In the last year and a half, I read about 30 books. Here are five great books that I read more than once and found myself going back again and again for practical applications.
Disclosure: the book links are Amazon affiliate links—when you purchase by following the link, at no extra cost to you, I get about 5% commission for each sale. You don’t have to use them, and I want to thank you for supporting my blog if you do.
I found great enlightenment in reading them, so I want to share with you.
The 5 books share a common theme:
- all of them gave me one or more idea to see things differently.
- all of them changed the way of how I do things.
Therefore, for each of the five books, I will give you
- the number of times I have read so far,
- one example of mental model from the book, and
- one application of the mental model
The books are listed in no particular order.
1. Atomic Habits
Author: James Clear
Listened to 3 times on Audible. Later I felt that I need to use it for reference, so I also bought the Kindle version.
There are no good habit or bad habit. The only difference between habits is when does the reward come.
For example, if my goal is to control my weight, then I would probably label
- eating a big burger (e.g. a Wendy’s Baconator) as bad, and
- working out in the gym as good.
But the difference is whether the reward is immediate or delayed because both activities result in multiple outcomes:
- Eating a burger provides an immediate reward. It is tasty, so I feel good now. But the delayed outcome feels bad.
- Working out is hard. Most of the time, pushing myself through a workout feels terrible when I’m in the gym, so the immediate outcome is bad. But the delayed outcome is good because I’m healthier.
When it comes to building habits, our body is geared toward instant gratification. Therefore, it’s harder to build a workout habit than an eating-a-burger-a-day habit.
The problem is that, when we build a habit, most of the time we want to reap long term benefits—how can I reconcile what my body wants with what my mind wants?
One solution he proposed in the book is to trick my brain into instant gratification by visuals to track the progress.
For example, I can put a calendar on my fridge and put a check mark whenever I finish my workout. Now, whenever I put a new check mark on the paper, I feel good. This is what I have been putting on my fridge since last November:
Each blue check mark means I wrote at least 25 minutes that day. Each red mark is a workout.
It has been 4 months since I started tracking on a calendar, and I still write and work out almost every day—which had never happened before in my life. I’d say that the trick worked. (Not sure how long it will last. We’ll find out. 😜)
2. High Output Management
Author: Andrew Grove, CEO of Intel in the 80s and 90s
Read once on paperback. Reading second time now. (I got the book 4 weeks ago.)
This is a book full of gold for data scientists. There is no way I can do justice to how good the mental model in this book is, but I’m glad that it exists.
His proposition: a manager’s output is the sum of
- the output of his organization, and
- the output of the neighboring organizations under his influence.
He said that a manager’s most precious resource is time. So he or she should think about what’s the best way to use their time to increase other’s output.
What’s the best way to increase a manager’s output? By engaging in activities with high managerial leverage. How does a manager become more effective? One way is to think about how to increase the leverage associated with the activities.
What are some high leverage activities that has high leverage?
- Performance reviews
What are some low leverage activities or negative leverage activities?
- Meetings without objective
- Delayed decisions
In my opinion, every data scientist is a middle manager, because our output depends on the success of business. Therefore, I need to learn how to manage, which I’m terrible at.
For one thing, the book has forced me to rethink about what I do daily—do I spend most of my day in low leverage activities? Do I look busy because I sprinkle my day with lots of motions that resulted in little output? What are the high leverage activities for data scientists?
For example, it is tempting to call a meeting whenever there is a question or whenever I want an update. But Andrew said in the book:
Before calling a meeting, ask yourself: What am I trying to accomplish? Then ask, is a meeting necessary?
So now I think twice. I also think twice before I accept a meeting.
Other than traditional business management activities, some high leverage data science activities I can think of:
- Write good documentation—saves tremendous onboarding time
- Packaging code for reuse—scalable to other teammates.
There are so many great mental model in the book besides managerial leverage. For instance,
- The basics of production
- The purpose of performance reviews
- The decision making process
etc. I plan to make another post or two about this book later.
3. Never Split the Difference
Author: Chris Voss, retired lead FBI hostage negotiator
Listened to 3 times on Audible
The author presented an idea:
Negotiation is about getting information.
