Today is the national stress awareness day (at least in the UK). So, I decided to share this book here. You can find the full details of customer reviews analytics at https://books.lanternoid.com/reviews/maybe-you-should-talk-to-someone-9781328662057-lori-gottlieb
Save your time and find the books you like fast. It's all free. If you are interested in a book, you can find it on https://books.lanternoid.com and request for its analytics. I will do my best to put them in the queue of requests and will add them to the website.
If you like this idea, please consider supporting us at https://books.lanternoid.com/subscribe
If you have any suggestions or ideas, please share it in the comments below.
“Algorithms to Live By” is a book by Brian Christian and Tom Griffiths that describes how insights from computer science can be applied to real-life problems. Examples include optimal stopping (referencing the secretary problem), the explore/exploit tradeoff (referencing the multi-armed bandit problem) and work prioritisation (referencing thrashing).
I enjoyed the book, and I imagine most readers of this subreddit would too (probably even if you already have a CS background). You may also be interested in the 80,000 Hours podcast episode interviewing Brian Christian about the book.
However, there’s some gap between “this is a fun and interesting book” to “this book will materially improve your life if you apply its lessons”. I’ve gradually become more sceptical of “life-hacking” over the last few years - there just seems to be something that makes cool, rational-seeming ideas that look good on paper not map to reality all that well. (Though it could be a problem with me specifically).
I’m backed up somewhat by Rob Wiblin, the host of the 80,000 Hours podcast, who, after his interview with Brian Christian, wrote an article called The ‘secretary problem’ is too bad a match for real life to usefully inform our decisions — so please stop citing it. To summarise, there are so many ways in which the traditional secretary problem deviates from real life that Wiblin thinks it’s basically useless in practice. Although there are some modifications to the model that can take deviations into account (which the book goes into), he thinks that in order to be of practical use the model would need to be far more complex than it is now.
I also get the sense that Scott Alexander himself is sceptical of this kind of “applied rationality”. For example, in his AMA, he said:
>My epistemic standards in my own life are "I don't really view my own life as a series of rationality proble... keep reading on reddit ➡
This was much better than expected. A deep dive into computer science - of algorithms and computation. Everyone can to some degree have an intuition about a mathematical reality, but this is often thought of being applied in physics. After this book, it's hard not to think about reality has the computation of information instead, at least as far as a subject is concerned, be an animal or computer.
Some of the insight of algorithms can be surprisingly practical, from how many interviews you should do before hiring someone, how much searching you should do before buying a house, or trying to decide if you should continue searching for a better and closer parking spot, at the expense of losing your current one. Most of these are called the "optimal stopping problem".
A surprising facet of many of these types of problems is that some are incredibly simple. For the optimal stopping problem, for example, in which you have to "quit while ahead", you should search for 37% of your total available time, without making any decisions. Then, at the very next time you have an option better than all of the previous ones, that's the one you should pick. Other problems, however, are incredibly complex, and some aren't even solvable at all. It's also mindblowing how heuristically some of the problems have to be approached, and sometimes randomness is better than any algorithm you can come up with.
While the subtitle of the book is the computer science of human decisions, in my opinion, it shined the most when applied to evolution, and how animals and us have solved computation problems and the adaptations that ensued. The chapters on over-fitting and game theory were fascinating.
The only problem I have with the book is that some of it was hard to follow. While the author did avoid the actual math behind many of these problems, it's still there, just at a more abstract and implicit level. However, I listened to the audiobook, which makes it much harder. Perhaps if one is reading it, it would be much easier to grasp. Nevertheless, the last third of the book was quite accessible.
Overall, a very interesting book, and definitely made me aware of what cognitive science can teach us at both a very practical level of everyday life, but also at fundamental problems of life itself. If you can, reading it would be preferred and you'll likely get a better understanding of many of the algorithms covered.
Thanks for reading. I have a book review newsletter, in which I sen... keep reading on reddit ➡
We are Brian Christian and Tom Griffiths, authors of Algorithms to Live By: The Computer Science of Human Decisions, which was recently named an Audible #1 bestseller and Amazon science book of the year.
We think about how to apply ideas and optimal strategies from computer science to everyday human decision making, from apartment hunting to dating, from sorting your socks to managing your time. Ask us anything!
Edit: We're signing off now -- thanks for the great questions, Reddit!
I bought Algorithms To Live By: The Computer Science of Human Decisions recently at Fully Booked. It's a self-help book. I'm liking it so far but I think it's really lacking a bit more on explaining some algorithms. Has anyone read this book as well? What are your insights?
I understand that the book touches several subjects of study including Economics, CS, OR, and Psychology; however, while I think the main focus of the book is on CS (I am biased), I read a comment on Goodreads by Brian Clegg (a writer with a degree in OR) saying that most of the concepts of the book are part of the field of study of OR. Can someone enlighten me?
By Brian Christian and Tom Griffiths. I think it would be an interesting discussion, if we're still doing book club episodes.