These days, we absorb information at a faster rate than ever before. We are constantly being subjected to images, news, facts, details and figures that it’s almost miraculous we can retain half of it. I don’t know about you, but no matter how advanced technology becomes, the feeling of picking up a brand new book is second to none. It gives your eyes a rest from blue light, your fingers get a break from masterfully balancing a tiny device in your hand and you suddenly feel more relaxed and less distracted than when you are reading from a screen. I recently asked my colleagues if they could recommend the best AI (non-fiction) books they have read and the answers I received came from various departments of our company. It doesn’t matter what you do for a living, anyone can understand technology if they have the right material to guide their understanding. If you are going to tackle a complex subject such as artificial intelligence, you are going to need to focus, so get comfortable, put your feet up and browse our recommendations for your next literary purchase: Superintelligence: Paths, Dangers, Strategies (Reprint Edition) Author: Nick Bostrom Technical Level: Low Publication Date: April 2016 Topic: Philosophy Have you started to hear about the AI singularity, the importance of ethical AI, the inevitability of superhuman intelligence arising in machines and wondered where all of these ideas came from? Many people currently working on AI ethics, such as Elon Musk, were heavily influenced by this seminal literature describing the effects that a superintelligent AI may have on our way of life. This book is for anyone interested in what the interplay between technology and society may look like in the next 50 years. According to our business analyst, the tone of the book shouldn’t be seen as a bad omen, but rather as an opportunity. Although many parts focus on issues such as the control problem, concerns surrounding value alignment and the orthogonality thesis, there is significant weight given to sections that discuss frameworks for international cooperation and institution design. The book ends on a call to action, urging philosophers to focus on problems that are not just interesting, but also useful. AI researchers need to be mindful of broad societal implications of their work and humanity as a whole need not be dismayed, but determined. Developing a strategy for building the right future always starts with asking the right questions, many of which can be found within these pages. Add it to your reading list: The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake Our World Author: Pedro Domingos Technical Level: Medium Publication Date: September 2015 Topic: Machine Learning Apparently our in-house legal counsel has something in common with Bill Gates, because they both highly recommend this book. If you’re interested in learning the basics of machine learning as a non-technical reader, this one is for you. It isn’t as straightforward and simplified as a “ML for dummies”, but it definitely offers basic insight into the techniques and the mathematics required to make this technology work. If you’re thinking of skipping this one because you’re looking to read about AI, reconsider! Don’t assume that machine learning is outside of the realm of artificial intelligence, because this field of computer science is what actually allows systems to learn, so rest assured, you aren’t missing out on what makes AI so fascinating. The author has the ability to often take a dry concept and add imagerie, concrete analogies and even in some instances, humour, to describe machine learning in a way that speaks to everyday people. More specifically, you can expect to learn about the five “tribes” of the ML world (analogizers, evolutionaries, Bayesians, connectionists, and symbolists), the concept of unsupervised learning and the philosophical questions about the future and the use of data for good and/or evil. Add it to your reading list: The Emperor’s New Mind: Concerning Computers, Minds and the Laws of Physics Author: Roger Penrose Technical Level: High Publication Date: February 2003 Topic: Physics Here’s what you need to know about the next author: he’s a physicist, he won the prestigious Wolf prize with his colleague, Stephen Hawking (you might have heard of him) and he firmly believes that machines, no matter how advanced, will never be able to replicate certain facets of the human mind. One of our software developers found the book’s support of Einstein’s assessment that quantum mechanics is incomplete because “it’s statistical, tell us nothing but “probabilities” about individual systems” particularly riveting. The way our mind operates goes far beyond fundamental theory and without it, machines are essentially limited and not limitless. In order to support his argument, the book dives into various subjects such as: Turing machines, complexity theory, quantum mechanics, black holes, white holes, Hawking radiation, entropy, quasicrystals and the structure of the brain just to name a few. Add it to your reading list: Machine Learning From Scratch: Practical Guide with Python (Data Sciences) Author:Alain Kaufmann Technical Level: High Publication Date: March 2017 Topic: Machine Learning Fresh off the press, this practical guide to machine learning and data science in Python focuses on real-world data analysis and its applications. For those with little programming experience looking to break into the field of machine learning, this book takes you through the essential methodology and Python libraries commonly used by industry professionals, including scikit-learn, numpy, and pandas. Speaking to one of our mobile developers, he mentioned how difficult it can be to address the big questions in data science if you don’t know the basics. This book addresses fundamental areas such as linear and logistic regression, using deep learning frameworks like TensorFlow, support vector machines, clustering, decision trees, and generative models. For those not interested in sifting through piles of decade old research articles and dusty textbooks, this book will give you a jump start. Add it to your reading list: Reinforcement Learning: An Introduction (Adaptive Computation and Machine Learning) Author: Richard Sutton and Andrew Barto Technical Level: High Publication Date: November 2018 Topic: Reinforcement Learning When our Chief Scientific Officer recommends a book, you would be wise to give it a read! This classic account of reinforcement learning, first published in 1998, is such a critical text in the machine learning literature, that the authors are publishing a second edition this year. Although the book delves deeply into the mathematical justifications behind reinforcement learning frameworks, the writing is accessible and clear. In addition to providing a comprehensive overview of key ideas like Markov decision processes, dynamic programming, Monte Carlo methods, and temporal-difference learning, the work presents insightful case studies on applications such as AlphaGo Zero, and speculates on what future research may look like. In a true gift to machine learning practitioners, the second edition of the book will come with a fully stocked github repo to guide the reader through the programmable figures outlined in each chapter. Those using the book for self-study can even receive answers to the exercises by contacting Dr. Richard Sutton, who is a world-renowned researcher in reinforcement learning and a researcher at the Alberta Machine Intelligence Institute. This is open-access publication at its finest! Add it to your reading list: What did you think of our recommendations? Share your favorites with us in the comments! Interested in starting your AI journey? Contact us today.