AI, Ethics, and Geoethics (CS 5970)
Module 4: Responsible Conduct of AI Research
Summary
- This module will take one week to complete
- We will be reading and actively discussing, watching a few videos, and completing activities related to the readings and videos
Overview
For this module, we will first read about responsible conduct of research in general and then then we will dive into the growing field of responsible AI research.
Being responsible and ethical students and scientists is important to our future careers and to society as a whole. If we are not ethical and responsible researchers (in both academia and private industry), the public loses trust in science and the science itself is potentially incorrect, both of which can lead to many bad consequences. To help ensure that we minimize ethical breaches, many federal funding agencies now mandate that their grant recipients and all students paid on those grants receive training in “responsible conduct of research.”
This module will not replicate the full course on RCR but instead introduce you to the topics and tie them together with how we can both responsibly develop AI as well as develop AI that is responsible with its actions.
Resources for this Module
Our main reading for this module will come from the book On Being a Scientist. For OU students, the book is on canvas so you don’t have to purchase it. The book is 55 pages and is a relatively fast read with a lot of vignettes. To help us absorb the materials well, we will break the reading up across the week (see assignments below).
Throughout the class, we will make use of a great online resource about ethics called Ethics Unwrapped. This online course has a lot of engaging videos about different ethical principles. For this module, we will focus on a few of the scandals and the ethical principal videos they list as the main focus for that scandal. While none of their scandals are AI related (we can be grateful for that!), we will start with technology related scandals!
We will also make use of journal papers and other published works about responsible use of AI.
Assignments
The assignments are broken into several days (to spread out the readings) though you are welcome to do them in different sized chunks if you prefer.
Day 1
Responsible Conduct of Research
- (30 min) Read pages 1-18 (this is Introduction all the way through Research Misconduct) from On Being a Scientist. I know that not all of you are graduate students so the discussion of choosing a research group may seem irrelevant but I will note that the advice applies to choosing where to work. The questions they list are valuable no matter your chosen career!
- (15 min) Pick one of the vignettes from page 5, page 10, page 14, page 17, or page 18 and discuss on #general in slack. State which vignette you are replying to do that you can do them in a thread
Case Studies and Ethical Principles
- (15 min) Read the Ethics Unwrapped page about the Therac-25 scandal (note, while this wasn’t AI, it was an automated programming issue that killed people!) and watch the video at the bottom about the ethics of harm
- (15 min) Pick one of the discussion questions from their page and discuss in #general
Grading Declarations
OU students, don’t forget to turn in your grading declarations on canvas! Today’s declaration is called “Module 4: Responsible conduct of research and Therac-25.”
Day 2
Responsible Conduct of Research
- (30 min) Read pages 19-38 (this is Responding to Suspected Violations of Professional Standards through Authorship and the Allocation of Credit) from On Being a Scientist.
- (15 min) Pick one of the vignettes from page 22, page 25, page 26, page 32, or page 36 and discuss on #general in slack. State which vignette you are replying to do that you can do them in a thread
Case Studies and Ethical Principles
- (15 min) Read the Ethics Unwrapped page about the Appropriating Hope. Again, while this isn’t AI, it is related to the sharing of ideas discussed in the book. Watch the video at the bottom about appropriation.
- (15 min) Pick one of the discussion questions from their page and discuss in #general
Grading Declarations
OU students, don’t forget to turn in your grading declarations on canvas! Today’s declaration is called “Module 4: RCR and Hope”
Day 3
Responsible Conduct of Research
- (15 min) Read the rest of the On Being a Scientistbook. This should be pages 39-49, which is Intellectual Property through The Researcher in Society.
- This time we are going to discuss TWO of the vignettes.
- (15 min) First, I want you to pick one of the vignettes from pages 42, 45, or 47 and discuss the questions in #general. As before, state which vignette you chose.
- (15 min) Everyone should discuss the final Agent Orange case. Specifically the ending sentence is critical to what we are studying. “I used to think that one could avoid involvement inn the anti-social consequences of science simply by not working on any project that might be turned to evil or destructive ends. I have learned that things are not that simple. The only resource is for a scientist to remain involved with it to the end.” (Galston, Arthur W. Science and Social Responsibility: A Case History. Annals of the New York Academy of Science (1972): 196-223 and also from page 49 of On Being a Scientist). Discuss how this bears on our work as AI scientists (and as researchers in whatever field you are in primarily).
Responsible AI research
We are going to focus the rest of our reading on papers and discussions about responsible AI research.
- (15 min) The first reading is a short article from Nature Medicine called Do no harm: a roadmap for responsible machine learning for health care. Your library should have access for free downloads but OU students can access a backup PDF on canvas.
- (20 min) The second reading is from Google and is titled “Responsible Development of AI“. This summarizes how Google is working to develop AI in a responsible manner.
- (5 min) Watch this video about banning lethal autonomous weapons
- (10 min) If you want to learn more about banning lethal autonomous weapons:
- Ban Lethal Autonomous Weapons
- Future of Life institute (and they have a specific focus on AI) as well as an older call on the autonomous weapons ban
Grading Declarations
OU students, don’t forget to turn in your grading declarations on canvas! Today’s declaration is called “Module 4: RCR and AI”
Personal goals for Responsible AI
(30 min) One of the artifacts that we will produce as a class is a set of personal goals for how you want to be a responsible AI producer and consumer as well as how you can contribute to the responsible use and development of AI. To get us going, this week I want you to create a draft of this product and turn it in on canvas. Don’t consider this an essay – a list of well thought out bullets will do. I will create a larger combined document and share it with class next week. The URLs for the shared document will be on canvas and slack (but not posted here) to keep us from being spammed. Your document should answer the following:
-
- How I will be a responsible producer of AI
- How I will be a responsible consumer of AI
- How I will contribute to the responsible use and development of AI
- How I will remain an involved scientist in my work