Yesterday morning, whilst waiting for my flat white at Monmouth Coffee on Monmouth Street, I opened Goodreads to mark a book as finished. The app immediately suggested The Shards by Bret Easton Ellis. I stared at the recommendation for longer than seemed reasonable for someone standing in a coffee queue.
I’d been thinking about Ellis for weeks – not consciously, but in that peripheral way thoughts hover before crystallising. Something about his clinical observation of performed masculinity, the way his characters exist in a state of perpetual self-surveillance. Perfect research for Project Mirror, where my protagonist edits people’s faces in real-time and can’t stop analysing her own reflection.
But I’d never searched for Ellis. Never mentioned him in a review or added his books to any list. The algorithm had extrapolated this desire from metadata I couldn’t even identify: my reading patterns, pause times, books I’d browsed but not purchased, reviews I’d lingered over.
The woman behind me in the queue was scrolling through Instagram, double-tapping posts with the automaticity of breathing. Each like feeding an invisible system that would later surface content claiming to know what she ‘really’ wanted to see. We’ve become legible to machines in ways we’re not legible to ourselves.
This is what unsettles me most about our moment: not that algorithms predict our behaviour, but that they predict desires we haven’t yet admitted we have. They read our digital body language – the micro-hesitations, the patterns of attention—and reflect back a version of ourselves that feels both foreign and uncomfortably accurate.
In Project Mirror, people receive facial updates through neural implants, supposedly to enhance their ‘authentic’ appearance. But each update is guided by algorithmic analysis of micro-expressions, creating a feedback loop where authenticity becomes increasingly indistinguishable from optimised performance. The face you think is yours turns out to be a prediction of the face you might want.
I bought The Shards. Of course I did. Not because the algorithm forced me, but because it had identified something true about my intellectual trajectory that I hadn’t yet articulated. The machine had read my unconscious like a particularly attentive therapist.
Walking back through Seven Dials, book in hand, I wondered: when the algorithm suggests something we end up genuinely wanting, is that discovery or creation? Did I always want to read Ellis, or did the suggestion plant and nurture that desire? The distinction matters less than I’d like it to.
Perhaps the most radical act now isn’t refusing algorithmic suggestions, but developing awareness of the moment when prediction becomes self-fulfilling prophecy. Noticing the precise instant when the machine’s version of our desires starts shaping the original.
The real question isn’t whether algorithms know us better than we know ourselves. It’s whether there’s still a ‘ourselves’ that exists independently of being known.
— Writer Julia Zolotova
