Abstract
The way we interact with computers has always shaped what we can do with them. Keyboards turned text into command, mice made pointing precise, and touchscreens put it all at our fingertips. Today we are building interfaces that operate on the body: closer to humans than ever before. As these interfaces grow more advanced, they carry the potential not only to expand what humans can accomplish but also to reappropriate power in an age of accelerating technology. Yet, the same technologies that empower can also control. The same data that enables autonomy can also subject us to new forms of oversight. The closer a device gets to the body, the more it demands trust and the more it reveals about what happens when that trust fails.
My technical project aimed to build a wrist‑worn gesture tracker capable of recognizing individual finger movements without requiring the user to hold or look at a device. The target capabilities were precise, low‑latency control: distinguishing between different finger presses and releases, translating them into standard computer commands, and doing so through a comfortable, wearable form factor. The solution fused surface electromyography, capturing electrical signals from the tendons that move the fingers, with a nine‑axis inertial measurement unit to track hand orientation. A machine‑learning model was trained on these multimodal signals to classify gestures with high accuracy, while the hardware was designed to integrate sensing and processing into a compact wrist‑worn package. The goal was to demonstrate that fine‑grained finger recognition could be achieved in a form factor closer to a watch than a laboratory instrument, opening the door to screenless, hands‑free digital interaction. From my technical project, I developed an understanding that engineering a wearable is not merely a technical problem. It is a negotiation with fragility, with the gap between what we design and how it lives on a person.
That understanding guided my STS research, where I examined the Dexcom G7 continuous glucose monitor. Like our gesture tracker, the G7 is a wearable that translates bodily signals into glucose readings so people with diabetes can manage their condition in real time. In 2025 the G7 became the subject of an FDA warning letter, a Class I recall, and a software correction. A component was changed without proper validation. A receiver’s alarms failed. An app silently stopped delivering data. Analyzing these failures through the lens of biopolitics, I argue they are not isolated engineering mistakes. They are symptoms of a system where regulatory frameworks measure safety through population‑level statistics while individual users bear the consequences that those statistics cannot capture. The same infrastructure that empowers patients to manage their own physiology also renders them subject to new forms of governance such as risk classification, data extraction, a form of care that measures success in averages while bodies fail in singular moments. The G7 reveals the paradox at the heart of digital health: we have built technologies that give people unprecedented control, yet that control is always mediated by systems they did not design and cannot fully trust.
Together, these projects taught me that human‑computer interfaces are never neutral. They carry a tension between empowerment and control, between autonomy and dependency. My technical work aimed to give users effortless control over their digital environment through a device that lives on the body; my STS work examined how such devices can also embed users in systems of governance they cannot easily escape. The closer technology gets to the body, the more it demands we consider not only what it enables but also what it asks us to surrender. If we want these interfaces to truly empower, we must design with intent and care: aware that every signal is a negotiation, every gesture an act of trust. The future of human‑computer interaction will not be measured by accuracy or latency alone. It will be measured by how well we close the distance between what a person needs and what a system provides.
