I got some time away from the kids and family. What do did I do with that time you ask? I selected a new font for the blog. Haha. That's funny you say, well F-ing yea it is! I love it.
Okay.... Let's start.
I will be starting my new job at Google / GCP / STS on Aug 12 and I am pretty excited to see under the hood. I spent the afternoon eating tacos and watching micrograd/makemore. Boy those math guys really worked something out for once. who knew.
Okay.... Now really start what we were here for....
Thinking in systems
The main take away that I have from this book is that we need to think not just in a reductive manner but also about the system as a whole. There are emergent behaviors of the systems that are only possible when we look at the interactions of the system instead of looking at the individual components.
To be honest, this playlist does a good job of summarizing the book as a whole: here.
Key terms and points
1. Stock: The reserve of the thing that we are accumulating. It is the thing that the system accumulates over time.
2. Inflow: Material / Information that enters the system leading to the increase in the stock if left to its own devices.
3. Outflow: Material/Information that leaves the system leading to the decrease in the stock.
The whole system can be imagined to be composed of a lot of feedback loops, these can be reinforcing feedback loop (that would increase the inflow over time) or a balancing feedback loop that would aim to stabilize or reduce the stock over time.
A system may have 1 or more stocks, 1 or more feedback loops. The result of altering the state of the feedback loops may result in increase, decrease or oscillations in the stock.
The result of the changing the feedback loops will also not be evident immediately, it may take some time for the effects to be observed. This is one of the main reasons for oscillations in the stock.
We should also think about the limiting factors. All systems cannot be scaled up indefinitely, they will all run into some limiting factors - human, technological or physical. Therefore unlimited growth is not possible and we need to be aware of the renewable as well as non-renewable resource (aspects of the stock).
Non-renewable resources = stock limited. taking out too much stock will cause exhaustion (fish in a pond)
Renewable resources = flow limited. - outflow is more than inflow would cause exhaustion
Why do systems work so well? This is primarily due to them having resilience. We want a rich structure of feedback loops at different levels so that they can compensate for different kinds of changes/breakdowns. This is a kind of systemic redundancy that allows things to evolve and develop "anti-fragility" / meta-resilience.
We want dynamic system resilience and not just static stability - a system that can only deal with resilience in a small problem space. In static stability, a common breakdown is information distortion - cancer can be thought of as a breakdown of the immune system's ability to respond to signals about bad cells.
Self-organization is also desirable as it's the capacity of a system to make it's own structure more complex in response to it's needs. For organizations, this means that it's people can organize themselves and use their imagination to solve novel problems.
Hierarchy works! It actively reduces the amount of information that needs to be exchanged at higher levels there by allowing for more complex structures to emerge. At the same time it encourages/enables lower levels to have denser connections.
End reading Chapter 3
No comments:
Post a Comment