Worth a read for someone thinking about broad implications of the technology

This is a book with some serious star power. With Henry Kissinger, former Secretary of State and National Security Advisor, Eric Schmidt, former CEO and Chairman of Google, and Daniel Huttenlocher, Dean of the College of Computing at MIT it is impossible to not be interested in the book. The book is an amalgamation of the thinking of these three people over the course of a couple of years and is far more focused on the structural, long-term impacts of narrow artificial intelligence on such topics as medicine, communications, and security / warfare.

The book opens with some very interesting examples of applied AI and how they created insights that were not attainable through human enterprise in the foreseeable future. One is an application of AI which resulted in a new antibiotic called halicin, in honor of HAL, which was not on an obvious path to discovery before the AI identified as it a potential candidate. The second revolves around cooling at Google’s data centers, a task some very talented engineers are focused on day in and day out. After bringing an AI into the fold they increased cooling efficiency by 40% the levels achieved by the human engineers alone.

The book goes on to investigate how global networks: economic, social, technological, etc. have fostered the development of AI as well as their potential to multiply its negative effects. Add in the high speed of decision making for an AI, the fact that its “logic” may not be interpretable, and that it can often outperform humans on their own and you get an interesting set of chapters. This part of the book is really focused on how individuals and societies should manage these downsides to hopefully get the upside of AI with less of the downside.

One of the last topics that I found really interesting was a section on what it will be mean to be human in a world of widespread AI. Today, many of us derive our self-worth from what we contribute in a career, as a parent, a volunteer, etc. and some, or all of that opportunity to contribute may be better done by an AI. Our ability to think critically and develop frameworks also define our humanity but it will be tempting to cede some of that to AIs both to reduce cognitive load and because it will be tempting to accept the AI’s decisions as better. As AI is used to assist or completely takeover many of our current tasks we will need to determine what make us happy and fulfilled, and what truly makes us human to remain engaged members of a collective society.

This book is definitely worth a read for anyone interested in AI and its potential impacts on humanity without conjecture or hype. It is definitely a how-to manual or a book that will help you with development of a specific AI, but it will push your thinking about how to think about systems broadly and how AIs can impact humanity, both positively and negatively, when introduced into those systems.