Here’s a snapshot of what’s on my reading shelf right now at the start of 2022. I’ve finished or half finished reading some of these, but I think they are interesting enough to share here.
Books left on the shelf at the end of 2021
Couples That Work
I’m halfway through this book by Jennifer Petriglieri, which marks my second time reading this book since it came out in late 2019 when I first purchased it. At that time, I was busy with my last year of college, and I read the book mostly as a “prepatory guide” if you will, for the future when both my partner and I will be working adults. Now that I’m a year into my job, the topic of how two can have it all is more pertinent than ever (‘all’ being - a great career, family, and relationship).
The book is a summary of the author’s interviews with numerous working couples in various stages of their lives on how they juggle their careers and everything else. I personally would have preferred to read more detailed case studies, though that would probably make the book less generally appealing. To spoil it for you, the gist of it is that there are various “models” of relationships (with respect to career and non-career management) that can work out for different couples. The key here is to communicate explicitly the plan, the roles, the boundaries, and the expectations (e.g. I veto any move to Montana).
The model that reportedly generates the most self-reported satisfaction is the double-primary model, in which the two partners are both primary breadwinners, and neither’s career generally takes a backseat for that of the other. I think this is what many naively aspire to attain at the start of their working lives, but eventually with family and more responsibilities, it obviously requires a lot more work and planning to sustain, especially when compared to other models e.g. the primary-secondary model, where one partner’s career takes on a more supplementary role in the context of the family dynamic.
You might already be doing a good number of the things recommended in this book. Overall I think this is a piece of good general light reading, and I strongly recommend it.
Performance Modeling and Design of Computer Systems: Queuing Theory in Action
In my short time at CMU I’ve unfortunately never had the opportunity in my schedule to take a class by Professor Mor Harchol-Balter. Reading this book is my attempt at making up for that. Off the bat, I think the topics covered in this book are quite interesting and very relevant. We always have a limited amount of compute available at any one time, and queuing is an annoying fact of life. Can we mitigate the problem by throwing more compute at it? If so, by how much? Does the “express checkout lane” strategy work? And when does it not work?
I’m only a few chapters in, so read it to find out yourself!
The Changing World Order: Why Nations Succeed and Fail
I originally got this as an audiobook, and there were just too many references to charts in PDFs that I had to just go buy the paperback copy to actually look at these charts. I was slightly disappointed he only narrated the first chapter. Ray Dalio in this book puts forth his model for the cyclical development of nations, states, and empires, and tries to model the existing state of the world using it.
My favorite part of this book is the lengthy historical perspective that Ray Dalio provides, which I think adds value regardless of whether one agrees or disagrees with the model and the theories put forth. A lot of things have happened in the short time I have been on this Earth, but a lot more has happened before that. Events that may seem out of the ordinary can and have happened many times in history, perhaps not in the last 50 years, but certainly in the last hundred, or few hundred. This does adjust one’s prior probabilities for future events significantly.
SpaCy is one of my favorite NLP frameworks. I especially like the work that has been put into the ecosystem, the documentation, and the surrounding tooling. I’ve been a superficial consumer of this library for a time, and I’ve been meaning to spend more time familiarizing myself with more parts of the library. This book by Duygu Altinok does provide that kind of tour, with sufficient depth to do quite a reasonable lot. It’s also written for SpaCy 3.0, which is the latest major version with many changes, in particular with regards to model training. I recommend this as a first book for SpaCy, if you are in need of a book. The online documentation and tutorials are pretty great already!
Python 3 Text Processing with NLTK 3 Bookbook
I can’t say I’m a great fan of using NLTK. Fortunately or unfortunately when I started my journey with NLP, SpaCy already existed, and that set the bar for what a “nice NLP package” should be like. That said, I’ve come to appreicate NLTK more lately, in particular for what it offered people at a time before SpaCy existed. There are still many things that NLTK does right now that SpaCy doesn’t do. For example, one task I’ve found myself often somewhat needing is lemmatizing using some dataset, such as WordNet. I’ve also come to appreciate that the two libraries exist to offer two different sets of functionalities. SpaCy is something you’d like to use in development and production - it’s well tested, well-documented, and has canonical ways of performing a a particular task well and fast. NLTK is something of a swiss army knife where you have to build more things yourself, but in return you get more control over the finer details. In any case, I’d like to learn more about what NLTK has to offer and what might we be doing that can be better done in NLTK compared to SpaCy.
人工智能基础 (Fundamentals of AI)
Once upon a time I had to do a poster presentation on my paper about the ethical considerations of self-driving cars, when a Chinese gentleman came along and asked me about my poster in Chinese. It dawned upon me that I did not know how to translate “self-driving car” in Chinese, or at least, I did not know the canoncial term for it. Naturally the conversation that continued was somewhat awkward.
So anyway, on a related topic, I came across the fact that China was doing this textbook for their AI curriculum for high school students (yes they seem to be light years ahead in this regard), and thought I should grab a copy, since my partner thinks that my Chinese ability belongs in high school or below anyway. I think this is a great high school textbook honestly, it actually does a good job in describing many important AI concepts at a high enough level with practical examples, without resorting to significant math. It covers topics like the history of AI, classifiers, computer vision, self driving cars, speech to text processing, and I’m just left wondering why governments around the world aren’t already translating this to be taught in their own schools (I bought this 2 years ago).
Anyway, grab a copy and see for yourself how behind we all are.