Abstract
On the one hand, automation of decision-making technologies may appear to be the only positive effect, given efficiency and optimization in the complicated setting. Nevertheless, both my capstone project and STS work demonstrate that as systems start to organize human decisions using data, they also transform autonomy in a manner that is not immediately apparent. My capstone project is to develop a Dynamic Schedule Planning Agent to assist students in managing academic loads, and my STS research explores the problem of ethical issues in sports performance tracking technologies. The technical project was one that I decided to do to solve a practical issue I have been having, which is dealing with deadlines in a variety of courses, whereas my STS research was inspired by the fact that I have a greater interest in the role played by data-driven systems in human decision-making and control. Despite the fact that one of the projects is academic, and the other one is athletic, both of them discuss the way in which engineering choices affect the way in which people interact with systems organising their behaviour. Collectively, they emphasize that the technical design is never neutral and should take into consideration its influence on user autonomy.
Dynamic Schedule Planning Agent deals with the issue of disjointed and inconsistent academic planning. Students tend to use various resources including Canvas pages and syllabi, and it is hard to have a clear picture of deadlines. The system addresses this by automatically deriving deadlines in syllabi and Canvas and creating a single calendar with a workload heatmap. It relies on natural language processing and extraction based on LLM to transform unstructured text into organized events and enable users to identify trends in their workload and predict the times of high stress. The point is not only organization, but better decision-making, to provide students with the opportunity to plan in advance and minimize stress at the last moment.
The findings of this project demonstrate that automation can greatly enhance the capacity of students to manage time and to know their workload. The test proved that the system was able to extract the majority of assignments and enable its users to find out quickly on the crunch weeks with the help of visualization tools. The heatmap allowed users to identify workload patterns in a few seconds, which is significant compared to checking various platforms manually. On the whole, the project shows that properly designed data-driven systems can increase user awareness and help them to be more effective in planning.
My STS study considers analogous data-driven systems, which are applied to a new area: sports performance monitoring. The main research question is how engineering decisions influence moral issues like surveillance, consent, and equity. The study is based on frameworks such as Actor-Network Theory, and the Social Construction of Technology to examine how tracking technologies gather and utilize data on athletes. These systems are aimed to maximize performance, yet they lead to power imbalances in which athletes do not have much control over the usage of their data. The study is important in that it demonstrates that the technologies aimed at making performance better can pose an ethical threat as well in cases when user autonomy is not given priority.
The results of the STS study indicate that these technologies are not only technical, but also sociotechnical systems with power relations. Athletes are not considered as sources of information but also as decision-makers, and the problem of informed consent and the accuracy of data is not properly considered. To illustrate, tracking systems may generate incomparable information, and there may be large variations among systems, which casts doubt on equity in assessment. On the whole, the study has found that ethical responsibility should be incorporated into system design, which should be transparent, equitable and user-controllable. Taken collectively, both projects reinforce the same main idea: data-driven technologies can either empower or weaken human autonomy based on their design and realization.
