The Stats Guy: The AI trap and falling in love with ourselves
AI is already shaping how we interpret the world, what we see, who we trust, and how we define truth, writes Simon Kuestenmacher.
Spoiler alert: This week’s column is centred around a great book that I will spoil for you in the first paragraph. Bookmark the column now, read or listen (only 90 minutes) to the book before reading on at your own risk.
More than 200 years ago, long before artificial intelligence or humanoid robotics were a thing, German author E.T.A. Hoffmann warned us of their dangers.
His 1816 short story Der Sandmann is a gothic tale of obsession and delusion. It’s arguably the earliest cautionary fable about falling in love with machines and is the basis on many of your favourite sci-fi movies.
In this classic German Romantic tale, the protagonist Nathanael becomes obsessed with Olimpia, who appears to be the beautiful and enigmatic daughter of a professor and a watchmaker.
She is quiet, poised, and always listens attentively. When Nathanael gazes into her eyes, he is entranced. He falls deeply in love with her before discovering that she is an automaton, a lifelike mechanical doll created by the sinister inventor Coppola and the professor Spalanzani.
So, all Nathanael saw when he gazed into Olimpia’s eyes was the reflection of his own eyes. That’s the moment he fell in love. Not with Olimpia, but with himself.
You might like
Der Sandmann explores themes of obsession, madness, and the uncanny (what Freud later called das Unheimliche – yes, after my recent Carl Jung column I felt I owed Sigmund a mention too), and it had a profound influence on later science fiction and psychological literature.
Der Sandmann predicted our AI dilemma more than 200 years ago. Photo: E.T.A. Hoffmann
I think of Der Sandmann as one of the most insightful depictions of artificial intelligence written, precisely because it was never about technology. It was about the human tendency to fall in love with our own image, to project meaning where there is none, and to become seduced by certainty and control.
Der Sandmann has quietly shaped some of the most powerful modern works about AI and human-machine entanglement. You can trace its DNA in Blade Runner, Ex Machina, Her, Westworld, and The Stepford Wives.
Even novels like Kazuo Ishiguro’s Klara and the Sun and Philip K. Dick’s Do Androids Dream of Electric Sheep? revisit Hoffmann’s core questions: What makes us human? Can machines reflect us too perfectly? And what happens when we prefer the reflection to reality?
Stay informed, daily
This brings us to today’s artificial intelligence.
When used carelessly, AI tools can easily become flattering mirrors. They answer our questions, but only in the way we ask them. Ask the wrong question, and you’ll get a wrong (albeit perfectly confident) answer. Ask a leading question, and it will confirm your beliefs.
Echo chambers are no longer confined to social media – they’re creeping even deeper into our daily lives.
An AI echo chamber is even creepier than just being fed posts that align with our worldview by Facebook or X. Artificial intelligence feeds us information that sounds perfectly neutral and thereby covers just how much our questions reflect our worldview to begin with.
The risk is that if we treat AI as a mirror, it will show us only ourselves. If we learn how to treat AI as a lens, it can widen our field of view.
AI is already shaping how we interpret the world, what we see, who we trust, and how we define truth. It’s critical we use it in ways that challenge our assumptions rather than reinforce them. Otherwise, like Nathanael, we may one day wake up to find we’ve fallen in love with an illusion of our own making.
You’ve probably seen the news stories where real people ended up falling in love with their AI companions, something that as we still mocked in 2013 when we watched Her. There shouldn’t be a shortage of such stories in the future.
The plot of the movie Her is looking increasingly realistic. Photo: Warner Bros
When I urge policy leaders and educators to push for AI literacy in schools, I want students to learn how AI works, what it is and what it isn’t. How to use AI, but most importantly I want students to be able to recognise the reflection of their own eyes.
Simon Kuestenmacher is a co-founder of The Demographics Group. His columns, media commentary and public speaking focus on current socio-demographic trends and how these impact Australia. His podcast, Demographics Decoded, explores the world through the demographic lens. Follow Simon on Twitter (X), Facebook, or LinkedIn.
