More on this book
Community
Kindle Notes & Highlights
Think of “feedback” as the information communicated in response to an action.
A critical requirement is learning how to filter feedback. Not all of it is useful. The more quickly you learn to identify good feedback and accept and incorporate it, the more progress you will make toward what you want to achieve.
There are two basic types of feedback loops: balancing and reinforcing, which are also called negative and positive. Balancing feedback loops tend toward an equilibrium, while reinforcing feedback loops amplify a particular process.
Complex societies require a fair amount of trust between members to function.
You start off trusting, and more importantly, you create a feedback loop that says you are capable of and willing to trust.
Too often we remove the learning process, including the inevitable failures and disappointments, in success stories.
Complex systems often have many feedback loops, and it can be hard to appreciate how adjusting to feedback in one part of the system will affect the rest. Using feedback loops as a mental model begins with noticing the feedback you give and respond to every day.
No one wants to be a bottleneck, which is easily conceptualized as that person who makes everyone else wait. We see this behavior in people who can’t delegate.
You cannot completely eliminate them because once you do, another part of the system will become the new limiting factor.
A bottleneck is something we can alleviate; a constraint is a fundamental limitation of the system.
Instead of addressing bottlenecks as they appear, your time might be better spent on a root-cause fix that makes a foundational improvement that leads many bottlenecks to disappear indefinitely.
When there are no severe bottlenecks, we may not think to improve things. After they emerge, we’re forced to be more creative and watch that we don’t transfer the pain somewhere worse. In the long run, they can make us better off.
The larger the company grows, the more work it takes to ensure information flows to the right places.
The average life span of an S&P 500 company is 24 years, and this number is decreasing over time.
Long-lived Japanese companies tend to be small. They’re owned and run by relatives and people with close relationships.
If you want to scale something up, you need to anticipate that new problems will keep arising—problems that didn’t exist at a smaller scale. Or you might need to keep solving the same problems in different ways.
An Astronaut’s Guide to Life on Earth,
why astronauts learn as much as they can. They don’t stop at what they need to know but continue lifelong learning to prepare for any eventuality they can think of. They reduce blind spots.
“Our core skill,” Hadfield writes, “the one that made us astronauts—the ability to parse and solve complex problems rapidly, with incomplete information in a hostile environment—was not something any of us had been born with. But by this point we all had it. We’d developed it on the job.”7
Our ego gets in the way of capitalizing on the margin of safety that is produced by knowing more than you need to.
“Early success is a terrible teacher. You’re essentially being rewarded for a lack of preparation, so when you find yourself in a situation where you must prepare, you can’t do it. You don’t know how.”10
Broad competence seems very costly compared to specialization, but it is more likely to save us in the outlier situations of life.
However, the rate of churn varies depending on factors like the availability of alternatives, ease of switching, and overall satisfaction.
When the purpose of an organization becomes preventing members from leaving, it turns into a cult.
People need the freedom to vote with their feet if things get too bad.
Replacing components of a system is an inevitable part of keeping a system healthy. New parts can improve functionality.
Increasingly, algorithms are designed to be directionally correct over perfect.
The Invisible Hook.7
By seeking profit above all else, pirates ended up designing a legal system that was far ahead of its time and arguably fairer than that of mainstream society in that era.
Articles covered the allocation of plunder (which was equal, aside from leaders receiving a bit more), bonuses for anyone who showed unusual bravery (to compensate the added risk involved), and what we could consider a prototype of disability benefits for anyone injured in battles.
We can also conceptualize them as if–then processes that are useful because they can help us ignore variables that don’t matter and focus on requirements.
The Death and Life of Great American Cities
Happy City,
Space alone does not create this environment. An overly large square can actively repel people as it feels overexposed and empty. A popular one must have enough activity and reasons to visit in a small enough space to create the density needed for spontaneous and ongoing interactions to occur.
Brasília, the federal capital of Brazil, was designed in the 1950s and ’60s to replace Rio de Janeiro as the capital. Architect Oscar Niemeyer and urban planner Lucio Costa crafted a vision of a utopian city from scratch. In the visual sense, Brasília is a stunning World Heritage site. From the air, it has a beautiful birdlike form. As a place to live, it functions less well.
As Aristotle put it thousands of years ago, “The whole is something over and above its parts, and not just the sum of them all.”
Weak emergence occurs in systems in which functions are based on identifiable rules. We can model weak emergence by identifying the underlying rules. Strong emergence does not have identifiable rules behind it, so we cannot model it.
Oppressive regimes thrive when people are too scared to be seen opposing them. Visible opposition inspires more people to ask questions and to join in fighting oppression. That’s why the government at the time went to such lengths to prevent dissent.
The Secret of Our Success,
People specialize because no one can know everything. Then they interact. And in that system in which the interaction occurs, something happens that otherwise wouldn’t.
In communications, to get to the essence of the thing is important because simple communications are easier to understand.
Attempting to build a complex system from scratch tends to be ineffective. It takes consistent, incremental progress from something basic that works.
Understanding the irreducible components of a system means you won’t waste your time trying to change what is unchangeable. You can master the minimum elements, then explore.
If you want people to pay attention, you need to keep raising the bar.
Plan for diminishing returns, and you just might avoid them.
This normal distribution tends to be the case for values that are subject to certain physical constraints, such as biological measurements.
But if we get too caught up in seeing the world as normally distributed, we can forget that long-tail values tend to have an outsized impact.
Normal distributions can be contrasted with power law distributions. The values in a power law distribution cluster at low or high values.
Wealth follows a typical power law distribution.
It’s important to consider how the consequences of our choices can multiply over time.
