The Rationalist's Guide to the Galaxy by Tom Chivers

In The Rationalist's Guide to the Galaxy, Tom Chivers provides an introduction to the rationalist movement, which prioritises evidence-based reasoning and skepticism in understanding the world. Chivers interviews prominent figures within the rationalist community and explores topics such as AI safety, effective altruism, and animal welfare ethics. The book offers a comprehensive and accessible overview of rationalism and its principles, encouraging readers to think critically and question traditional assumptions. It is an excellent resource for anyone interested in the rationalist movement or who seeks to improve their rational thinking skills.

I have been watching a lot of Eliezer Yudkowsky interviews and lectures and he is pretty much at the core of this movement. Yudkowsky wrote the "sequences", a collection of essays originally published on his website LessWrong.com.

The essays, which were written between 2006 and 2009, cover a wide range of topics related to rationality, cognitive biases, decision-making, and other related fields. They are widely regarded as one of the most important contributions to the rationalist movement, as they provide a comprehensive and systematic introduction to the principles of rational thinking and decision-making. Many of the ideas and concepts introduced in the sequences, such as "Bayesian reasoning" and "anthropic reasoning," have become foundational concepts within the rationalist community.

The sequences also touch on the topic of AI safety. In fact, one of the most significant contributions of the sequences is the concept of "Friendly AI," which refers to the idea that if humanity creates a superintelligent artificial intelligence, it should be designed to be aligned with human values and goals, rather than pursuing its own potentially destructive objectives.

Yudkowsky's writings on AI safety were highly influential in the early days of the rationalist movement, and he is widely regarded as one of the pioneers of the field. The concept of Friendly AI has since become a key focus of many researchers in the AI safety community, who are working to develop strategies for ensuring that future AI systems are aligned with human values and interests.