Issues With the CFAR (Center for Applied Rationality) Reading List, by Section
(there are three sections)
The CFAR has a reading list for rationality texts. While most of these books are good, many are inappropriate for a reading list devoted to rationality. I have done a line-item assessment of the choices as of 01/05/23 below. I have also made an alternative recommendation list.
Section 1. Popular
ENDORSE: Thinking Fast and Slow (Kahneman). This is a foundational book on rationality, and this fact is widely known. I don’t need to elaborate further.
OPPOSE: Rationality: From AI to Zombies (Yudkowsky) — you may find this counterintuitive, because "rationality" is the first word in the book's title. However, Yudkowsky is confused about what rationality *is*, and his definition tends to include many of his pet interests, such as evolutionary psychology and metaethics; these are interesting, sure, but evolutionary psychology is descriptive psychology, not rationality, and metaethics is philosophy, not rationality. (You would viscerally understand why metaethics is not the study of rationality if you've ever endured a multi-hour debate about moral realism. Likewise with evo-psych; "did humans evolve to make racist judgments?" is an interesting question, and something evolutionary psychology would study, but has few if any implications for the study of rationality itself.) The book itself is just the LessWrong sequences, and as I've mentioned many times, the LessWrong sequences are only somewhat about rationality. AI, for example, is not rationality; we may apply principles of rationality to AI, but we can also apply principles of rationality to sports nutrition, or to the management of record labels. The study of AI itself is not the study of rationality.
AGNOSTIC: The Scout Mindset (Galef). I have not read this and because it is a popular book, I am not quite able to glean what I'm getting from the table of contents like I can with an academic book.
OPPOSE: Superforecasting (Tetlock). I have read this, and while it's an interesting book about how advanced predictions are made, the majority of the text has quite little to do with rationality and more to do with how properly selecting the right non-experts can outperform experts.
(I could make a similar case; I am not a drummer, but won the digital "World's Fastest Drummer" contest in late 2022. There are several dozen rhythm gamers who I believe would place in the global top 100, if not top 10, at WFD-style competitions with minimal training — even when using sticks. In this case, the rhythm gamers would be "non-experts" because I've identified superior selection criteria for this task, much like how Tetlock developed superior selection criteria for identifying prediction-makers compared to the FBI's or whichever organization it was. This is the majority of the book, and it's an interesting argument, but I don't think it has much to edify the reader on rationality itself.)
OPPOSE: Focusing (Gendlin). Note: I have not read this, and only read reviews, summaries, and the table of contents. This is a psychological self-help book for people who could use improvement in their ability to focus. This is in a similar genre to Robert Leahy's "The Worry Cure", which deals with eliminating anxious thoughts, or Menno Henselmans's "The Science of Self-Control", which overviews many studies on self-control and what we can learn from them. I have read Leahy's "Worry Cure" and Henselmans's "Science" and can say they are great books, but I would oppose their inclusion in a list of rationality books on similar grounds: simply improving yourself psychologically is not a study of rationality, nor necessarily an application of what is known from the rationality literature; this is applied psychology, not applied rationality. To put this differently: “12 Rules for Life” would also be an inappropriate choice, and for similar reasons.
OPPOSE: Influence (Cialdini). I have read this book, and while I think it's an excellent list of persuasion techniques, persuasion is not rationality nor even applied rationality; it is applied psychology, and the same objections to including "Focusing" apply here. Simply adding an applied psychology book because it helps achieve your goals would be like if I added the famous "Don't Talk To The Police" lecture because not getting arrested is also in your best interest, or if I added "The Personal MBA" by Kaufman — also a good book — because knowing how to make money in your best interest, or if I added self-defense weaponry and brazilian jiu-jitsu guides for the same reason, or to take this to its logical extreme if I compiled a list of ways to rob people because the clearance rate for burglary is only about 15%, which means 85% are successful, and with the average convicted criminal's IQ being significantly (as much as 1SD in some cases) lower than the mean, imagine how much more a smart person could... etc, you get the idea. Simply "that which helps you" is not rationality.
ENDORSE: What Intelligence Tests Miss (Stanovich). I have not read this. I have, however, heard it extensively discussed and have a very good idea of what it is about. The book *is* about rationality, and how one might test for rationality, and appropriate for a rationality reading list.
AGNOSTIC: Decisive (Chip/Heath). The table of contents give me the impression that this is a book that improves your day-to-day decision-making. While helpful, this is not strictly speaking rationality, and this is not hair-splitting. For example, I could make a book of similar techniques like "write the most important things on physical notes" and "make sure the most important things you need to remember are in the most visible places." These would help you make better decisions because you would be more organized, but you have not actually *overridden an irrationality* so much as applied a technique that has the *consequence* of a more rational behavior. Nonetheless, I have the impression that I could be wrong about how much this addresses rationality itself, so I'm weakly categorizing it as "agnostic".
Section 2. Academic.
Universally endorse all of this section.
I will caveat this by noting that this section has a heavy emphasis on decision-making rather than reasoning, and especially probabilistic decision making. Reasoning is as important, and sometimes more, than decision-making; when making big decisions, reasons usually precede decisions. Further, reasoning is often if not usually social, while decision-making can be as much asocial as it is social. It is for example imperative to be a good reasoner if you are going to do to do business or politics at all, and avoiding these things is for many not a realistic option in the adult world. Therefore, I would include more texts such as those I've included in my personal reading recommendation list.
Section 3. Staff Picks
Universally oppose, with the exception of Mullainathan and Shafir's "Scarcity".
While these are fun, insightful, informative and helpful books, they engage rationality little if at all. For example, "Algorithms to Live By" is quite interesting, but appears to make the case that the human brain is more computational than we'd give it credit for, rather than focusing on rationality-advancing content as a dominant theme. "Scarcity" is an exception because the book’s second section appears to show how scarcity drives people to irrational behavior.