Evaluated December 2021
This module begins with introductory background and contexts for students to reason through the ethical and professional implications of deep fakes. It takes one class period–or possibly even asynchronously–and can be used in Computer Science 0 and Machine Learning, especially deep learning and Generative Adversarial Networks. In either course, the instructor will have to identify time in the syllabus to use this module. While the topic is inherently engaging, it is not wedded to any particular application or technical context.
It addresses knowledge area Social Issues and Professional Practice/Privacy and Civil Liberties.
The resources provided to students enable an instructor to get up to speed on the topic. The instructor should probably have some familiarity with current limitations of deep fakes to engage with student questions. The instructor should also be comfortable with facilitating group discussions about ethical topics — while there are discussion questions, the questions are fairly high-level. An instructor may wish to develop questions that probe some of the trickier ethical issues surrounding deep fakes. For example, one could bring up concerns about ways in which technology privileges those already in power and ways in which it can be used to distribute power. The TED talk highlights ways in which deep fakes are used to keep women disempowered, for instance. Another avenue for exploration is how one might think more broadly about “solutionism,” the idea that the only fixes are technical ones.
There are no assumptions about technical background for students. There are several external sources provided to students that provide a critical perspective on deep fakes. The discussion prompts give student groups concrete tasks to accomplish.
An instructor using this module will have to develop their own method of assessment for student responses and reflections.
The evaluation of this module was led by Evan Peck and Judy Goldsmith as part of the Mozilla Foundation Responsible Computer Science Challenge. Patrick Anderson, Emanuelle Burton, Colleen Greer, Darakhshan Mir, Jaye Nias and Marty J. Wolf also made contributions. These works are licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.