they can find for the human readers will do the work themselves, or find human helpers who can. Those who don’t, won’t. (It was always thus.)
P.S. A colleague who works at MIT points out that the editorial operation at their Admissions office apparently holds itself to a higher editorial standard than The Wall Street Journal‘s editorial page:
“Unless explicitly stated otherwise, all blogs on the MIT Admissions website are written without the use of generative AI/language models. We view the blogs as a relationship between our bloggers and you, our human readers. The bloggers show our respect for that relationship by writing in their own voice and sharing their own perspective, without being filtered or rearticulated by generative AI, so that you can get to know some MIT students for real, and share a little bit in their subjective and lived experience.”
No, Gigot?
***
UPDATE: In the WSJ, Paul Gigot writes a detailed defense. He says that in the newsroom, “The rule is that everything my colleagues and I write must be original work. No letting AI draft a column or editorial. Our readers need to know we’re not subcontracting our ideas or prose to the amalgamation of a machine.”
But it’s more complicated for guest columnists because they “aren’t professional writers, yet they can provide value for readers owing to their expertise and standing in public life.”
Many contributors, especially prominent politicians or CEOs, employ speechwriters or hire a Beltway writing shop. How different is that from using AI to streamline an essay?
What matters is the ultimate argument the author makes and stands behind. In the case of Mr. Druckenmiller, his views on the bond market and U.S. fiscal incontinence are worth listening to. The press corps wants me to disavow Mr. Druckenmiller, though his views on financial markets have been formed by decades of experience.
First, Druckenmiller didn’t say he used AI to “streamline” an essay, he said he used it to write the essay. But how different is using a speechwriter to compose an op-ed from using AI? Depends on how you use the speechwriter. If it’s the old “write down my ideas as if I had them” routine—not much different, maybe. If it’s the kind of thoughtful, creative back-and-forth that led John Kennedy to call his speechwriter Ted Sorensen “my intellectual blood bank”—well, then I’d say it’s a lot different. And if it’s in between, it’s in between.
Again: The core problem here wasn’t the AI. The problem was a piece that could have been more compelling and candid if humans had involved themselves—in the writing, and in the editing. Instead, Druckenmiller used AI to crap it out, and it sounds like Gigot ran the crap, straight.
