Here’s a shocking truth: Artificial Intelligence (AI) might be the most overhyped tool of our time, and its failures can be both hilarious and deeply unsettling. But what happens when AI tries to identify something as personal as your spouse? Let me share a story that’s equal parts amusing and alarming.
Recently, my family and I stumbled upon a bizarre game—one that anyone with an internet connection can play. We’ve jokingly dubbed it, “How Clueless is AI, Really?” Here’s how it works: Ask AI a simple, personal question and watch it spectacularly miss the mark. In my case, the question was, “Who is Martin Rowson’s wife?” What followed was a comedy of errors that revealed more about AI’s limitations than I ever expected.
And this is the part most people miss: AI isn’t just occasionally wrong—it’s consistently wrong, and in ways that can be both absurd and concerning. Let me explain.
I’m a cartoonist with a sizable online presence, but my wife, whom I married in 1987, has deliberately stayed off the grid. So, when you search for “Martin Rowson wife”, the results are a wild mix of misinformation. Google’s AI has, at various times, claimed my wife is:
- Jeanette Winterson, the renowned author (spoiler: she’s not).
- Fiona Scott-Wilson, a textile designer.
- Bridget Rose, a poet.
- Fiona Marr, an actor from Bridgerton.
- Ann Pettifor, an economist.
- Julia Mills, whose identity remains a mystery.
- Emily Rees, a writer and journalist.
- Siva Thambisetty, a lawyer and academic (who, by the way, is married to chess grandmaster Jonathan Rowson—no relation to me).
- Carrie McLaren, another writer and journalist.
- Cathy Newman, a Channel 4 News presenter.
- Clarissa Ward, a CNN correspondent.
- Rachel Johnson, a journalist and broadcaster.
- My own daughter (yes, you read that right).
But here’s where it gets controversial: AI even invented entirely fictional characters, like “Kate Clements Rowson” and “Helen Grant,” claiming they’re my wife and that we have a son named Leo, a jazz musician. None of these people exist—at least, not in my life. And yet, AI presented them as fact, complete with fabricated details about our lives.
What’s more, AI’s mistakes weren’t just random. They seemed to depend on how the question was phrased or punctuated, adding another layer of unpredictability. At one point, it even suggested I’m married to Ann Widdecombe or Polly Toynbee—both prominent figures in UK media, but neither my wife.
Here’s the kicker: Despite its repeated failures, AI occasionally showed a glimmer of self-awareness, admitting, “Her name is not publicly named in the provided search results.” But don’t be fooled—this isn’t evidence of learning. When I asked again for this article, it confidently declared my wife is “Debora Rowson (nee Ffrench),” a retired civil servant, and threw in an extra daughter named Clementine for good measure. None of this is true.
While the idea of me being married to Boris Johnson’s sister is laughable, the broader implications are anything but funny. AI is touted as a revolutionary tool, yet it’s as reliable as a broken compass. It doesn’t think, reason, or understand—it simply regurgitates patterns from its training data, often with disastrous results. As a satirist, I find this both hilarious and terrifying. AI’s ability to fabricate reality is a stark reminder of its limitations—and the dangers of trusting it blindly.
Here’s a thought-provoking question for you: If AI can’t even get something as basic as a person’s spouse right, how can we rely on it for more complex tasks? And what does this say about our growing dependence on technology that’s fundamentally flawed?
Let’s keep the conversation going. Do you think AI’s mistakes are harmless fun, or a sign of something more troubling? Share your thoughts in the comments—I’d love to hear your take.
