Enhancing Talent Discovery: An AI Chatbot Solution; Ethical Decision-Making Frameworks in Autonomous Vehicles

This thesis combines two projects: a technical project that is based on improving workforce talent discovery through an AI-driven chatbot, and an STS research project based on integrating ethical decision-making models into autonomous vehicles. Although the technical project directly addresses an application of artificial intelligence during my internship, my STS research analyzes the interaction of artificial intelligence with ethical values. My interest in pursuing these interrelated topics is driven by an understanding that the effectiveness and societal acceptance of technology are not just dependent on its capabilities but also on the ethical principles governing its use.

Organizations are often at a dilemma with analysis and characterization of different talent pools, which leads to problems making informed decisions while hiring. During my internship last summer, I developed an AI-driven chatbot that, through many key metrics such as education trends, hiring rates, gender diversity and others, helped provide workforce insights. It streamlined the discovery of talent for hiring managers in one go so they could instantly access and act upon high-value insights much more efficiently. The chatbot significantly increased recruiters' access to actionable information, transforming dense workforce data into easy-to-understand visuals and summaries. Early testing showed hiring managers could easily acquire valuable talent insights, greatly reducing the amount of manual effort required beforehand. Despite this progress, future developments should address deeper system integration, expand dataset coverage, refine visualization tools, and perform extensive user testing. Ultimately, this project demonstrated AI’s powerful role in simplifying decision-making, promoting efficiency, and reducing bias in organizational hiring processes.

Autonomous vehicles are revolutionizing transportation in this current time, but their integration into society comes with serious ethical concerns. The central problem is how to instruct or program AVs to make decisions in a matter of a second in cases of high danger that can be considered ethical. This research delves into concepts of ethical decision-making and what is coded within AV software. I will be asking the question, “How must AVs be programmed to make ethical driving decisions in cases of risk and harm distribution”? To study this issue, in this research the Social Construction of Technology (SCOT) and risk management frameworks are employed to consider how society's values influence AV decision-making and how risk is quantified in real-world uses. Through case study, literature review, and a survey, this research will compare ethical frameworks such as utilitarianism and deontology while also observing how they influence AV behavior. This research will show the ethical implications of AV programming, roles of regulatory policy, and possible bias in decision-making algorithms currently. This research aims to connect technical and ethical concerns to ensure that AV systems reflect the values of society and promote safety.

Working on both projects was important in building a picture of the applications of artificial intelligence. The technical Capstone was concerned with emphasizing the practical use of creating effective AI systems and user-centered design solutions while the STS research further developed my awareness of more general ethical responsibilities and social impacts associated with such technologies. Both methods at the same time established a connection between them that I might have not been able to identify had I worked on them separately. The use of ethical theory into the design of artificial intelligence systems reinforces the idea that successful technological solutions must include societal values and ethical expectations. Working on both projects highlighted the need for transparency, accountability, and public engagement in technological progress. The STS study encouraged me to actively consider potential biases in data-driven systems. Lessons learned from studying algorithmic biases in AV ethics connects with the mission of the AI chatbot project; to attempt the removal of biases by recruiters in the hiring process. These projects show that engineering successful technologies such as autonomous vehicles or AI chatbots requires the consideration of social and ethical aspects. The skills acquired through this research experience, both technical and ethical, will affect my approach to future engineering endeavors.

School of Engineering and Applied Science

Bachelor of Science in Computer Science

STS Advisor: Bryn Seabrook