Saturday, June 17, 2017

Lipson, Hod (2016): Driverless - Intelligent Cars and the Road Ahead

What is it about?

This book is about autonomous cars - their past, present and future.

However, the book not only discusses autonomous cars themselves, but to quite a notable degree covers the underlying technologies as well. Perhaps the most notable such examples are artificial intelligence and especially "deep learning", machine vision (with optical, laser and radio technology) and computer and software technology in general.

In addition, the authors very welcomingly discuss also the derivative social effects which would result from widespread adoption of autonomous cars (e.g. loss of certain jobs, changes in urban landscape, economies of urban, sub-urban and rural areas etc. Furthermore, the book does a good job in discussing some of the most obvious ethical questions such as how to value human life, because such a value is needed when an autonomous vehicle is in an emergency situation forced to choose between two or more courses of action all of which involve fatalities and/or damage to property.

Was it good?

The book is very interesting and reads quite well. At places, the authors get close to that stylistic line which irritates me in contemporary non-fiction books (overly colourful language, hyperbolic analogues etc.), but don't get there.

I especially appreciated the extended discussion on underlying or enabling technologies (e.g. how a neural network works, what affects its performance and how this have developed over the past decades, and how neural networks can be and are being employed in machine vision/sensing such as in autonomous vehicles), as this gives one a substantially deeper understanding about the current state and foreseeable future of autonomous cars.

Moreover, I equally liked the discussion concerning societal and ethical issues. This, actually, sets this book apart from may other contemporary non-fiction books especially on technical subjects, because the authors explicitly admit that there currently seems to be to much simplistic hype around autonomous vehicles.

The main take-away for me?

The main take-away for me certainly was an increased understanding about the technical complexity of making autonomous cars reliable and eventually "mainstream". For example, detecting a human progressing slowly while carrying a large dense object as human is very, very difficult to pull off with machine sensing. Yet, this must be routine with 99.9999% accuracy if autonomous vehicles are to become ubiquitous.

Who should read the book?

The book certainly requires some interest in the subject and an engineering mindset (neural networks, laser distance detection...), but anyone at all wondering about how advanced autonomous cars currently are and how (or whether) they become commonplace, should certainly read the book.

The book on  Driverless

Wednesday, June 14, 2017

Line-In Publishing: Sociology - Exploring Human Society

What is it about?

This is a basic text on sociology, like a basic undergraduate textbook on "Sociology 101" course.

Hence, it starts with describing what sociology is (as an academic and intellectual discipline) and how sociological research is conducted, and then proceeds to discuss key themes in sociology such as the society, organizations and organizational behavior, the family, crime etc.

As a very notable feature, all the major topics are discussed through three sociological lenses: functionalist theory, conflict theory and symbolic interaction,

Was it good?

The book is extremely accessible - it discusses the topics in a very down-to-earth way and with no particular hurry or an use of highly specialized vocabulary. At times, this may convey a slowly progressing impression, but every once in a while I found myself really contemplating quite basic issues such as my conception of the family, or how very basic processes operate in my workplace. Thus, the very basic nature of the book actually turned out to be a benefit for me.

Moreover, the choice to run all the key themes through the three lenses - functionalist theory, conflict theory and symbolic interaction is a very good choice: in this manner all the themes appear in different light depending on which of the lenses one uses to look at issues.

The main take-away for me?

My take-away perhaps is at a meta level. Namely, that sociologists have accumulated quite a bit of well-reasoned and researched knowledge which is fully applicable but undervalued in many walks of life. For example, I would claim that at most workplaces people are perplexed by issues (say, for example, difficulty of making changes in work processes) which would be crystal clear and obvious for a sociologist - and even for a sociologist living half a century ago.

Who should read the book?

I would recommend the book for absolutely everyone. Especially if one has not studied sociology before, this book is a stellar place to start.

The book on Sociology

Wednesday, May 31, 2017

Lefèvre, Edwin (1923): Reminiscences of a Stock Operator

What is it about?

The book is a biography of a stock trader (or speculator), reportedly covering the life Jesse Lauriston Livermore.

