You are reading a newsletter I shared with my mailing list. If you like it (and I hope you do), you can get a newsletter like this every Sunday by signing up here.
Hello! Hope you had a great week.

This week, we start with weightlifting. Turns out there's quite a lot this brawny sport can teach us about learning and decision making in a complex world.

Next, we look at the limits of decision-making algorithms. What Clinton's campaign found out about the limitations of algorithms, and how we can remind ourselves of our computers' fallibility.

Then, the Mental Model of the Week - The Map is not the Territory. A way to remind ourselves that the world is not as simple as we think it is. 

Here's the deal - Dive as deep as you want. Read my thoughts first. If you find them intriguing, read the main article. If you want to learn more, check out the related articles and books.
1. What weightlifting can teach us about learning and decision making

I normally steer completely clear of clickbait like "Management lessons from India's cricket team" or "The Barcelona school of leadership". These "lessons" are usually just platitudes. Cliched "words of wisdom", spouted so often they're useless.

But this article - Strength training is learning from tail events - is very different. Which is to be expected, given that it's written by Nassim Taleb.

He uses strength training and weight lifting to illustrate three insights:
  • Learning happens at the extremes. You build strength only by working your muscles to failure. So it is with other disciplines. You learn most at the edge of your abilities. That's what deliberate practice is.
     
  • In complex systems, don't look only at individual parts. Whether it's your body or a competitive market, there are too many moving parts and variables, with unpredictable effects. You can't look at individual muscles or factors in isolation - you need to take a systems approach.
     
  • The newer a theory is, the less robust it is. Nutrition fads come and go, but timeless wisdom stays true - Lift weights, and you'll grow strong. Similarly, in life and business: if you hear a new theory, don't trust it yet. See if it withstands the test of time. The longer an idea lasts, the longer you should expect it to last. [This is also called the Lindy Effect.]
Turns out, weightlifting (and Taleb) can teach us quite a lot!
2. Clinton, Ada, and the perils of blindly trusting your computer

A LOT of people were shocked by the US election results. Including Hillary Clinton. After all, her computer had told her she would win!

As this Washington Post article says, Clinton's campaign relied heavily on an algorithm named Ada. This complex algorithm mined tons of data to guide almost every critical decision her campaign made. 

So, given their total faith in Ada, Clinton's team was understandably bemused. 

Could they have used a more accurate system? Is a more accurate system even possible?

I don't think so. For three reasons:
  • Algorithms don't predict results with certainty. It may come as a surprise, but computers can't see the future (yet). They're merely predicting the probable outcome. Something predicted with 75% probability will be wrong 25 out of 100 times. 
     
  • A rule-based system can't always codify complex reality. As Taleb says above, reality is too complex to always be distilled to 10, 20 or even 100 factors. Every once in a while, there will be unforeseen implications.
     
  • Garbage in, Garbage Out. Or GIGO, as we used to say in Computers 101. Every algorithm is subject to the biases of its creator. Most of the establishment underestimated the power of the rural Rust Belt. Unsurprisingly, Ada did too. As Techcrunch says, machine learning algorithms are biased by the data they are fed

What's the learning for us? Before trusting your Facebook news feed, your Google search results, or your complicated financial model, remember: They may be telling you what you want to hear.

[Aside: The aftermath of the election is a great example of self-serving bias. Clinton aides blame loss on everything but themselves.]
Mental Model of the Week: The Map is not the Territory

What it is:
We use maps, principles, mental models, learnings from experience, etc. to help us navigate the world around us. But it's important to remind ourselves - the map is not the territory.
  • The map doesn't include every feature of the territory. Even a very detailed map of London won't include every thin street.
  • The territory is different: Sounds banal, but a map of London won't help at all, if you're in Mumbai.
  • The territory may have changed. An 1850 map of London won't help you in 2016 London.
This sounds trite, but we often forget this, as we see in the following examples.

Examples in business:
Rules to follow:
  1. Start from first principles. Always begin with, "What do we know to be absolutely true?"
     
  2. Beware of false rigor. Just because something is described concretely doesn't mean it is concrete. 

Further reading:
That's it for this week! Hope you liked the articles, and the mental model.

Since you got this far, here's a bonus: remember the theory that happiness doesn't increase with income? Turns out it's wrong. Money can buy happiness, after all. Sigh, there go my plans of retiring early.

See you next week!

PS. If you like the stuff I send you every week, I'd be honored if you could forward this to a friend so they can subscribe. Thanks a lot!
Sunday Reads is a hand-curated weekly newsletter, bringing you the most thought-provoking articles on business, strategy, entrepreneurship and everything in between. Think of it as a little brain exercise, to help us all become better at what we do.

If you enjoy what you read here, please consider forwarding it to spread the word.
Share
Tweet
Forward
Share

You are reading a newsletter I shared with my mailing list. If you like it (and I hope you do), you can get a newsletter like this every Sunday by signing up here.
Copyright © 2016 Jithamithra T, All rights reserved.


Want to change how you receive these emails?
You can update your preferences or unsubscribe <<Email Address>> from this list

Email Marketing Powered by Mailchimp