Recommended Reading

Artificial Intelligence by Melanie Mitchell

Why is this book recommended?

If you want a more technical overview of AI, but not so technical that you can’t follow along, try this book. It covers the basics of AI while also providing a historical perspective on AI, but it doesn’t cover Gen AI. Melanie makes it intuitive and super entertaining and weaves in interesting stories throughout.

Melanie is a highly accomplished computer science researcher and professor; you may need to distill her perspective to apply it to your business context.

Empire of AI by Karen Hao

Why is this book recommended?

Maybe you don’t like gossip, but maybe you do. As controversial as this book is, it is recommended to help you understand that there are power dynamics in AI. This can significantly affect the decisions businesses make in AI models they invest in and companies they will or will not work with. Reading other perspectives will help you form your own, and you will make better business decisions as a leader.

Caution: the book’s reporting may seem overly one-sided, opinionated, and negative.

AI at the Edge by Daniel Situnayake & Jenny Plunkett

Why is this book recommended?

We keep talking about LLMs, but rarely talk about AI at the edge, where we execute AI tasks directly on local devices, such as smartphones, tablets, or a local server, instead of sending data to a remote cloud for processing.

There are a considerable number of real-world applications for AI at the edge. This book offers interesting real-world examples and best practices for implementing AI, both in general and on devices. If you’re a practitioner, read all the way through. For leaders: even if you read the first few chapters of the book, it will spark ideas.

Zucked: Waking Up to the Facebook Catastrophe

Why is this book recommended?

Left or right, the extent to which certain companies prioritize profits at the expense of data privacy and societal well-being continues to intensify.

This book examines how Facebook’s growth-at-all-costs culture and product decisions led to widespread privacy issues and societal harm, resulting in political influence, misinformation, and loss of public trust.

While it barely scratches the surface of the broader dangers posed by modern AI, it provides a valuable foundation for understanding how even “old-fashioned” technology, when scaled without accountability, can cause harm, knowledge that is essential when thinking about the risks of AI. The author, Roger McNamee, was an early investor in Facebook and later became a vocal critic.

I would pay particular attention to how the privacy abuses and misinformation occurred, and how your organization can avoid similar situations when using AI tools and applications.

Superintelligence: Paths, Dangers, Strategies

Why is this book recommended?

This book explores the possible paths to the development of superintelligent AI and the challenges that could arise if machines eventually surpass human intelligence. It examines concepts such as the AI control problem, intelligence explosion, and the difficulty of aligning highly capable systems with human values before they become too powerful to direct.

While the scenarios discussed are highly speculative, and today’s large language models are neither human-level intelligent nor superintelligent, I would pay particular attention to the discussions around AI alignment, unintended consequences, and the challenges of controlling increasingly capable systems. These concepts provide valuable context for thinking about how AI should be designed, governed, and deployed responsibly within any organization.