Assignments

All weekly assignments must be done individually. If you wish to use material from either the course readings or readings you find yourself, please make sure to cite properly.

Each week, there will be an assigned topic for that week's writing; these will also appear in the course Slack page. Your submissions will be due to the LEARN dropbox for the course by 11:59 on the Mondays before the Tuesday class meetings. Students must complete six of these essays in the term; see the grading pane for more details.

Late essays will not be accepted except under extreme circumstances; please contact the instructor in those situations. Otherwise, please just respond to the next submission prompt.

Example papers that have been previously submitted to courses similar to this offering of CS 492 are available on the course's LEARN site, under the Sample Papers tab under Content.

#	Due Date	Subject	Prompt
1	16 Jan	Introduction and case studies	What is a case of ostensibly value-neutral technology that might have a different impact on a historically disadvantaged group? How would you assess that impact?
2	23 Jan	Algorithmic bias: theory	Suppose you were hired as a consultant to Luka, the company that owns Replika. You have been brought in because of concerns about Replika's sexist or inappropriate behavior with customers. What changes--technical or otherwise--would you recommend that Luka make to the product?
3	30 Jan	Race and the criminal justice system	Choose 1 of: Suppose you were the IT manager at a university that was going to put in place proctoring software like the ones described in the article in this week's reading. What would you advise your bosses to do to ensure that the concerns in that article were being taken seriously? Which, if any, of the concerns would you not worry about? Would you prefer a system like COMPAS to a potentially biased judge? What research would you do to identify which of these is a better approach?
4	6 Feb	Indigenous people and technology	Some of the Maori activists quoted in the data sovereignty paper argue that their own community should be able to control access to data about their language and culture. Others argue that these materials should be made freely available so that models get developed to incorporate these data, and that members of their culture can share in some of the benefits of this technology. Which side of this debate do you think is correct, and why?
5	13 Feb	Gender discrimination and violence against women	It is widely claimed that having more diversity on product development teams both yields better products and avoids serious pitfalls. Do you think that if Apple had had a more diverse team developing the AirTags product, they could have avoided the situation where their product was being used by stalkers? What do you think of company programs to try to increase diversity among their product teams?
6	27 Feb	Big data, privacy and the right to be forgotten	Suppose you’ve been building an archive of historical sexually explicit images of men created in 1970s in the US. Someone contacts you and says that an image in your archive is of their late father, and asks you to remove it. Which principles do you use in deciding how to respond to the request, and what is your eventual decision?
7	6 Mar	Discrimination on the basis of sexual orientation and gender identity	You are responsible for identifying the recipient of large entrance scholarships for a prestigious university program. One of your colleagues has designed a machine-learning algorithm that uses applicants' social media profiles (e.g., posts, likes, friends, groups, etc.) to predict educational success and worthiness, and recommends that you incorporate this system into a pilot study for the upcoming application cycle. Would that be ethical? Why or why not?
8	13 Mar	Discrimination against disabled people	Suppose that you have a family member with a mental health issue. Would you encourage them to use a mental health chatbot as part of their treatment? Why or why not?
9	20 Mar	Global issues including government surveillance	How can technology companies responsibly gather data in developing countries or Indigenous communities? Are there particular principles they should keep in mind?
10	27 Mar	Fixes: inclusivity in education and tech	Develop and present your own curricular module for an existing 300- or 400-level technical course to address the issues raised in this course and the readings discussing tech curricula. What material would you cover? What readings (not already on our course’s syllabus) and/or assignments would you give? [Note: we are not asking you to redesign CS 492. This prompt is instead consistent with the Harvard project, where individual courses get individual modules.]
11	3 Apr	Fixes: law, regulation, and other approaches	Describe an AI application that might be used by a government agency to assist with automated decision making. Perform an algorithmic impact assessment (AIA) on the application using the Government of Canada tool from the readings. (Obviously, since this is a simulation exercise, you’ll need to make some reasonable assumptions with respect to responses to the various questions and explain those.) What are the findings from the AIA about the intended AI application? What did you learn from this exercise? Is the AIA a useful tool?