5 Must-read books on Ethics in AI

Ethics in AI is a hot topic of discussion everywhere these days. We see it discussed in conferences, news stories as well as on social media. The answer to the question about why Ethics in AI is essential lies with the fact that AI systems are replacing human processes. Automation and AI algorithms introduce a level of opaqueness and a lack of transparency around the decision-making process. In some cases, such as self-driving cars, it can affect human lives. Artificial Intelligence applications can also perpetuate and amplify cognitive biases and introduce inequality in outcomes. We have witnessed examples of AI recruiting systems gone wrong in addition to questionable facial-recognition technology.

Many ethical AI frameworks have been published recently. The IEEE framework focuses on the ethical design of autonomous and intelligent systems. Europe’s Ethics guidelines emphasize the need for human-centric AI systems. Big five consulting companies — Deloitte, Accenture, etc. have provided their versions of Ethical AI design principles. Most of them focus on the risks of the outcomes of the predictive systems. When evaluating the results, it is important to consider strategies for the mitigation of those risks. If there are concerns that the unresolved, there needs to be scope to prevent the systems’ deployment.

It is no surprise that many books have been published on this topic. Most of these books have focused on highlighting the negative consequences of machine learning in general and biased data more precisely.

I have a long list of books yet to read but based on what I have read thus far, here are my top five recommendations:

1. Weapons of Math Destruction by Cathy o’ Neil: In this book, renowned mathematician Cathy O’Neil provides multiple examples of machine learning or statistical modeling gone wrong. Examples range from practical teacher evaluation algorithms to automatic systems that judge job applicants. While Cathy focuses on the negative impacts of big data analysis, it would be an eye-opener for anyone looking at real-world examples of biased systems caused by wrongful statistical modeling techniques.
2. 21 lessons for the 21st century by Yuval Noah Harari: Although this is not your typical AI-focused book, Harari presents a broader view on the effect of AI on society, future jobs, fake news, and topics like the Universal Basic Income (UBI). He presents a provocative book around our current culture as influenced by technology and how it changes the narrative globally, the impact of automation on developing countries, and how

“those who own the data own the future.”

His views on disruptive technology (biotech, infotech) and its consequences (climate change, mass unemployment) provide fodder for policymakers and governments to level the global playing field.

3. Hello World: Being Human in the Age of Algorithms by Hannah Fry: In this book, Hannah deftly handles explaining the math and statistics behind AI algorithms and explains how they are transforming fields such as health, justice, transport, and the arts. While Hannah presents some horrendous stories of algorithms gone awry, she also states that

“People are less tolerant of an algorithm’s mistakes than of their own — even if their own mistakes are bigger.”

4. Automating Inequality by Virginia Eubanks: Virginia presents with research examples of the social implications of technology and its impact on marginalized populations. Based on her experience working on technology and poverty for over two decades, Virginia presents dire cases of how the ethical and technical complexity of automated decision-making affects the poor and working-class with life-threatening consequences. While the book is riveting, it is intended to awaken institutions and policymakers to have conversations around these societal inequalities caused by automation.

5. Artificial Unintelligence, How Computers misunderstand the world by Meredith Broussard: Meredith states,

“My goal in this book is to empower people around technology.”

She does an excellent job of it while emphasizing the point of algorithmic bias, accountability, and representation in a male-dominated tech field. The book presents an in-depth look at the social, legal, and cultural impact of AI for the public with an urgent call to action to design and implement technologies that benefit all of society.

Originally published at https://www.linkedin.com.