My Q2 2023 Reading List

Selected books I’ve read in Q2 2023, and my thoughts about them.

• Information Theory, Inference, and Learning Algorithms. This is not a book to finish — I read a few chapters every now and then, sometimes just skimming through, sometimes trying for real to understand it all and do exercises as well. Because this is a hard book, with a lot of content, densely packed. I really like it, because of the broad scope — there are not many books combining all those topics together. And this is not an ad hoc topic combination, the author really spots connexions and analogies — such as how inference relates to compression! Last but not least, the language is very neat and even witty, e.g., “exact marginalization is a macho activity”.
• Burn. I was looking forward to the book’s seemingly controversial message “if you do more sports, you won’t burn more calories”, as generally both sports and metabolism are my favourite topics. I’ve learned a number of things here, e.g., about various ways of measuring how many calories one has burned, and how does the body maintain its energy balance. I even did check out some of the cited studies — some of which I felt don’t put the exact same message as the author takes away from them. Additionally, what really burned me during the reading was the use of imperial units all over (which the author himself does not seem to like but targets American audience), and the bit verbose and personal account. Nevertheless, this is a good book, and the message of “all diets work since they all taste bad” is funny. Bonus points go to bashing other books and some paleo diets, tracing all their origins to a single, flawed, cultural atlas from long ago.
• The Man in the High Castle. A bit disappointed by this one — I received strong recommendations, but ultimately, I liked some of the other works by the author more — Ubik, for example, which conducts a similar play on the reader, with the surprise and shock materialising better. Also, the potential of alternate history did not feel fulfilled. What I did like here was how author changed the language based on which person was currently the main actor, which was very spot on given the number of cultures that have a role in the book. And the idea of “algorithmically” generating a book — fun in today’s light of various large language ML models.
• Finding Ultra. This definitively surprised me. I got it just because “looks like dealing with long distance running”. However, it is more an autobiography of an apparently manic guy who went through competitive swimming, alcoholism, law career, vegan eating, and ended up with ultra triathlons. Everything he does has an extreme tag to it, so I don’t think this book is anyhow followable or relatable, even for me. Still, it was an ok read.
• Planet Remade. A good one — I was looking more for a scientific discussion of what solutions are possible, however this one was more a historical one about what has already happened, and how the discussion about climate change and geoengineering evolved in the past hundred or so years. It is a pragmatic, realistic book — neither doom nor calm. It is strongly arguing for doing actions, sometimes in a nicely poetic manner. What I liked the most was a chapter about volcano eruptions — there are a lot of interesting consequences to those.
• Hunt, Gather, Parent. This was a very influential book, especially for my wife, as we are having a newborn. It shares some ground with the Burn book reviewed above, both studying indigenous cultures, albeit with a different aim. There were some quite deep ideas, e.g., about evolutional benefits of having offsprings that take very long time to become at least somehow self-sufficient. The beneficial lessons of the book are not restricted to just parenting, but to general people interactions — sadly, I’m not overly social type, but the message of not pushing resonates anyway.
• The Art of Doing Science and Engineering. As a first-hand account of a prolific scientist, this was a good read. I liked the confident tone, the way which his opinions seem to have been justified retrospectively (the book came almost 30 years ago), the underlying messages, and the connexions across domains.
• Radical Longevity. Not my favourite. Compared to, e.g., Sinclair’s Longevity book, which reads smoothly, is motivating, and wasn’t overwhelming, this one is zealous in exhaustive listings of all the dos and don’ts, weird in some recommendations (like homeopathy), and not really holding together. Still, there are some listings and tables which I think I’ll find practical as a reference or inspiration.
• Fast Python for Data Science. It is a technical book, but reads quite smoothly, and contains a good mix of general advice and directly applicable things or examples. It is neither too easy nor too advanced, and the focus on data applications makes good sense as it makes the book very compact.
• Architecture Modernization. I was reviewing this book for Manning, in quite a tight deadline, so had to do a speed read. I think that contributed a bit to my feeling of the book being somehow repetitive, verbose, and trying to address too many topics. It is actually not just about architecture, but about product and organizational strategies, domain designs, platforms, … So more like a “all non-architecture things an architect should know to have an impact”. This made the book a worthwhile read in the end.
• Earth System Science. Not that it was bad, but at this stage of my knowledge it didn’t give me much new information. Even as an introductory one, I think it is beaten by Archer’s Global Warming. Still, it is short and compact, so a possibly useful reference or a refresher.
• Under a White Sky. A short read with anecdotes of how humans are capable of destroying the environment even when starting with good intentions. Could be viewed as a counter-read to Planet Remade. Not that this book is pessimistic — its tone is objective, neutral, explorative. And similarly, Planet Remade is not naively optimistic or urging without caution.
• Mars Trilogy. I have to compare this to the Remembrance of Earth’s Past — both are trilogies, both progress to the point of changing genres or sub-genres by the last book, both deal with space. But where Remembrance is more thriller-y or space operaic, Mars Trilogy is more about engineering, terraforming, society, without science-beyond-believable. Also, the overall tone is rather happier, a solar punk. However, the Mars Trilogy was quite dragging and even boring, especially the last volume. Geriatric. Still, it was an important read to be made.
• Warhammer books — Regular dosage, with Heresy Series (Battle for the Abyss, Mechanicum, Tales of Heresy, Fallen Angels, A Thousand Sons), Dawn of Fire Series (Indomitus, Throne of Light, Iron Kingdom), as well as Ahriman: Exodus.