Machine Learning Intern Journal — My Favourite AI Books
As the title indicates, this is the journal of a Machine Learning (ML) intern at the impactIA Foundation. I’ll be attempting to keep a weekly journal of my activities in the Foundation to keep track of my progress and leave a roadmap for the interns who come after me.
Books, books — books! I come from a family of avid readers. Week evenings after work, weekends throughout the day, holidays on the beach or in the mountains, you’ll find us with books an arm’s length away. When I start approaching the end of a book, I have to take a new book with me because I fear not having anything to read. Anyways, I love books and this week I’m going to give you a list of my favourite books (non-fiction, fiction and textbooks) related to Artificial Intelligence. I’ll also include some that are on my to-read list (denoted by an asterisk, *).
NON-FICTION
Weapons of Math Destruction — Cathy O’Neil
In this day and age of ever increasing technological power, algorithms are used to make decisions for us (where we go to school, whether we get a car loan, how much we pay for health insurance). Naively, we believe these to be fair. But in this book, Cathy O’Neil shows us why the opposite is true. Weapons of Math Destruction (WMDs) are opaque, unregulated and incontestable, even when they’re wrong.
A truly eye opening book, which as an emerging computer scientist I take on with great responsibility. It is up to us, the future designers of these algorithms, to do better. As O’Neil points out, “Data is not going away. Nor are computers — much less mathematics. (…) these models are constructed not just from data but from the choices we make about which data to pay attention to — and which to leave out.”
Another thought which comes to mind is the necessity for technologically savvy government officials. How can we expect appropriate and useful regulations to be put in place when when we have government officials asking Mark Zuckerberg “How does Facebook make money?”.
Although the title screams “nerd alert”, this book remains accessible to anyone interested in how their lives are being dictated by algorithms.
Life 3.0 : Being Human in the Age of Artificial Intelligence — Max Tegmark
I came across this book when I started listening to Lex Fridman’s AI Podcast (I can’t recommend it enough, Fridman is an incredible interviewer and has amazing guests on his podcast).
How can we grow our prosperity through automation without leaving people lacking income or purpose? What career advice should we give today’s kids? How can we make future AI systems more robust, so that they do what we want without crashing, malfunctioning or getting hacked? Should we fear an arms race in lethal autonomous weapons? Will machines eventually outsmart us at all tasks, replacing humans on the job market and perhaps altogether? Will AI help life flourish like never before or give us more power than we can handle?
This book asks a lot of questions and does a great job at presenting different answers and view points.
Homo Deus : A History of Tomorrow—Yuval Noah Harari
I read Sapiens, Harari’s previous book and was blown away. An incredible account of humanity’s history. This book instead focuses on the future of humanity.
Harari puts forward a provocative and terrifying vision of a potential future, one where a small number of elites upgrade themselves through biotechnology and genetic engineering, leaving masses behind and creating the godlike species of the book’s title. If advances in biotechnology continue improving at a rapid pace, what would stop the elite of upgrading themselves out of reach of the masses?
Bill Gates wrote a brilliant review of this book, and I agree with his comments on the ending of the book. Harari presents a new form of religion called ‘Dataism’, in which the greatest moral good is to increase the flow of information. Dataism “has nothing against human experiences,” Harari writes. “It just doesn’t think they have intrinsic value.” The problem is that Dataism doesn’t really help organise people’s lives, because it doesn’t account for the fact that people will always have social needs. Gates says: “Even in a world without war or hunger or disease, we would still value helping, interacting with, and caring for each other.”
De l’autre côté de la machine — Aurélie Jean
Un super livre qui explique extrêmement bien la réalité de l’intelligence artificielle, les bias algorithmiques, et le progrès humain. Facilement accessible, c’est un livre qui me servira comme guide pour expliquer des concepts avancés aux non-techniques.
The Future of Humanity — Michio Kaku
This is the first book I read by Kaku, and you can be sure I’ll be reading more of them (I’m currently reading Hyperspace)! Although there’s a lot of physics in this book, there’s also a fair bit about AI’s role in our future (sending robots into space for exploration, the future of jobs and dangers of AI). He’s got a brilliant way of making complex theories accessible to the general public. I wasn’t expecting such heavy physics theories, but I thoroughly enjoyed reading about general relativity, quantum theory and string theory, and this book reignited my interest in physics!
Rise of the Robots : Technology and the Threat of a Jobless Future — Martin Ford
I read this one a while ago, but it was one of the first books I read about AI and I remember it having an impact on me. In Rise of the Robots, Ford details what machine intelligence and robotics can accomplish, and implores employers, scholars, and policy makers alike to face the implications. The past solutions to technological disruption, especially more training and education, aren’t going to work, and we must decide, now, whether the future will see broad-based prosperity or catastrophic levels of inequality and economic insecurity.