The book is set in early 20th century (the book was originally published in 1923), and is organized chronologically and organized around highlight events (often particularly successful or unsuccessful or otherwise "teaching" trading campaigns).

Was it good?

The book is extremely good. Not only is it written in a very entertaining and personal style, but it also dispenses quite poignant observations of the human condition.

Indeed, one could say that the psychological observations of human nature and general psychological tendencies are the most valuable content in the book.

And even if one is not interested in any such observations, the book makes quite entertaining reading nonetheless.

The main take-away for me?

Besides the insights about inherent human psychology, a thought that I constantly kept on having throughout the book is "that wouldn't work today". And, in fact, the author (the narrator) admits in the very last pages of the book that his exploits had become increasingly difficult already in the 1920s: there was more stocks traded, more information to digest (impossibly much already in the 1920s), more stringent regulations on insider trading (a very central phenomenon in Livermore's exploits, though he himself was not an insider) and so on.

In any event, the book illustrates very nicely how the speculators got their name and stereotypical character in the early 1920s: already then the speculators were wholly uninterested in the "real" economy, only looking for how the "stocks acted" for the purpose of turning a profit on stock price developments.

Who should read the book?

If one is at all interested in financial economy and the stock market in particular, reading the book - despite its peculiar historical context - is time most assuredly time well spent.

The book on Reminiscences of a Stock Operator

Graeber, David (2014): Debt - The First 5,000 Years

What is it about?

The book describes how people have perceived and interacted through debt until about early 1970s, from the start of recorded history.

The book most certainly has an agenda. This agenda is to suggest that the modern notion of debt, an impersonal numerically expressed money sum involving parties which have little to no personal connection, is an anomaly in the historical record.

Moreover, the author suggests that this calculating, impersonal and mathematical understanding has brought about all kinds of undesirable social effects such as greed, self-centeredness and so on. After all, according to the author, social relations are significantly more healthy if debt includes a social aspect even if debts are generally expected to be paid back in a way or another.

Was it good?

The basic setup, the main argumentative line, and the conclusions certainly are highly interesting, and the author's train of through and evidence-based reasoning is credible. One certainly can't miss the basic message or its support from the historical record.

However, delivering this message takes nearly 500 pages. That's a lot. Most of the pages are used for quite detailed historical descriptions from different eras, which to my taste started to be a bit too much to my taste.

Once again, this book too would be significantly more enjoyable, if the middle 400 pages were compressed into, say, 1/4 of their current length. If appropriately done, I can't see the basic message being diluted a bit.

The main take-away for me?

Well, I presume that the basic message of the book is the main take-away. Namely, how people perceive debt (like any institution or social convention) has significant societal effects. In the case of debt, when debt is being perceived as an impersonal mathematical construct, the morality concerning debt and economic behavior more generally is very different from a society where debt is between people who know each other and interact on a regular basis.

Who should read the book?

In its current form (length), it is not very easy to recommend the book - unless one reads just the first and last 50 pages or so, and cursorily scans everything between. In any case, the basic message is one which should be heard wide and far in the Western world, as it provides a nice contrast to how things are today - and shows that they could be otherwise too.

The book on Debt

Thursday, May 4, 2017

Dormehl, Luke (2017): Thinking Machines - The Quest for Artificial Intelligence--and Where It's Taking Us Next

What is it about?

The book is an excellent and quite accessible overview of artificial intelligence, or AI (what it is, what approaches there are to AI, what AI currently can and can not do), including a historical overview of the origins and early developments of AI.

In addition, and importantly, the book has a good deal of forward-looking discussion about how AI conceivably could develop in the (near) future, and what kinds of questions this could bring about, especially with respect to ethics and legislation (e.g. responsibility questions in driverless cars).

Was it good?

I thoroughly enjoyed the book. The informational contents are - at least for me - in a good balance in terms of basics and "frontiers", and especially the case illustrations (e.g. IBM's Jeopardy-winning AI the Deep Mind) nicely make the discussion concrete.

However, what I really appreciated was that the author successfully resisted the temptation to launch into science fiction-like speculations towards the end of the book (through the notion of conscious AI, the singularity etc. were covered). Instead, all the future-looking and ethics-related discussion is firmly rooted in what current and realistically foreseeable technology could enable.

