Across industries and disciplines, across corporations, government agencies, and nonprofits, and across various roles within various organizations, many of us are being encouraged to use generative AI tools—or find ourselves using them, since they are increasingly baked into various products and services.
The effectiveuseof generative AI requires acertain level of AI literacy.AI ethics literacy is a part of AI literacy. It helps developers build better products, helps users of such products make better use of them, and helps all of us understand how various tools impact us, and how societies are being reshaped by such uses.
Below is a list of AI ethics issues (not exhaustive, and likely to be updated as generative AI tools are deployed in novel ways in new contexts). For each of them, we have pulled together relevant materials (articles and video recordings) from the center’s website. We will add materials as Markkula Center staff and Santa Clara University faculty scholars continue to write and present events related to AI literacy.
We hope that engagement with these resources will generate a deeper and broader conversation.
Preface
1. Environmental Impact
2. Transparency
3. Labor
4. Socio-Economic Inequality
5. Ethics of Deployment
6. Social Life (with Chatbots)
7. Moral Deskilling
8. Anthropomorphism
9. Hallucinations / Accuracy / Truth
10. Bias (In and Out)
11. Prediction
12. Authenticity
13. AI Slop
14. Privacy
15. Cybersecurity
16. Consciousness / Sentience
17. AGI and Superintelligence
18. Effects on the Human Spirit
Preface
Using generative AI well requires a level of AI literacy. Without understanding both the technology’s affordances and its limitations, we cannot make good use of it. Without understanding (at least at a general level) how large language models generate what they do, we can’t decide when they are fit for the purpose. And without understanding the benefits and harms they might bring, and to whom, we can’t conduct the ethical analysis that should precede any technology’s adoption.
AI ethics literacy is the part of AI literacy that highlights the ethical issues related to the development and deployment of generative AI tools. AI ethics comprises more than privacy, intellectual property rights, bias, and job displacement (important issues with which many people are now more or less familiar). The many beneficial uses of AI are also, of course, a key part of AI ethics literacy, but those are often touted loudly by all the entities developing and deploying AI tools; to the extent that this compendium rings more of a cautionary note, it’s in response to the volume of the AI-related hype.
Amid the growing calls for AI literacy, we hope that this compendium will serve as a resource for developers, users, students, teachers, parents… in truth, for all of us who live in societies that are rapidly adopting generative AI.
Environmental Impact
Overview
Video: AI’s Environmental Footprint—Development and Usage
Water
Green AI
Energy
Case Study: The Environmental Impact of Generating Images With AI
Business Implications
Video: AI and the Environment: Business Implications
From “Surprise and Delight” to “Startle and Dismay”
Regulation
Video: AI, Regulation, and Environmental Law
AI for Sustainability
Public Health
[Links to be added as articles, cases, videos, etc. get published]
Transparency
Labor
Labeling, Augmenting, Evaluating Data
[Links to be added as articles, cases, videos, etc. get published]
Replacement/Augmentation of Human Work
Cheating
Generative AI and a More Human Education, Part 2: Text
A Moral Panic: ChatGPT and the Gamification of Education
Case Study: AI-Writing Detectors
Socio-Economic Inequality
[Links to be added as articles, cases, videos, etc. get published]
Ethics of Deployment
ChatGPT and the Ethics of Deployment and Disclosure
This is an Experimental Blog Post
Case Study: Emotion Recognition and Smart Cities
Ethics Case: Open Source AI: To Release or not to Release GPT-2
From “Surprise and Delight” to “Startle and Dismay”
Social Life (with Chatbots)
AI, Death, and Mourning – Markkula Center for Applied Ethics
Can Chatbot Friendships Make us Happy?
Case Study: Chatbots and Agency
Moral Deskilling
Video: Shannon Vallor: De-Coding our Humanity: Dignity and Fullness in the Digital Age
Anthropomorphism
Can Chatbot Friendships Make Us Happy?
Case Study: AI, Death, and Mourning
Hallucinations / Accuracy / Truth
Certainly, Here is a Blog Post
How Must Journalists and Journalism View Generative AI?
Case Study: AI-Writing Detectors
Video: Shannon Vallor: The AI Mirror
How we can Protect Against Deceptive AI Political Ads
Bias (In And Out)
On Biased Humans and Algorithms
AI and the Pink Elephant in the Room
Prediction
Pride, Prejudice, and Predictions About People
Authenticity
Authenticity in the AI Content Era Will not Come Cheap
On Recommendation Algorithms and What Makes Us Boring
Case Study: AI, Death, and Mourning
Perplexity AI vs Journalism: The Risks we Need to Anticipate
AI Slop
[Links to be added as articles, cases, videos, etc. get published]
Privacy
Data Harvesting
Case Study: Emotion Recognition and Smart Cities
Data Leakage
Copyright
Generative AI and a More Human Education, Part 1: Art
Video: Generative AI Ethical, Legal, and Technical Questions
Cybersecurity
Agentic AI’s Cybersecurity Challenges
[Links to be added as articles, cases, videos, etc. get published]
Prompt Injection
[Links to be added as articles, cases, videos, etc. get published]
Data Leakage
Consciousness / Sentience
Case Study: Chatbots and Agency
AGI and Superintelligence
[Links to be added as articles, cases, videos, etc. get published]
Effects on the Human Spirit
Cura Personalis and Generative AI in Education
Generative AI and a More Human Education, Part 2: Text
Video: Shannon Vallor: De-Coding our Humanity: Dignity and Fullness in the Digital Age
Image: “Datafication”/ Kathryn Conrad / Better Images of AI / CC-BY 4.0