CSCI 14
Ethics of Technology Winter 2020

Cross-listed PHIL 14 / STS 14
This is not the current course catalog

Class Details

A prominent company recently realized the machine-learning algorithm trained on its past hiring data had learned a bias against female candidates and so was unsuitable for resume evaluation. But given competing definitions of fairness, how should we decide what it means for an algorithm to be unbiased? Machine vision algorithms are systematically less likely to recognize faces of people of color. Since many face recognition algorithms are used for surveillance, would improving these algorithms promote justice? Deep fakes may pose serious challenges to democratic discourse, as faked videos of political leaders making incendiary statements cast doubt on the provenance of real videos. Do the researchers developing these algorithms, often academics funded by National Science Foundation grants, have an obligation to desist? In a field filled with such vexing questions, the ethical issue most commonly addressed by the media is whether a self-driving car should swerve to hit one person in order to avoid hitting two. In this class, we will go beyond the headlines to explore the ethics of technology. We will discuss issues such as transparency, bias and fairness, surveillance, automation and work, the politics of artifacts, the epistemology of deep fakes, and more. Our discussion will rely on articles from the course packet, enlivened by discussions with experts in the field over Skype. Students will apply their ethical knowledge to write multiple newspaper length op-eds arguing for their views. If students choose to submit these op-eds for publication, the instructor will coach them on appropriate procedures and venues. Adjunct Instructor Bio: Kathleen Creel ’10 is an advanced doctoral student in the Department of History & Philosophy of Science at the University of Pittsburgh. Her research focuses on epistemic and ethical issues in computer science and its scientific applications, such as transparency in machine learning and the ability of algorithmic decisions to provide reasons.
The Class: Format: lecture
Limit: 15
Grading: pass/fail only
Requirements/Evaluation: 3 op-eds for a total of 10 pages
Prerequisites: none
Enrollment Preferences: based on a written paragraph expressing interest
Materials/Lab Fee: $20
Notes: This course is cross-listed and the prefixes carry the following divisional credit:
PHIL 14 STS 14 CSCI 14

Class Grid

Course Catalog Archive Search

TERM/YEAR
TEACHING MODE
SUBJECT
DIVISION



DISTRIBUTION



ENROLLMENT LIMIT
COURSE TYPE
Start Time
End Time
Day(s)