My vote for the foggiest assertion of the pandemic to date is that the U.S. has an abysmally high number of COVID-19 cases — more than 29 million as of March 8, 2021 — because of testing. “If you don’t test, you don’t have any cases,” former President Donald Trump said during a televised White House roundtable on June 15, 2020. “If we stopped testing right now, we’d have very few cases, if any.” I wonder how many people who heard that had the same first thought as I did: Correlation is not causation.

This isn’t to say that correlation — the idea that two or more things are associated in some way — isn’t valuable. Indeed, there is big money in correlation. In order to peddle subscriptions, Pandora doesn’t need to know what causes people who listen to The Grateful Dead to also listen to Phish. To bulk up its sales, Amazon doesn’t need to know what causes people who buy a Paleo diet book to also buy beef jerky.

Though associations gleaned from big data drive recommendation engines and bolster corporate revenues, they have their limitations. Imagine trying to control a viral pandemic by refusing to test people for the virus.

The passive observation of data has limited value, because, as Judea Pearl reminds readers several times in The Book of Why: The New Science of Cause and Effect, data is profoundly dumb. “Data can tell you that the people who took a medicine recovered faster than those who did not take it, but they can’t tell you why,” writes the director of UCLA’s Cognitive Systems Lab. “Maybe those who took the medicine did so because they could afford it and would have recovered just as fast without it.” Read the rest here: https://www.strategy-business.com/blog/Seeing-doing-and-imagining

DNF at 21%. I just couldn't follow the logic. It's a great idea and maybe the edited version flows better.

I'm confident in saying that this is a 5-star read for anyone with any knowledge of the kind of complex math that employs letters and squiggly symbols. Alas, it's been a long time since I took advanced calculus so my low rating comes from a total lack of understanding a solid half of what the book talks about.
However! And this is very important! What I did understand is fascinating and very well explained through simple to grasp examples that give a more human angle to the complex topics being discussed. The author's decision to include the histories of theories and biographies of the people involved is a brilliant stroke that made the theory accessible to everyone, at least if only on a surface level.
I can't really say that I would recommend this to any level of reader - unless you're looking for a serious challenge that wakes up some neurons - but I definitely recommend it for anyone interested in AI or just how modern sciences are looking at cause and effect to shape a new future.

Happy thanks to NetGalley and Basic Books for the perplexing and educational read.

Great book, especially if you finding you have to convince people that causal relationships can be modeled from natural experiments (aka experience).

Sorry, but I need to bow out from this review because my education and experience is not sufficient to permit me to give it a fair reading. Thank you for allowing me to have a look at it! Best wishes to the author and publisher.

Causality is an important topic in data science and AI -- one which Pearl rightfully points out is not at all addressed by recent advances in deep learning. Unfortunately, this book's treatment of the topic doesn't really work. It's not quite a popular science book, because you need more than general knowledge to follow a lot of the material, but it's also not a textbook, because the examples and techniques are not fully explained. It's somewhere in the middle, and it's mashed up with a very critical history of Pearl's forebears and intellectual rivals.

Pearl is correct that models of causality and counterfactual thinking will be essential building blocks to achieving any kind of strong AI, and it seems that he and his colleagues have taken some initial steps toward developing the necessary formalisms, but it's unclear whether these are a big leap forward or just baby steps. Pearl clearly claims the former, but the evidence is weak. It appears that considerable human intervention is needed in the creation of even the simplest causal diagrams in Pearls's formalism, which means these techniques can't yet be fully automated.

I suspect we're still at least a decade away from even beginning to create a strong AI. This, of course, is no fault of Pearl's. I hope that this topic continues to draw scientific attention, and that someday a better popular writer can explain it to us more clearly.

This was very confusing. I was lost about halfway through the book. Things were being said that had nothing to do with what I was looking for.

This book is a good reference to understanding the issue of causality. The author provides a lot of examples while also tracing the history of causality. The book will appeal to the curious mind, the one that keeps asking the question "why". The book kindly keeps mathematics at a bare minimum and helps reduce the complexity in it. That said I think the book will appeal only to its direct audience.

One of the things I have begun to really understand in the past ten years is that the onslaught of information is ramping up and it will not slow down. Getting information and making informed decisions isn't as easy as it was 20 years ago. Now one google search can send you down a rabbit hole that can last weeks if not months. The Book of Why discuss information and how we as people view information and how to manage the information we do receive. The book is well written and well researched, very engaging, and doesn't get boring with too much technical jargon that can become overwhelming. Highly recommend.