Reading List

Algorithmic Bias

We are compiling this list for anyone who wants to learn more about the case for why and how we can make AI systems more inclusive. For a more comprehensive list of the work we came across, learned from, or bookmarked, go here.

First of all, what is algorithmic bias?

• Machine Learning for Everyone

• Mozilla's 2019 Internet Health Report

• Machine Bias: An investigative piece by ProPublica on the racism and the use of algorithms for criminal sentencing decisions

• An Examination of Hiring Algorithms, Equity, and Bias: Report by Upturn on bias in automated hiring.

• AI Now Report 2018

• 'The Algorithms aren't Biased, We Are': An important biased that at the end of the day, this is a human problem - machines don't work on their own.

Why should AI be inclusive?

• Against Black Inclusion in Facial Recognition: provides another perspective on advocating for greater representation in tech - what if it's a technology we don't want to be used at all?

• On Diversity in Tech

• Norms for Beneficial A.I.: A Computational Analysis of the Societal Value Alignment Problem

• TED talk - Machine intelligence makes human morals more important: why we "cannot outsource our responsibilities to machines," and must "hold on ever tighter to human values and human ethics."

So what can we do about it?

• Ethics Alone Can't Change Big Tech

• From Principles to Action: How Do We Implement Tech Ethics?

• Ethical OS Toolkit - refer to page 43 for Risk Zone 4: Machine Ethics & Algorithmic Bias

• Big Brother's Blind Spot: highlights the importance of being aware of the nuance associated with bias and inclusion.
  

• AI Now's Algorithmic Impact Assessment

• We Need a Change in Computer Science Curricula