Meeting # 3: 11.30.23

🎯 Objectives

Know the potential pitfalls of AI.
Understand the impact that pitfalls of AI might have in education and our communities.
Iteratively prototype teacher guidelines on the use of AI in the classroom.

3️⃣ Pitfalls of AI (20 min)

❓ Why:

Understand the impact that pitfalls of AI might have in education and our communities.

ℹ️ How:

Explore real-world examples that showcase the pitfalls of AI

Gender Shades
Human vs. Gen AI Bias
Career Bias
Mortgage Lending
Predictive Policing Software
Wisconsin: "High Risk" Labeling Failure

Gender Shades

Human vs. Gen AI Bias

🔗 Bloomberg Article

Bias: Gen AI vs Humans by Andrew Fenstermaker

Career Bias

ℹ️ Canva Magic Media: Image generation functionality built on Stable Diffusion which gets its raw data from LAION-5B, the world’s largest openly accessible image-text dataset, with more than 5 billion images and captions found on the internet.

ℹ️ Bing Image Creator: powered by DALL·E 3

ℹ️ Adobe Firefly: trained on content from Adobe Stock, openly licensed content, and public domain content where the copyright has expired

AI Examples of Bias (Presentation) by Andrew Fenstermaker

Mortgage Lending

🔗 The Secret Bias Hidden in Mortgage Approval Algorithms

Mortgage Lending by Andrew Fenstermaker

Predictive Policing Software

🔗 The Verge: Predictive Policing

Predictive Policing by Andrew Fenstermaker

Wisconsin: "High Risk" Labeling Failure

🔗 The Markup: False Alarm: How Wisconsin Uses Race and Income to Label Students “High Risk”

High Risk Labeling by Andrew Fenstermaker

AI in Education