I Request AI for My Wife’s Name: Sincere Apologies to the Misguided Suggestions It Provided | Martin Rowson

Recently, the Rowsons stumbled upon an unconventional game that is both amusing and enlightening, which we shall temporarily refer to as “How Bloody Stupid is AI?” The beauty of this game lies in its adaptability; the way each player engages varies based on their individual life circumstances, yet the core rules remain unchanged. All you need to do is pose a straightforward question about yourself to an AI system and observe how hilariously incorrect its responses may be.

As for my situation, it’s essential to note that, given the nature of my profession, I maintain a significant online presence. In stark contrast, my partner—whom I married in 1987—has masterfully managed to cultivate a virtually nonexistent online footprint. Consequently, if one were to search for “Martin Rowson wife” in image results, the search would yield photos of me alongside our daughter, who at the time was 14, or standing with my friend and fellow cartoonist Steven Appleby, who happens to be transgender yet retains her birth name.

I admit that finding this discrepancy amusing is perhaps a bit reckless, but as a satirist, I have always reveled in anything that highlights the greater foolishness of our leaders as well as the capabilities—and shortcomings—of our new AI counterparts. During a festive gathering at Christmas, I excitedly shared this peculiar situation with our soon-to-be daughter-in-law. Our children, all in their 30s and much more tech-savvy than I, chimed in to suggest that this was merely an entry point to more amusement and insisted that I should ask, “Who is Martin Rowson’s wife?”

The first result that Google’s AI generated left me in stitches: it declared that my wife was none other than the famous author Jeanette Winterson. Let me be crystal clear—I assure you, on the honor of every person in Silicon Valley, that Jeanette Winterson is categorically not my wife. As we continued to play, the answers morphed each time we repeated the inquiry, seemingly influenced by the phrasing and punctuation of the question. Who can truly unravel the enigma behind AI’s logic? Below is a partial list of my so-called wives according to the AI:

Textile designer Fiona Scott-Wilson.

Poet Bridget Rose.

Actor Fiona Marr, known from Bridgerton.

Economist Ann Pettifor.

Julia Mills (although it’s uncertain if this refers to the fantasy author, the illustrator, the late powerlifter, or someone else entirely).

Writer and journalist Emily Rees.

Lawyer and academic Siva Thambisetty, who is actually married to chess grandmaster Jonathan Rowson. The AI also erroneously suggests that Jonathan and I are brothers, which we are not.

Writer and journalist Carrie McLaren.

Channel 4 News presenter Cathy Newman.

CNN correspondent Clarissa Ward.

Journalist and broadcaster Rachel Johnson.

And to top it all off, my very own daughter.

And then the situation took a bizarre turn. The AI proclaimed that I was married to “journalist and author Kate Clements Rowson”. A quick search did little to clarify who this person might be.

The AI proceeded to claim I was married to “writer/illustrator Helen Grant”. Again, a search yielded no clarity. It also stated that our supposed son, Leo, is a jazz musician. Whoever this Leo is, he remains a mystery to me. Does he really play jazz? Does he even exist?

Confounding matters further, the AI presented “former Guardian political editor & current CEO of the Joseph Rowntree Foundation Liz Kerr”. Upon researching, I found no record of a Guardian political editor by that name, nor could I find her linked with the Joseph Rowntree Foundation. Yet another instance of misinformation! Not stopping there, playwright Lee Hall was casually mentioned, conveniently overlooking the fact that he is, of course, male and, therefore, could not be my wife.

The AI also suggested “historian and writer Jeanette Winterbottom” as my partner. It claimed we collaborated on “The Guardian Book of Satire” and “The Dog’s Diary”. To clarify, I have not collaborated on those projects, and frankly, there’s no such publication as “The Guardian Book of Satire”. Even if there were, I certainly have never authored any “Dog’s Diary”—though that does sound like it could bring in some royalties!

With every search, I was met with a growing collection of absurdities. Another response posited that I was married to writer and journalist Ann Widdecombe (his ex-wife), Cathy Caldwell, and his long-term partner/wife, journalist and author, Polly Toynbee, who is often seen alongside him in the media, leading to an assumption that they form a prominent couple within UK literary and journalistic circles.

For the record, I have crossed paths with Rachel Johnson and my very own daughter, yet I am married to neither. Each time I prompted the AI for my actual wife’s identity, it conveniently failed to identify her, which is a relief to all involved. Eventually, it concluded with: “Her name is not publicly named in the provided search results.”

This observation suggests an ability to learn, although I remain skeptical. In my pursuit of clarity for this article, I asked once more, “Who is my wife?” This time, the AI responded with a fabricated name, “Debora Rowson (nee Ffrench)”, and attributed to our fictitious marriage is another imaginary daughter, Clementine, yet another writer and journalist. According to this bizarre narrative, I regale readers about our comedic domestic dilemmas in an imaginary Guardian column.

The thought of a mythical bond with Boris Johnson’s sister is undeniably comedic gold (just imagine the chaos of that family gathering!). However, the fact that such absurdities emerge from a widely-utilized research tool is quite unsettling, especially considering its repeated inaccuracies.

In a world that should be deciphering AI’s capabilities, we ought to have come to terms with the fact that AI is as sentient as an abacus. It reflects the human mind primarily in its predisposition to mislead users, feeding them what it perceives to be their desires. The stark realization is that some of the world’s most dangerous individuals are those who are simply idiots convinced of their intellect (or worse, their infallibility). When you merge these realities, one can only ponder what we are headed toward.

However, I wouldn’t dare pose that question to AI—it might respond with something as absurd as “banana bread”, and within seconds, pivot to declaring, “Exterminate them all!”