AI, Ethics, and Geoethics (CS 5970)


Module 2: Why Do We Need AI, Ethics, and Geoethics?

Summary

  • This module should take 2 weeks to complete
  • We will be watching several videos and completing introductory reading as well as engaging in active discussion

Why do we need ethics for AI?

Since this class is an elective and you chose to enroll, you probably already have enough interest in using AI ethically to have some answers to the question of why we need ethics for AI.  If you have been reading the news lately, you will see multiple examples of how either AI or another automated system made a mistake that affected someone or a group of people in a negative way.  Unfortunately the news does not tend to show us any of the potential wins of using AI or automated technology! 

In addition to studying the need for ethical development and use of technology for AI in general, we will have a focus on the geoscience specific related applications.  This doesn’t mean other applications are not of interest!  It just means we will definitely focus some on geoscience applications such as weather and climate.

For this initial module, we are going to focus on an overviews of the wide variety of impacts and ways that AI and automated technology can negatively impact humanity.  Much of these impacts are unintentional but that does not make them any better or more tolerable!  While you read and listen to the videos, think about ways in which these negative impacts could have been avoided in the first place.  Would hiring a more diverse team have been enough?  While it may be a good start, I doubt it is the full solution.  We really need to think about the potential uses of our technology as we develop it and ensure that it is used wisely.  

Readings

The readings for this module (and the future modules) are available in the OU Bizzell library for OU students only.  They are also available online in kindle format, though unfortunately I have been unable to find them as an e-book (which would let the library share the readings online more freely).  The two books we will be using are listed below.  Do not start reading just yet!  Wait for the assignments below.

Videos

There are two videos for this module.  OU students will have access to the movie through the OU library.  For students outside of OU, there are online screenings available for the movie.  Just search around and you can find them.  Consumer Reports was offering free screenings for anyone for awhile.  The second video comes from the recent American Meteorological Society 20th Conference on Artificial Intelligence for  Environmental Science.

  • The full-length movie is Coded Bias (wait to watch it until you read the assignments below!)
  • The talk is from Dr Nicol Turner-Lee and was presented as a keynote at the AMS AI conference in January 2021. Because the talk is currently behind a paywall, OU students will receive special instructions on watching the talk through canvas.  Participants outside of OU can email Dr McGovern for access.

Assignments

Week 1/Days 1-3

  • If this class was in person, I would screen the movie in class over several class periods.  My idea is that we would watch and interrupt for discussion frequently, thus getting through only part of the movie each class. Since the class is online, we are going to simulate this through slack!  
  • (4-5 hours) Assignment for watching and discussing Coded Bias (it involves reading reading as well)
  • OU students: There are two grading declarations for this assignment: 
    • “Module 2: Coded Bias movie”
    • “Module 2: Coded bias discussion”

Week 2/Day 4

Week 2/Day 5

Week 2/Day 6

Ongoing (throughout the semester but starting these two weeks)