AI ≈ ML

“all models are biased, thats how they work”

Joanna Bryson - https://www.machine-ethics.net/podcast/9-joanna-j-bryson/

cognative bias / social bias

Algorithmic Bias ≈ negative outcomes for a group using a technology system

negative bias

discrimination,
inequality, denigrate,  injustice, dogma, stereotyping, inaccess,
harms, differentially impacted,
unfair

“Fairness: This refers to the equitable treatment of individuals, or groups of individuals, by an AI system.

When properly calibrated, AI can assist humans in making fairer choices, countering human biases, and promoting inclusivity.”
   

https://www.ibm.com/cloud/learn/ai-ethics

“Algorithms can enhance already existing biases”

https://ethics-of-ai.mooc.fi/chapter-1/1-a-guide-to-ai-ethics

Provocations

What does “good” bias discovery / mitigation look like?

Fairness responsibilities
in job roles during production?

Data protection and
bias mitigation

What is the role of
positive discrimination and equity in data science?

Aligning the end user's perception of fairness

Bonus questions

What role should the legisaltion have in Bias / Fairness?

Minimum Fairness metrics in general and specific contexts?