I never saw it this way. But a great negotiator knows how to discover information that he or she didn’t have when the process started e.g.
- What the other side really wants (often not what they say)
- The “Black Swans”—information that would completely change the course of negotiation when discovered.
Therefore, much of the book was devoted to how to get more information out in a negotiation from the other side. From this principle, many thoughts in the book follows.
For example, he said that
No deal is better than a bad deal.
Saying “No” is always better than meeting in the middle. If you accept that negotiation is about getting information, then by meeting in the middle, it probably means that
- you didn’t work hard enough to find a great deal.
The better alternative is to gather more information and either
- turn it into a great deal, or
- walk away.
I used to struggle with the ambiguity in business problems. The problem is that, for most data science projects,
- the business also does not know what they want.
This book changed how I view data science projects and negotiations. Now I spend as long as needed to negotiate and discover what the business really need before I build the solution.
As a non-native English speaker, some of the conversation techniques in the book blew my mind. But I guess if anyone has gotten negotiation down to a science, it would be the FBI hostage negotiators.
PS. I shared this book with a friend when he was looking for a new job. He had one offer, but it only looked “ok-ish.” It was a job that looked like what he wanted to do, but the compensation seemed low. He was about to take the offer but decided to use the framework in the book to delay the commitment and learn more about the job.
In the end, he said no to the offer. Two months later, he got a great position in a company he liked with a great pay raise.
4. The War of Art
Author: Steven Pressfield
I read 5 times on paperback.
The book is structured around what the author called Resistance—the “thing” that prevents anyone from sitting down and start on doing creative work.
Such resistance do exist. When I am writing a blog post, Resistance tells me
- Your writing is terrible.
- Who is going to read it?
- Your English sucks.
- You code sucks.
- Don’t you think you are in above your head?
And the way to beat resistance is to turn pro.
Just like we go to work every day whether we feel it or not, to beat Resistance, I can’t wait for the inspiration to come. Like what Steven said: sit down and do the work.
Someone once asked Somerset Maugham if he wrote on a schedule or only when struck by inspiration. “I wrote only when inspiration strikes,” he replied. “Fortunately it strikes every morning at nine o’clock sharp.”
That’s a pro.
Honestly, I read the book for the first time in late 2017, but I still had a long lapse where I couldn’t produce anything (you can see it from my publishing history.) Maybe it was because that I didn’t play for money like a pro—another point in the book.
It wasn’t until I read Atomic Habits that I found a way to combine the “turning pro” mentality with a habit that builds on it.
But this blog won’t exist if I didn’t read the book. Before I penned my first post in English, I didn’t think I was good enough to write non-academic articles. The book gave me the push I needed, and the publication of the first post on this blog—How I Learned Data Science Skills as a Math PhD was a direct result from reading the book.
5. Show and Tell
Author: Dan Roam
Read cover to cover about 6 times on hardcover. Numerous time for reference when I draw.
I learned about this book when I went to David Neal’s talk at Music City Code in 2017. David likes to use drawings for his slides, and he recommended this book for anyone who wants to try it.
Presentation is about bringing change to the audience
One thing I learned in graduate school was that:
- Academic talks suck.
Well, probably not all of them, but most of them suck. When I think about it, most talks suck because they were about the presenter—their ideas, their experiments, and their accomplishments. They didn’t try to bring change to the audience. They didn’t help the audience
- see what they see, or
- change the way they do things.
Therefore, whenever I’m making a presentation today, I want my audience to walk away with a change. They should be different after listened to my talk. It could be their view, their thought process, their emotion, their ways of doing things—but a great talk starts from bringing change to the audience.
Like what Dan said in the book,
If we don’t change our audience in some way, what’s the point of making our presentation?
When I start working on a talk or a business presentation, I would keep asking myself—what is the one change I want to see in my audience? And work backwards.
The book also has many practical advices on how to draw stick figures. I probably won’t even attempt to draw stick figures if it weren’t for David’s talk and this book, but I have learned that simple figures communicate much better than stock photos or bullet points, so I try.