Abstract
With recent strides in technology, including but not limited to the development of AI and LLMs, technology is encroaching more and more into the daily lives of ordinary people. In modern society, the drive to innovate can often push technology forward faster than our ethical understanding of it. Though not focused on the same topic, both my STS research paper and technical project aim to integrate robots and technology into daily life in a respectful and thoughtful way. The technical project is defined by its tight constraints, requiring the robotic device we built to fit within a model city and its similarly scaled down traffic regulations. My research paper, on the other hand, examines the potential costs and consequences of the widespread release of consumer humanoid home robots through the lens of the social relational model.
In the technical portion of my thesis, we created a small remotely controlled robotic rover (known as OSCAR) that is designed to navigate a model city while collecting and depositing simulacra of trash and recycling. The small scale of the city along with the wide range of features we hoped to include made the design and manufacturing of this device quite difficult. The final device includes a Mecanum wheel based omnidirectional drive system, a directable collection system with 180 degrees of rotational freedom, and a scissor lift depositing mechanism. In the design, the focus was on mechanical operation rather than electronic control through a microcontroller. For example, the collection module operates the grabbing and lifting though a single motor connected to a custom gearbox that changes configuration after making contact with a bin.
In my STS research paper, I investigated the phenomena of humanoid home robots that have begun being produced by several large companies. These devices are marketed as helpful assistants that can do chores or tasks, but much of the marketing also sells the idea of them as companions or even friends. By combining the social-relational model with the computers as social actors theory, we can analyze some of the potential consequences of the widespread adoption of this technology, and specifically the use of anthropomorphization in the design of these devices. The social-relational model breaks down social interaction into three distinct aspects: the user effect, the agent effect, and the context. By examining how subject traits affected their acceptance or trust in these robotic devices, the user effect provides concerning results: the most accepting users are those least familiar and most emotionally vulnerable. For the agent effect, the robot trait of anthropomorphization increases social bonding and intimacy with the users. Finally for the context, having embodied autonomous devices, powered by LLMs to enable convincing social interaction, increases not only user acceptance, but also trust, perceived security, and information disclosure. In summary, widespread adoption of this technology as it stands could give companies unprecedented access to both users’ physical and social lives.
As engineers, there is sometimes a feeling that innovation is the most important thing, that sometimes things should be created just for the sake of creating. Unfortunately, when technology outpaces our understanding of its consequences, people get hurt. Without constraining a full size autonomous trash collection robot to safely follow given restrictions, it could easily damage or harm other vehicles or pedestrians. Similarly, if home robots are given unrestricted access to the social lives of vulnerable users, corporations could easily take advantage of that situation for their own gain. STS strives to mitigate these problems by teaching current and future engineers to consider the impact their work could have beyond just the surface level. As engineers, we can’t make mistakes and rely on others to catch or fix them, the onus is on us as the designers and creators to create things to make the world a better and more just place.