Switching from reading blog posts to reading books (digital and physical ones) substantially improved my mental health since sometimes blog posts abstract to thousands of hours of information in small (of what I call) information bombs that make us believe we learn when in reality it’s just filling the gaps with noise or craters, these bombs I mentioned have the logical structure as follows:

  1. as far as we know X, everyone knows in the industry this must be done as Y, so we should
  2. intuitively, this should be done following X
  3. any general assumption as a source of truth

These logical (but not harmful) structures assume every reader has a level of knowledge in the same range as their target audience, not even disclosing to the audience in the first place what they are trying to achieve while trying to be as broad as possible. For example, trying to bring down-to-earth complex topics that are just that, complex topics that, in the end, the simple explanation ends up being as complex to understand as the original one.

The Feynman way of learning is entirely antagonistic to this way of learning; articles are just pills of information trying to be as quick as possible for the reader to read and move on.

Link aggregators such as slashdot or hn make us addicted to some data. Not information per se, but data.

The reader who ingests knowledge without knowing these gaps in his mind will find it useless to try to put into practice what he just read since they don’t have the right tools to do it.

Try to do it yourself. Pick the top five articles of the month in any aggregator you visit frequently, summarize what you have learned from them, and apply those learnings to a piece of work you have been doing.

Now, why books?

Books might be awful, and articles might be really good; however, whenever you select a book, you “try” to spot that one you like, reading the content and trying to understand what the author is trying to put into words.

At least personally, I googled the author multiple times before buying it from him, and just from there, I bought it. I read. I devour. I put the knowledge (if it’s any good) into practice, and I move again.

I use articles on the opposite side to support an argument or discussion. They might be really poorly written, although if they align with my way of thinking, I might be interested in reading them and marking them as a bookmark. They might be helpful later in the future in a potential discussion whenever I try to use them as an example.

Articles have knowledge (what’s knowledge anyway?). Of course, they do. However, the danger with these is the amount of assumptions the authors make before understanding which kind of readers they might have. It’s not an issue on the author, of course, but before reading one, we should be prepared to ingest complete nonsense for our current state of mind.

Any good book before starting it has an “Audience” part before any technical part. This tries to encapsulate if this book is for you.

Of course, this is not a problem by itself since information has been like this since the origins of magazines, but reading a quick small article might make us believe that we understand what the author wanted to express or follow a topic when, in reality, we are just following some author’s POV trying to put into his own words the topic he is attempting to describe.

Following the “Let’s build X” article might do more harm than good (honestly trying to find a place in mind where it’s not) since usually, the person writing is not a teacher, the person reading those have the knowledge required, and the person writing might just want to “show” themselves what they’re capable of.

Being a teacher is something only some know how to do. Being a teacher is doing the Feynman way and going from first principles to a Nobel prize.

Fortunately, a bunch of folks on the internet write excellent technical content, and most of those are behind a paywall for a reason (@cmuratori (Computer, Enhance!) first comes to my mind). In addition, I’d like to mention that it’s good to see long-form discussions with experts such as (Foucault vs Chomsky and Jonathan Blow about compilers)

Personally, I’ll stick with books for long-form content and articles (link aggregators) as short-form real-time data (kind of like this since I don’t read newspapers). I wouldn’t even know about the xz exploit if it wasn’t for hackernews