Invisible Women: Data Bias in a World Designed for Men — Caroline Criado Perez
Not sure where to begin, this book is a phenomenal piece of research that exposes the (now) evident lack of sex-disaggregated data in our man-default world. The book is dense with a plethora of references to studies (which at some times can be a bit overwhelming). All in all, I can’t recommend this book highly enough to *every single human* out there, and especially to men. It will really make you think. Being involved in the growing field of AI, I really hope I can help tip the balance towards an equilibrium, where we collect sex-disaggregated data and create a world that caters for women’s needs as much as it has always for men. And of course that starts by listening to women and working with women. A truly mind blowing book.
Superintelligence* — Nick Bostrom
You know a book holds promise if it’s praised by Elon Musk and Bill Gates. This one has been on my list for a while and I really should get around to reading it. Superintelligence asks the questions: what happens when machines surpass humans in general intelligence? Will artificial agents save or destroy us? Nick Bostrom lays the foundation for understanding the future of humanity and intelligent life.
The Master Algorithm : How the Quest for the Ultimate Learning Machine Will Remake Our World *— Pedro Domingos
This book is also highly recommended by Bill Gates. In the world’s top research labs and universities, the race is on to invent the ultimate learning algorithm: one capable of discovering any knowledge from data, and doing anything we want, before we even ask. In The Master Algorithm, Pedro Domingos lifts the veil to give us a peek inside the learning machines that power Google, Amazon, and your smartphone. He assembles a blueprint for the future universal learner — the Master Algorithm — and discusses what it will mean for business, science, and society. If data-ism is today’s philosophy, this book is its bible.
FICTION
It’s while writing this blog that I’ve realised I barely read fiction novels about AI… maybe that’s because I know how misleading it can be. However, there are some exceptions, books that capture fascinating and complex situations and pose important questions. I’ve added 4 classics that are on my list (I’ve already seen the movies, but I nevertheless want to experience the books) but I won’t write about them.
Machines Like Me — Ian McEwan
Machines Like Me was a surprising one, I really wasn’t expecting to be pulled into the story so strongly. As an AI student, I’m always a bit wary about intelligent androids, but this book did a very good job at creating interesting situations which gave rise to probing questions about what makes us humans. And although we’re far away from such intelligent androids, I think it’s important to already be asking and attempting to answer these questions.
I loved the alternate version of history, one where Alan Turing didn’t die in 1954 or opt for chemical castration in lieu of prison after his conviction for homosexuality. Instead, he is alive, knighted, and a driving force behind AI and computational biology. My favourite fake history has to be the inclusion of Demis Hassabis (DeepMind CEO) as a colleague of Turing, helping him create an AI capable of winning at Go and laying down the fundamentals of AGI (Artificial General Intelligence). Other notable wishful historical rewrites include John F. Kennedy’s near-death in Dallas, John Lennon not being assassinated and regrouping with The Beatles, or jimmy Carter winning a second term instead of losing to Ronald Reagan.
Do Androids Dream of Electric Sheep?* — Philip K. Dick
The Moon is a Harsh Mistress* — Robert A. Heinlein
I, Robot *— Isaac Asimov
2001: A Space Odyssey* — Arthur C. Clarke
TEXTBOOKS
Artificial Intelligence : A Modern Approach — Stuart J. Russell & Peter Norvig
This book is a university textbook on artificial intelligence, written by Stuart J. Russell and Peter Norvig. It is used in over 1100 universities worldwide and has been called “the most popular artificial intelligence textbook in the world”. A truly incredible piece of work, and an essential tool for my learnings in AI throughout my university studies and beyond.
Reinforcement Learning : An Introduction — Andrew Barto & Richard S. Sutton
I was introduced to Reinforcement Learning (RL) for my Bachelor project, when I was offered to join a Postdoc researcher at my university to contribute to a project involving biologically-plausible temporal difference learning. This was really a turning point in my studies, as it was the moment where everything seemingly clicked into place, and I had found my calling. RL is a beautiful learning paradigm and I truly believe it is the way towards building robust and general intelligence. It seems to me like the closest thing to biological intelligence, and has actually served in understanding more about biological systems. This is referred to by DeepMind as the ‘virtuous cycle’, where neuroscience gives to AI and AI gives to neuroscience and so on.
Theoretical Neuroscience : Computational & Mathematical Modeling of Neural Systems — Peter Dayan & L. F. Abbott
Another textbook I was given while working on my Bachelor project, this one with a heavy focus on computational neuroscience. The title is quite informative.
Deep Learning* — Ian Goodfellow, Yoshua Bengio & Aaron Courville
The text offers mathematical and conceptual background, covering relevant concepts in linear algebra, probability theory and information theory, numerical computation, and machine learning. It describes deep learning techniques used by practitioners in industry, including deep feedforward networks, regularization, optimization algorithms, convolutional networks, sequence modeling, and practical methodology; and it surveys such applications as natural language processing, speech recognition, computer vision, online recommendation systems, bioinformatics, and videogames.
Finally, the book offers research perspectives, covering such theoretical topics as linear factor models, autoencoders, representation learning, structured probabilistic models, Monte Carlo methods, the partition function, approximate inference, and deep generative models.
