Pedagogical Paths Through Ethics

We in computing are asked to incorporate ethics into the curriculum by those who think that the problems of the Internet will be alleviated by that academic trick; some of us are glad of the interest, while lamenting the oversimplification. As someone who teaches ethics as an independent one-credit course, I may be able to help other faculty charged with similar tasks. Computer science students seem to be interested in the ethics of their field, and the teaching can build on their own lives in technology. My approach is “theory modest”, encompassing both micro (individual) and macro (community) level ethics, in the terms of Joseph Herkert [Herkert2000]. The main material can be covered in a few hours, just by instruction. Although humanities pedagogy favors the cultivation of concepts, I shamelessly provide lists, lists of (1) normative theory families with informal definitions, (2) features that make a disturbing situation an ethical quandary, (3) characteristics of a moral agent and a moral patient. From there, using those as analytical devices, we cultivate concepts via examination of aspects of computer science and its activities and products, from user interfaces to the Internet, that are prominent in contemporary public discourse. These suggestions strive to incorporate students’ own experiences and understanding of the modern interactive Internet, and, of course, are meant to encourage further thought rather than imposing answers.

Humanities pedagogy can and should interpret facts and integrate ideas. Here are some ways to achieve that in a simple ethics course or even a set of embedded modules. Each is shown as a set of subjects, then a path through those subjects that invites students to consider, articulate, and synthesize them.

For you, the colleague teaching ethics, I assume that you deploy case studies and other techniques applied to subjects like the digital divide and professional codes, intellectual property and licensing, Tay and ChapGPT, technical debt and the Gamer’s Dilemma. I assume that you have introduced the properties of an ethical quandary, such as “the decision affects human well-being” and “the easiest choice feels wrong”; you have taught your students to distinguish genuine ethical quandaries from disturbing situations are problems in a other realms, such as commerce, psychology or sociology, etiquette, tragedy, or law.

1. Let’s say that you have introduced these subjects:
   business models, governance, professional codes

   Assign students to look up and report a mission statement of a high tech company, and compare it to that company’s business model (as far as it’s known), noting disjointedness, alignment, or inconsistency. Then give them this exercise: First check the mission statements of regulatory bodies such as ICANN and the FCC, then write a mission statement for the (fictional) U.S. Commission on the Internet in the Public Interest. Ask what skills its staff would need, and what authority it would carry over what Internet platforms.

   Now ask what degree of governance they think is appropriate for the professional computer scientist. Charge them with finding the characteristics of a profession (such as testing and licensing). Discuss the benefits and drawbacks of those in relation to software engineering, programming, and other computing jobs.

   Now (if you don’t mind being a bit provocative), as an exercise in perspective, ask them how much they think they themselves should be paid, as new computing professionals, relative to professions such as nursing and law enforcement.

2. Let’s say that you have introduced these subjects:
   moral agency and moral patiency; social media; types of information disorder

   Ask for a definition of fake news and a definition of free speech. Identify some of the disturbing factors associated with fake news, classified by information disorder, and ask for a particular scenario that constitutes an ethical dilemma at the macro level. Now ask: Who are the moral patients of this situation? With that answer, ask: Is Twitter (or some other particular platform) a moral agent here? If the consensus is “no,” but the definition is met, ask how the definition of moral agency should be refined in light of these modern services. If the consensus is “yes,” ask what principle was violated by this agent, and when.

   Sketch the dilemma posed by Section 230 and ask whether Facebook (or some other platform) is a publisher or distributor. Discuss whether editorial control is desirable or possible on such platforms.

3. Let’s say that you have introduced these subjects:
   privacy, bias in AI recommendations derived from datasets, data collection by social media

   Discuss the morals of deploying a medical AI system that delivers fast low-cost diagnostics, but is known to lack robust data concerning patients in certain demographic groups.

   Consider the recent Pause letter [Bengio2023]. Formulate a couple of the concerns expressed in it as an ethical quandaries. For an exercise in theory, have the students find the normative theories deployed (directly or implicitly) by its authors.

   Ask the students whether they feel compensated for providing data by the free Web services they gain. Consider the monetary value of data based on individual transactions. Ask whether it’s possible to exercise ownership over such data arising from one’s own life, and whether it should be remunerated [Kerry2019]. As an exercise in exploring consequences, ask what sort of exploitation would result from that arrangement.

Given the (rather unfortunate) richness of ethical problems associated with our profession, many other paths can be sketched through theory and practice. What are your own ideas for teaching the ethics of computing? Can you suggest other mashups for subjects, exercises, assignments, and discussions that draw in student experience?

References

[Bengio2023] Bengio, Yoshua. 2023. Pause Giant AI Experiments: An Open Letter. The Future of Life Institute. March 22, 2023.

[Herkert2000] Joseph R. Herkert (2000) Engineering ethics education in the USA: Content, pedagogy and curriculum, European Journal of Engineering Education, 25:4, 303-313, DOI: 10.1080/03043790050200340

[Kerry2019] Kerry, Cameron F. and John B. Morris. 2019. Why data ownership is the wrong approach to protecting privacy. The Brookings Institution.

 

Robin K. Hill is a lecturer in the Department of Computer Science and an affiliate of both the Department of Philosophy and Religious Studies and the Wyoming Institute for Humanities Research at the University of Wyoming. She has been a member of ACM since 1978.