Assignment 02: Ethics and AI

Learning Goals

  • Practice applying ethical frameworks
  • Practice discussing ethical issues with partners

Grading Walk-Throughs

These instructions are subject to change as I have not yet met with the TAs.

This assignment will be graded as “Nailed It” / “Not Yet” by a TA. To complete (“Nailed It”) the assignment, you must

  1. Complete the assignment and submit your work to gradescope.
  2. Meet with a TA during their mentor session hours.
  3. Complete the walk-through with all group members. I prefer all partners to be present during the walk-through, but you can each meet with the TA separately if needed.
  4. Walk the TA through your answers. Do not expect to make corrections during the walk-through.
  5. The TA will then either
    • mark your assignment as 100% on gradescope, or
    • inform you that you have some corrections to make.
  6. If corrections are needed, then you will need to complete them and conduct a new walk-through with a TA.

If you have concerns about the grading walk-through, you can meet with me after you have first met with a TA.

Overview

For this assignment you will:

  1. Form groups of four
  2. Choose a fictional case study
  3. Answer a few questions prior to reading the study
  4. Read the study
  5. Answer the prompts found in the study

Tasks

I’d recommend having one partner start by opening up gradescope now and acting as a scribe (the group’s note taker).

Case Studies

Here is a quick summary of the six available case studies provided by the Princeton Dialogues on AI and Ethics. You should choose one prior to reading the PDF; your group is welcome to start over and choose a new one if you decide you don’t like the one you’ve chosen after you’ve started reading it.

Case Study 1: Automated Healthcare App

A team of medical researchers and computer scientists develop an app that utilizes artificial intelligence technologies to make diabetic care easier, more holistic and more accessible.

Case Study 2: Dynamic Sound Identification

An R&D company develop an app that can identify–among other things–personal information about those speaking, links to websites selling a product being advertised on television, encyclopedic entries about bird calls in the wild and other relevant resources.

Case Study 3: Optimizing Schools

A public high school contracts a data science company and gives them access to student behavioral data (attendance, purchases, library usage, movement on campus, etc.) so that they can predict students at-risk of dropping out.

Case Study 4: Law Enforcement Chatbots

A country’s federal law enforcement agency teams up with University researchers to develop a chatbot that could be used to identify cybercriminals.

Case Study 5: Hiring By Machine

A group of military veterans create a non-profit company that creates products for veterans transitioning back to civilian life. The company strongly believes in open-source and hiring veterans.

Case Study 6: Public Sector Data Analytics

A once-prosperous city contracts with a consulting group to use an algorithmic, data-driven approach to reduce crime.

Initial Prompt

Prior to reading the case study in full, your group should choose at least three questions to discuss. You do not need to have a complete understanding of the case study or the questions prior to thinking these through. Try to put yourselves in the developers’ shoes and guess at any issues that might arise or any background learning that your team might need to conduct.

Essentially, apply three of these questions to your limited understanding of your chosen case study.

  • Questions on oversight and accountability
    1. Which laws and regulations might be applicable to this project?
    2. How is ethical accountability being achieved?
  • Questions on data privacy and anonymity
    1. How might the legal rights of organizations and individuals be impinged by our use of the data?
    2. How might individuals’ privacy and anonymity be impinged via aggregation and linking of the data?
  • Questions on data availability and validity
    1. How do you know that the data is ethically available for its intended use?
    2. How do you know that the data is valid for its intended use?
  • Questions on model bias
    1. How have we identified and minimized any bias in the data or in the model?
    2. How was any potential modeler bias identified and then, if appropriate, mitigated?
  • Questions on model transparency and interpretation
    1. How transparent does the model need to be and how is that transparency achieved?
    2. What are likely misinterpretations of the results and what can be done to prevent those misinterpretations?

These questions are from Integrating Ethics within Machine Learning Courses | ACM Transactions on Computing Education by J. Saltz et al.

You should now add some notes to gradescope about your initial discussion.

Discussion Questions

Case studies include anywhere from three to seven discussion questions. Please read the case study individually, but pause at each discussion question and discuss them as a group. You’ll find a place on gradescope to jot down some notes about your discussion, which means you’ll want to assign one person to have it open and act as a scribe.

Now read through your chosen case study PDF.

Each case study PDF includes a “Reflections & Discussion Questions” section near the end. You should read through these individually, and you can optionally choose to discuss them as a group. However, you did not need to report on your discussions for this part of the case study PDFs.

Submitting Your Assignment

You will submit your responses on gradescope. Only one partner should submit. The submitter will add the other partner through the gradescope interface.

Additional details for using gradescope can be found here: