Some highlights / takeaways / thoughts / comments from #RWRI 14, Day 3:
1. Biggest takeaway overall through three days is: "Science isn't about naive evidence, it's about understanding properties."
1. Biggest takeaway overall through three days is: "Science isn't about naive evidence, it's about understanding properties."
It's pretty easy to run a linear regression on two variables, or develop an algorithm, and call it science. It's very difficult (and sometimes impossible) to deeply understand the properties of the parts of a system and how they interact with each other.
2. Deterministic systems can produce unpredictable results. In complex systems, a slight tweak to the probability can yield completely different results. Or, sometimes, the end result will be the same no matter how large you tweak the probability.
The point being, again, that you can't figure out the end result of a system in Extremistan (i.e. most of life) with only a general understanding. You have to deeply understand the properties, interactions, etc. in order to truly understand a system.
Take the stock market, or Bitcoin price, for example. At any given time, the price comes from an incalculably complex web of actions, interactions, randomness, etc.
Yet, you'll see people / news organizations give a single reason for why the price moved up or down, which is obviously a load of BS.
3. "Nothing top-down has ever survived." In a top-down structure, there is a concentration of power at the top. It can work for a while, depending on the people in charge. But eventually, someone will royally screw up, and it'll all come crashing down.
This isn't to say that nothing should ever be centralized. Rather, systems should be built to where even if the people 'in charge' screw up, the system will still survive.
If a system has some form of centralization, the system's survival should never be dependent on the competence (or survival) of its centralized group.
A simple heuristic for how to build a system for survival:
"Never do something one level higher than you need to."
Put a different way: The only things that should be performed by any given level of a system / hierarchy are the things that cannot be done at the level below it.
"Never do something one level higher than you need to."
Put a different way: The only things that should be performed by any given level of a system / hierarchy are the things that cannot be done at the level below it.