The main take-away for me?

After reading the book, I probably understand and appreciate more the "mundane" applications of AI (e.g. movie recommending systems, autonomous driving software), and how anticipated developments in the near future may influence our lives and force us to rethink to a degree the premises in our legal systems (e.g. what about if a credit screening system is found to discriminate against a group of people, but because the system is implemented with a neural network, nobody can discern how those credit screening decisions are made?).

Who should read the book?

If one is at all interested in information technology and "big data" or the topic of artificial intelligence in particular, but is not entirely sure what the fuss is about, this book is well worth reading.

The book on Thinking Machines

Tuesday, April 25, 2017

Arbesman, Samuel (2016): Overcomplicated - Technology at the Limits of Comprehension

What is it about?

The basic position of the book is that the technological environment in which we live has become so complex that in many cases we no longer fully understand let alone control how technology works. Especially this is true in large technical systems such as the Internet, or electricity or water distribution.

Thus, complex technological systems exhibit emergent behavior which is at times surprising to us because due to their complexity, their behavior can not be modelled or otherwise anticipated in all possible circumstances.

As examples, complex technological systems cause phenomena such as sudden swings in financial markets (computerized systems trading by themselves at a very high speed) and large-scale electricity outages (a single failure causing a cascade of effects which diffuse system-wide).

According to the author, this "overcomplication" is to a part due to fundamental restrictions in our capacity to process and understand information, and to a part due to the very nature of complex systems/phenomena (c.f. complexity theory).

As a possible way to at least partly tackle the problem, the author suggests that we should increasingly view technology with biological rather than mechanistic metaphors.

Was it good?

The book is quite good and pleasant to read. Well, probably the contents could have been squeezed into one third if not even tighter with the basic message left basically intact -- North American non-fiction books for a general audience seem to suffer from a kind of a syndrome of "having to be of a reputable length".

In any event, the book contains a number of illustrative examples (stock markets, electricity grids, commonly-used software applications, computerized functionality in cars etc.), which make one to really appreciate what the author has to say.

The main take-away for me?

In a way, the book didn't include anything particularly new to me, which I assume to be the case if one is read anything touching on complexity theory and/or technological systems. In any event, the human side of the discussion (how we, as humans, may be inherently incapable of grasping large complex systems) was a welcome and refreshing addition to a "standard treatment". Also, the biological vs. mechanistic metaphors of viewing technological systems (though not worded in this manner by the author) was thought-provoking--the fundamental metaphors with which we view the world surely are quite consequential.

Who should read the book?

The book on Overcomplicated

Monday, April 24, 2017

Lohr, Steve (2015): Data-ism - The Revolution Transforming Decision Making, Consumer Behavior, and Almost Everything Else

What is it about?

The book is, loosely speaking, about big data and its various possible and actual uses in different walks of life, especially in business. Data-ism basically refers to the all-pervasive importance of data especially in the future -- to the tune of "data is the new oil".

The book quite welcomely also includes ethical discussions, especially towards the end of the book, about how much data about our activities in, say, in the Internet can reveal about our preferences - including such preferences which we are unaware ourselves.

The book is quite heavily built around case studies, which for the most part travel with the author throughout the book, from theme to theme. In addition to the case studies, the book also includes general discussion and technology description, but it most often is motivated by an opening case.

Was it good?

The book has its merits and its drawbacks. On the plus side, the book is quite accessible and well-balanced overview of how data can be put into use in different walks of life, and what potential problems this brings about or has brought about. The case studies are also quite interesting and serve a clear purpose.

However, I increasingly like the "make it vivid" style of writing in general audience non-fiction books. In this book, for example, the descriptions of the outer appearances of featured people is striking to a degree of being annoying. Who cares, what is the texture of someone's moustache and how it plays along with the colour of his tie, if one wants to read about predicting customer behavior with social media data.

The main take-away for me?

The main take-away perhaps once again is the increased appreciation of what can be done with data, and how many applications there are for data accumulation, processing and consequent decision-making.

Who should read the book?

The book on Data-ism