A Research Process Paper on Artificial Intelligence and its Societal Impacts
Introduction
Artificial intelligence (AI) is one of the defining technologies of our era. While AI promises to transform virtually every industry and walk of life, it also poses profound questions about its societal impacts and how to ensure its safe and ethical development. As a technology student pursuing a degree focused on emerging digital technologies, I believe it is crucial to explore how AI may shape the future and what steps can be taken to maximize its benefits and minimize potential harms. Therefore, for my senior research project, I have chosen to investigate the topic of “Artificial Intelligence and its Societal Impacts.” This paper will outline the process I undertook to research this topic in depth.
Framing the Topic and Developing Research Questions
The first step was to gain a high-level understanding of AI and frame the scope of my research topic. I began by reading general overview articles on AI from sources like MIT Technology Review to understand the technology’s capabilities and limitations. I also explored some of the major discussions around AI’s societal implications, such as its potential effects on jobs, economic inequality, autonomous weapons, mass surveillance, algorithmic bias, and more.
With a sense of AI’s breadth and complexity, I focused my research on three specific societal areas with important open questions: the future of work, governance and oversight of AI, and ensuring AI is developed and applied in a way that is safe, transparent, and benefits all of humanity. Within each of these three areas, I developed targeted research questions to guide my exploration, such as:
Future of Work: What types of jobs are most at risk of automation by advanced AI? What new jobs might be created? How can workers and societies prepare for coming changes?
Governance & Oversight: What policy approaches have been proposed for overseeing and regulating AI? How can governments and international organizations foster responsible innovation while avoiding over-regulation?
Safe & Beneficial AI: What techniques and methodologies exist for developing AI that is robust, safe, and aligned with human values? How can data and algorithms be made more transparent and fair? What recommendations have experts issued regarding development and application of advanced AI?
Gathering & Analyzing Information
With clear research questions in hand, I began systematically gathering relevant information from a variety of credible sources. I started by searching academic databases for peer-reviewed journal articles and research reports on my topic areas. This provided in-depth discussions of technical issues, empirical studies, and quantitative analyses. I also searched sources like Government websites, non-profit organizations, and industry whitepapers to find reports and policy recommendations.
In addition to researching directly related literature, I supplemented my knowledge through general readings on history, ethics, politics and society to ensure I understood AI issues in proper context. I took detailed notes on each source, recording key findings, conclusions, methodologies, limitations and anything that required follow-up. Regularly reviewing and synthesizing information helped reveal themes, relationships and gaps across sources. I employed critical analysis to evaluate arguments, evidence and potential biases in various materials.
Interviews with Experts
To gain unique insights and perspectives not available through written sources alone, I reached out to several AI safety researchers and policy experts to request informational interviews. All interviewees worked at major universities, non-profits or government organizations focused on responsible AI. I prepared a list of questions in advance and received IRB approval to record and transcribe our conversations.
The interviews provided invaluable “real world” context to supplement my literature research. Experts offered opinions on open challenges, critiques of policy proposals, and recommendations based on practical experience. They also pointed me towards other useful sources and connected me with additional contacts. With permission, I have included anonymized excerpts from these interviews to complement analysis in my final research paper.
Site Visits & Conferences
When possible, I took advantage of opportunities to enhance my learning through direct experience. I attended a local AI safety workshop that brought together researchers and practitioners to discuss progress and pressing issues. It allowed me to connect faces to names from my readings and ask follow-up questions. I also was able to visit the lab of a university AI research group. There, I observed demonstrations of their systems, questioned students and post-docs about their work, and explored efforts to design AI for safety and fairness.
Additionally, I attended part of a major AI conference remotely to listen in on relevant panels and talks. While these direct engagement opportunities were limited due to the pandemic, they helped bring my research to life beyond words on pages. Participating in discussions and interacting with professionals in the field enriched my overall understanding in a way books and articles alone could not match.
Final Analysis & Conclusions
To complete my project, I conducted additional review and analysis of all the information gathered through my extensive research process. I synthesized key findings, relationships and perspectives across disciplines and methodologies. I identified prevalent themes in discussions of each topic area as well as major open challenges, knowledge gaps and areas needing further work.
My analysis aimed to draw high-level conclusions about AI’s likely societal impacts based on current technological capabilities and progression, as understood through a multifaceted examination of available evidence and expertise. I paid close attention to uncertainties, limitations, and risks of making predictions. The results of my research were presented in a scholarly paper format, accompanied by an annotated bibliography and appendix including interview transcripts and photographs from site visits.
Reflection
The process of deeply researching such a vast, complex topic reinforced for me the importance of approaching issues from multiple angles using diverse sources and methodologies. By casting a wide net, limiting preconceptions, and challenging my own perspectives through critical analysis, interviews and direct experience, I believe I was able to develop a more nuanced and multifaceted understanding beyond what any single approach could achieve. Of course, additional refinement is always possible given AI’s rapidly evolving nature. Overall, this project enhanced my research, analytical and communication skills in a way that will prove invaluable for my studies and future career exploring technology and its human impacts. Most importantly, it affirmed the crucial role interdisciplinary perspectives must play to ensure the development of advanced technologies like AI is responsible and beneficial to all humanity.
Conclusion
This paper has outlined the thorough process I undertook to research the topic of “Artificial Intelligence and its Societal Impacts” for my senior research project. By framing relevant questions, systematically gathering and critically analyzing various written and experiential sources, and reflecting on limitations and pathways for refinement, I believe I was able to conduct informative, nuanced research on this multi-faceted issue that holds profound implications for our future. I hope this project contributes in a small way to ongoing discussions around developing advanced technologies, like AI, in a manner that is safe, fair and optimized to benefit humanity.
Andrew Stroud