UC San Diego

From fine-grain filtering of noise required to selectively attend to a specific person in a social setting, to the rapid visual pattern recognition required for complex tasks like radiological diagnosis, human perceptual and cognitive systems are capable of incredible things. Built up over years of evolution in a physical world, people have developed wonderful abilities. However, these capabilities begin to show their limits as society enters into an increasingly digital world, unbound by physical limitations. Human attention, perception, and cognition do not scale up as the volume of information scales up, and given the rapid explosion of digitization of modern work and life, we are simply not able to keep up. Thankfully, this explosion of digital data has co-occurred alongside advances in sensing hardware that can 'see' and 'hear' as people do. Furthermore, advances in computing technology now allow machines to detect signals and recognize patterns in these data in a way that is similar to how trained people can. This presents a new and untapped opportunity for humans to benefit from computing in ways that can help them sense things that happen in the world, develop an understanding of patterns in large volumes of data, and augment our collective capabilities to scale up in the digital age. This dissertation recounts several works (i.e. Lab-in-a-Box, ChronoSense, GutCheck) that seek to better understand how computing can augment the human ability to sense interpersonal behavior and scale up methods for understanding social interactions. I showcase the potential for these approaches through their ability to surface new insights in the domains of human factors, communication, and medicine. Through the lens of applied research and human-centered design, I use these works to argue for an impending revolution in human-computer teamwork, showcasing the promise of computing to evolve from being just a tool to becoming more of a mediator and teammate within group activities. These technologies are not without their unintended consequences, and I close with a discussion of future work needed to proactively mitigate adverse outcomes.