It feels like every day we learn about a new way AI could upend our lives. From threats of the technology replacing jobs en masse to invasive new facial recognition systems, AI and computer ethics have become national issues. As a response, some elite universities have implemented programs in ethical computing directly into their computer science pedagogy, such as EthiCS at Harvard and Socially Responsible Computing (SRC) at Brown. Incorporating ethics into the curriculum is an adaptation to the increasing value of computer science skills across job fields and growing concerns about issues like algorithmic bias, data misuse, and the environmental impact of AI.
Princeton has dipped its toe into ethical computing through the Dialogues on AI and Ethics research program and a couple select courses — COS 350: Ethics of Computing, and COS 109: Computers in Our World. However, ethical computing has not been formally incorporated into the main intro COS sequence (COS 126, 217, and 226), which is the foundation of further computer science study and popular with students across disciplines. To address ethical concerns about machine learning and respond to the growing relevance of computer science to a variety of fields, Princeton’s Department of Computer Science (COS) should incorporate lessons, discussions, and readings about ethical computing into its introductory sequence.
A myriad of students across departments partake in the intro COS program. This spring, 263 people are enrolled in COS 126: Computer Science — An Interdisciplinary Approach, and over the past four semesters, the course has consistently enrolled over 250 students. And according to the COS department, over half of Princeton students in previous years — representing all majors — have taken the course. Since the intro COS sequence is popular, it has a platform. Let’s use it to craft a holistic, ethical education that helps students responsibly use computer science across industries.
In some ways, it seems that Princeton’s COS program has already recognized the importance of catering the intro COS sequence to majors and non-majors alike. After all, the first course most students take, COS 126, is subtitled “An Interdisciplinary Approach.” As someone who took COS 126 last semester, I did enjoy how the course’s assignments utilized concepts from fields like music and linguistics to teach students computer science concepts.
But equally important as applying computing to various disciplines is recognizing the broader ethical impact of these applications in a societal context. In this regard, I found COS 126 missed an opportunity. Two of the course’s assignments directly related to ethical computing issues of today: Students implemented a primitive machine learning algorithm as well as a basic language model built on principles similar to ChatGPT.
As someone already interested in computer ethics, these assignments exemplified how the “learning” and “thinking” we ascribe to machine learning differs greatly from that of humans. But I had to connect those dots on my own. By discussing these concepts in lecture or precept, the COS program could involve all students in an interactive examination of how deeply ethics is enmeshed in computer science, contextualizing class concepts within the various fields students may pursue in the future.
The perils of COS without ethics have become even clearer recently. Let’s face it: The presidential administration currently attempting to curb civil rights and liberties in favor of investing in AI and other technology is heavily influenced by Mark Zuckerberg, Elon Musk, and Sam Altman, three men from elite universities who either pursued computer science in their studies or outside of school.
Altman has even been on the record saying that the advent of artificial intelligence could “require changes to the social contract,” implying that humans should adjust to AI, rather than the other way around. Princeton has a responsibility to avoid producing students with a similarly monomaniacal view of technology’s importance — and enhancing the intro COS sequence is a good first step.
Some would argue that, with an already large COS curriculum and the department’s reputation for difficult courses, implementing ethics-centered coursework would only overwhelm students and hinder their learning. But the successful efforts of other universities demonstrate that inclusion of computer science ethics enriches and balances the curriculum.
At Brown, the SRC program has a list of courses — including introductory ones — that have incorporated ethical content, such as mini-lectures, class discussions, and extra credit assignments. At Harvard, the Embedded EthiCS program offers a variety of weekly teaching labs for lower- and upper-level courses alike, with relevant content tailored for each. Both of these programs engage with AI ethics in a way that naturally follows from course content, rather than tacking on extra coursework.
Princeton’s COS department can and should follow the lead of its counterparts at other universities. Just like a special lecture was dedicated in COS 126 to Machine Learning, someone versed in computer science ethics should visit intro courses once or twice to explore the ethics related to each class’s topics. A few precepts per semester could also be designated to ethical discussions, given that they already serve the purpose of expanding upon course content. And in the case of COS majors, perhaps Princeton should implement an ethical computing class requirement to receive a degree. Ultimately, this should be an easy choice for a university like Princeton which prioritizes the liberal arts and, especially, the service of humanity.
No practice or discipline exists in a vacuum, and that’s important to recognize now more than ever. To meet the moment and its own principles, Princeton must expand the content of introductory COS programs to address the ethics of computing. Ethics doesn’t have to be an afterthought or tangent of our computer science education; we have the tools we need to apply advanced computer science concepts to a variety of disciplines while remaining explicitly and boldly humanistic in the face of the evolving technological horizons of today.
Get the best of the ‘Prince’ delivered straight to your inbox. Subscribe now »
Shane McCauley is an assistant Opinion editor from Boston intending to major in Anthropology (and minor in Computer Science). He can be reached at sm8000[at]princeton.edu.
