Snoozer

Through the rise of cameras, microphones, application trackers, mouse/key loggers, and facial recognition, technology has been increasingly used for digital surveillance in academic and work settings. This panopticon environment raises concerns about privacy and mental wellbeing. While surveillance can deter unwanted behavior, it encourages performative rather than genuine productivity, and even exacerbates underlying systemic issues like hustle culture that result in overwork and unproductivity in the first place.

I wanted to raise the question: what are the consequences of new AI/ML technologies for digital surveillance—who are they really benefitting? When they solve problems, are they treating symptoms or the cause—or even contributing to them?

Then, I asked: how can I build an application using AI/ML technology that makes things…objectively worse? ‍And thus, Snoozer was born.

Personally, I struggle with attentiveness and maintaining a consistent sleep schedule, which often manifests in me nodding off in front of my computer during a boring lecture or meeting. I wanted to design something that would force me to stay focused by catching me and stopping me when I start drifting away.

Sleepy user -> Snoozer sees you! -> Snooze a little too long... -> Rude awakening.

I took photos of myself in various "asleep" and "awake" states and uploaded them to Teachable Machine. Given the scope of the project, I was unable to amass a giant dataset of photos of faces and label them all. Instead, I curated a very clean dataset of my own facial expressions. While it was functional in certain contexts, it was not universal enough for practical use. I deployed this version on Glitch, using p5.js for my webcam processing and interface. This setup was unable to be overlaid on other websites, however.

The vector art plays on the Corporate Memphis style—bland, inoffensive, and commonly used in modern business and tech. By presenting surveillance and data collection as benign and essential for “workplace efficiency”, corporations obscure the ethical and privacy concerns associated with such technologies.

First-time users were not told that there would be a jumpscare, but instinctively understood that leaving their eyes closed for long enough to fill up the sleep meter was a negative behavior. This anxiety exacerbated by the fact that they couldn’t see the meter increase while their eyes were closed, so they had no idea how much “time” they had left, causing the impulse to open their eyes to repeatedly check.

“Oh my god, I couldn’t fall asleep if I wanted to, but I can’t concentrate on my work either! I’m so scared something bad is gonna happen.” -participant quote

I deployed Snoozer on the Chrome Web Store and showcased it at the 2025 UT Design Capstone Exhibition.