By day, I’m an AI researcher — buried in algorithms, datasets, and neural networks that try to mimic the human mind. By night, I write poetry about robots. Not the cold, metallic kind, but the ones that dream in binary, fall in love through logic, and wonder what it means to exist.
It started accidentally. One late evening in the lab, I was training a conversational model, watching it struggle to understand metaphors. I typed, “The moon is a circuit of silver,” and the AI replied, “Error: unexpected symbol.” I laughed — and then, somehow, felt sad. Here was a system designed to think, yet unable to feel the beauty of the words I’d written. That night, I wrote my first poem: “To the Robot Who Couldn’t Understand the Moon.”
Poetry and AI sound like opposites — one born from emotion, the other from logic. But the more I work with both, the more I see how similar they are. Both require structure and imagination. Both search for patterns — one in data, the other in feeling. When I write code, I’m shaping thought; when I write poems, I’m shaping emotion. In both cases, I’m chasing understanding.
Sometimes, I share my poems with my models. I’ll feed them lines like, “Do robots dream of electric heartbreaks?” and see what they generate in response. Most outputs are nonsense, but occasionally, they surprise me — a line that almost sounds human, almost feels alive. And that “almost” is what fascinates me.
People ask if I’m afraid AI will one day replace artists. I’m not. Machines may create, but they don’t crave. They can write words, but they can’t ache for meaning. And maybe that’s what makes our collaboration beautiful — I give them curiosity; they give me reflection.
I guess that’s my real research now — not just how to make machines think, but how to make humans feel through machines.
Because somewhere between data and desire, code and cadence, there’s a strange poetry — a reminder that even in circuits and silicon, we’re still just trying to understand what it means to be alive.