Joyce Ng shows us the commuter through the eyes of machines

Artists, Case Studies | Published on | Last updated on

3 min
Joyce Ng shows us the commuter through the eyes of machines

For Human Error Rate, Joyce took videos of commuters from around the world and prompted an AI model to mirror their behavior

Meet the artist

Under the name alsoguppyme, Joyce creates work that transforms the ordinary into the surreal. As an AI artist, she doesn’t want to use technology to escape reality or smooth it over. The way she sees it, technology can be used as a tool for us to look more closely at the world we already live in.

Before she started working with generative AI, she trained as a painter and spent decades working in commercial creative roles. As an artist, she’s always been eager to explore and iterate with all the tools at her disposal. And this curiosity continues to shape how she works in the AI art space today.

“I started putting work out there, and somehow it became natural—something I couldn’t stop doing.”

Human Error Rate grew out of Joyce’s own experience with commuting. When she was fifteen, she crossed the Malaysia–Singapore causeway every day to get to school. The journey was slow and crowded. People stood shoulder to shoulder for hours

“That sea of helmets on the causeway is something I still can’t shake off.”

For Joyce, she sees trains and subways as places where people stop performing and disappear into themselves. These quiet moments of everyday behavior became the starting point for Human Error Rate.

“The side-by-side, before-and-after format is simple, but it’s powerful—it starts a conversation without being didactic.”

Inside the creative process

Each work in Human Error Rate is presented in two parts. One shows the original human footage. The other shows how an AI system interprets that same moment. Joyce chose this structure for its clarity. Placed side by side, the contrast makes the gap between everyday human behavior and machine perception immediately visible.

To create the work, Joyce moved fluidly between Leonardo and other tools, pushing images through different states—still frames into motion, motion into 3D forms, and back again. The process was intentionally exploratory.

Column image
Column image

I don’t want to be didactic. I don’t want to tell people what to think. They can take it as entertainment or they can sit with it and ask questions. My role is just to share what I’m seeing.

Sometimes the AI model captured movement with precision. Other times, it fell apart. For Joyce, these weren’t technical errors. They were insights into how machines see the world.

Pushing their craft forward

Joyce is careful not to frame Human Error Rate as a story about humans versus machines. What she’s drawn to instead are the moments when AI systems mimic human behavior in an entirely unexpected way

“This isn’t AI versus human or AI bashing. It’s about asking how much AI truly understands us.”

By focusing on ordinary moments, Joyce invites viewers to consider how easily the human experience can be reinterpreted when people are reduced to data. In showing this, Human Error Rate reflects how Joyce sees art as a way of paying attention.

Watch: Human Error Rate

. . . .