Skip to main content

The Undoing Project

It’s a Michael Lewis book.
That alone, is enough for me to tell you to go read it.

Kahneman and Tversky’s work has probably been the biggest influence on my life in recent years, since Taleb.1
We cannot think in probabilities in our daily lives.
We keep fooling ourselves, with various biases.
And the intuition we have, is because we are really amazing biological machines. And even that is subject to error. Unless the intuition is backed by extensive experience. And even then we can easily be fooled.
That basically is the gist of their work (to me, so far).
I flunked nearly every experiment in Danny Kahneman’s Thinking, Fast and Slow.2
We need to think slowly, through every implication, when it comes to the few big decisions in life.

So how did this work come about?
The Undoing Project tells us the story.
Of Kahneman and Tversky’s friendship and years of collaboration.
Of how a MacArthur Fellow and and Nobel Laureate are in the end, only human, and even they could not undo the events in their lives and their relationship.
Of how the world is patently unfair and never treats people equally.

It’s a lovely read.


  1. Taleb was the one, who introduced me to Kahneman’s work, in the first place 

  2. A Lindy book, if there ever was one 

Change is the Only Constant

If Taleb convinced me that Mathematics was beautiful philosophy, Ben Orlin is the one made me fall in love with it.
Change is the Only Constant is beautiful and funny at the same time.

It’s the story of Calculus over the ages and through domains.
It weaves through life and time, through people, interesting and otherwise.
And the way Ben tells it, it bears no resemblance to the dry crap that is taught in schools and colleges.
It’s beautiful and wonderful, but not paramount and still subject to the vagaries and complexities of life and nature.

In their more insufferable moods, the “hard” sciences like to boast and crow, as if “hard” means complicated and “soft” means simple. This is, of course, exactly backward. The softer the science, the more complex its phenomena.
Physicists can predict what atoms will do. But gather enought atoms, and the calculations grow unwieldy. We need new, emergent laws—chemical laws. Then, gather enought chemicals, and complexity overwhelms us again. We need biology to step in with new theories and rules. And so on down the line. At each tipping point, the role of math evolves: from certain to tentative, from deterministic to statistical, from consensus to controversy. Simple phenomena (like quarks) follow mathematical rules with slavish fidelity. Complex phenomena (like toddlers) less so.

Loved it!


A Hundred Days of Code, Day 022 - Getting into the Groove

Did the same time as yesterday.
Only about an hour.
Was much more prodcutive though.

Getting the hang of how to sit and program and work through things I do not know.
Gaining a bit of experience with the workflow now.
I have the basics in hand. I know what I want to look up.
So check problem, work a bit, look up, try, fail, repeat, gain incremental success, work some more.
Love the feedback loop too.
With other stuff I try, I have to wait days, weeks, months.
Here, it’s immediate.

Beginning to love the work, as I get more familiar with it.
Tomorrow is another day :)

A Hundred Days of Code, Day 020 - Setting up an Editor for Python Development

Had given myself a day, to see if I could get a good Python development environment using Elpy and Emacs.
It does work.
Just not well enough for me.
At the end of the day today, I was happy I learnt so much about Emacs.
But that is not my focus right now. Python is.
Emacs knowledge can come slowly and organically.

So I have kept Emacs as my regular editor for nearly everything text.
And switched to the community edition of Pycharm for all my Python projects.

I am also grateful to all of you, who gave me problems to practice on.
These will help me grow.


A Hundred Days of Code, Day 019 - Future Exercise Addendum

Ok, think I have the problem of writing code, licked.
Will start with one problem from all the Lerner Courses, that I still have to do.
And another problem, I ask my friends to give me.

This should prove for a promising start.
If you, dear reader, have problems in Python that you’d like me, a budding Python programmer to solve, please mail them to me at jason at this domain.
It’ll add to my pool of problems and give me plenty to practice.