Three Top books about Artificial Intelligence and Machine Learning

In this post I am going to talk about three books about AI and Machine Learning that I find a must read for every lover of these topics. The experiences obtained from reading these books vary widely: one of them will talk about where Artificial Intelligence is heading, the different paths towards it, the possible dangers, and the ways to handle these dangers.

One of the books will dissect the functioning of the brain, and resume the similarities of what goes on under the hood of our neural systems with what happens inside a Machine Learning model.

The last one speaks about the different Machine Learning families, the ultimate goal of Machine Learning (from a research perspective mainly), where are we in relation to this goal, and what is the most likely way that we will reach it.

If this sounds interesting, read on to find about these fantastic pieces of literature, and don’t worry, I only speak superficially about the content of the Books to encourage you to read them, there are absolutely no spoilers 🙂

Superintelligence by Nick Bostrom

The Unfinished Fable of the Sparrows:

It was the nest-building season, but after days of long hard work, the sparrows sat in the evening glow, relaxing and chirping away.

"We are all so small and weak. Imagine how easy life would be if we had an owl that could help us build our nests!"

"Yes!" said another, "And we could use it to look after our elderly and our young"

"It could give us advice and keep an eye out for the neighbourhood cat" added a third.

Then Pastus, the elder-bird spoke: "Let us send out scouts in all directions and try to find an abandoned owlet somewhere, or maybe an egg. A crow chick might also do, or a baby weasel. This could be the best thing that ever happened to us, at least since the opening of the Pavilion of Unlimited Grain in yonder backyard."

The flock was exhilarated, and sparrows everywhere started chirping at the top of their lungs.

Only Scronkfinkle, a one-eyed sparrow with a fretful temperament, was unconvinced of the wisdom of the endeavour. Qouth he: "This will surely be our undoing. Should we not give some thought to the art of owl- domestication and owl-taming first, before we bring such creature into our midst?"

"Taming an owl sounds like an exceedingly difficult thing to do. It will be enough to find an owl egg. So let us start there. After we have succeeded in raising an owl, then we can think about taking on this other challenge." Replied Pastus.

"There is a flaw in that plan!" squeaked Scronkfinkle; but his protests were in vain as the flock had already lifted off to start implementing the directives set out by Pastus.

Just two or three sparrows remained behind. Together they began to try to work out how owls might be tamed or domesticated. They soon realised Pastus had been right: this was an excellently difficult challenge, especially in the absence of an actual owl to practice on. Nevertheless they pressed on as best they could, constantly fearing that the flock might return with an owl egg before a solution to the control problem had been found.

This tale starts _SuperIntelligence, a book by the Swedish philosopher Nick Bostrom, that debates the issue of machines becoming more intelligent than humans in the near or not so near future. The Unfinished Fable of the Sparrows_ refers to the issue of the control of an entity which we are not completely familiar with, and whose powers might be beyond our individual reach.

Although the time is not precisely known, it is widely accepted between the top personalities of the field of artificial intelligence, that the day will come when machines reach the same level of intelligence as humans.

When this day comes, will it stop there? Or will machines, using their newly achieved human-level intelligence, surpass this threshold and achieve even higher levels of consciousness and knowledge than humans, allowing them to dominate the world, or even the entire universe?

What is intelligence? what forms of intelligence exist? what is the most likely way in which machines can achieve this super-intelligence? Will we be prepared for this event? If not, how can we prepare? What are the possible outcomes? Is there any specific motivation for the behaviour of an artificial intelligence?

These questions and a lot more are answered by Bostrom in the approximately 320 pages of text, diagrams and images that compromise this wonderful book. If you want to know the answer to any of them, or just educate yourself on the possible shortcomings of Artificial Intelligence, this book is absolutely a must read that leaves no stone unturned.

How to Create a Mind by Ray Kurzweil

"If a machine can prove indistinguishable from a human, we should award it the respect we would to a human – we should accept that it has a mind"

In How to Create a Mind, Ray Kurzweil, director of engineering at Google, exposes his theory of the functioning of the brain, based on the existence of hierarchical pattern recognisers, from which we project reality from its lowest level of abstraction to its highest dimension. You have probably noticed that I have used the term "his theory".

This is because even tough in the last decade there have been some amazing achievements in the fields of neurology and neuroscience, there is still a lot of fog around our knowledge of the brain, and for this reason what is described in this text is no more than that; a theory.

Throughout the book we can find many analogies between the underlying processes that are carried out inside our brains, and the fundamental processes that sustain a wide variety of technologies, mainly machine learning based, like virtual assistants like Alexa and Siri, or self driving cars for example.

There are a lot more similarities between the approximately 350 gram mass that we have inside our heads, and the computer from where this article has been written than we know of.

Kurzweil also exposes the improvement of human capabilities over the centuries with the use of technology: from the first axe that was used to cut a tree to the smartphone that we all carry around nowadays, and that are basically an extension of our own body. Who knows, maybe in a couple of decades we will use devices similar to these but that are directly integrated into our biology.

Lastly, to end the book, there are a couple of chapters dedicated to the philosophical and metaphysical aspects of the mind like consciousness, free will, or the concept of identity.

Although being oriented towards a reader with a technical background or interest, this book can definitely be enjoyed by anybody who wishes to wonder into how the brain could work, and to enter the mind of one of the greatest technological thinkers of our time.

"Because important things come in a case, you’ve got a skull for your brain, a plastic sleeve for your comb, and a wallet for your money"

The Master Algorithm by Pedro Domingos

"All knowledge – past, present and future – can be derived from data by a single, universal learning algorithm"

The Machine Learning world, can be broadly divided into 5 different continents, each continent representing a specific family of methods or algorithms that differs from the rest either on the paradigm that originates it, on the underlying methods to each family, or on the ways that the algorithms work. Because of this, each of them is good on solving a specific problem, and has a couple of specific use case applications.

However, they also all have one thing in common: discovering hidden insights in data and using these insights to generate some kind of value. The five families mentioned in the previous paragraph are: Bayesians, Connectionists, Evolutionaries, Analogizers, and _Symbolists._

The book starts with a small introduction to machine learning, followed by the main motivation of the work of the author: finding an universal machine learning algorithm that can be used to solve any kind of problem. Algorithms that combine two or more of the aforementioned families have been developed already, gathering the virtues of their constituent families, but none of them has managed to unite all of them.

Then each family is described, with their history, pros, cons, and main algorithms. Lastly, Domingos speaks about the upcoming future of AI, describing the virtues and capabilities of this universal algorithm, but largely leaving aside the possible dangers and challenges that were mentioned in the previous books.

Although this text is a very good read even for a non – machine learning practitioner, it is most enjoyable if we do have some kind of knowledge of each family and have used different kinds of algorithms like Naive Bayes, SVMs or Neural Networks at least to some point. Unlike the two previous books, it is quite focused on Machine Learning, however everything is very very well explained with a lot of examples, analogies and diagrams.

"One Algorithm to rule them all, One algorithm to find them, One Algorithm to bring them all and in the darkness bind them, In the Land of Learning where the Data lies"

Closing words

As always, I hope you enjoyed the post, and that I have convinced you to read at least one of the books. They are all fantastic pieces of work, and I could have written an in-depth article about each of them, but from my point of view, they fit very well together as a pack, and if you read them all you will get an incredible overview of the current state of Artificial Intelligence.

Here you can find links to each one of them.

_Feel free to connect with me on LinkedIn or follow me on Twitter at @jaimezorno. Also, you can take a look at my other posts on Data Science and Machine Learning here. Have a good read!_

To find more books on Artificial Intelligence like the ones review in this article, check out the following repo: