> This is truly one of the gaping holes in human cognition, this idea that a single participant in a larger system can make a change and then model the results of that change without accounting for the reactions of all the other actors
The parent's sentiment may be a tired old theme in the context of a game company that produces a long-awaited sequel to a much-beloved classic, meddles too much with the winning formula, and ends up killing the goose that lays the golden eggs, but I feel that this idea needs more airtime in politics.
The best, surest way to get a complex system that works is to begin with a simple system that works and make a series of many incremental changes.
This should be a fundamental axiom in politics -- that anything you do, from creating a top-secret surveillance agency, to sweeping healthcare reform, to the war on drugs, to the legalization of gay marriage, is not guaranteed to have the effect its creators intended, at the magnitude they intended, without any other potentially larger effects, some of which may range from undesirable to catastrophic.
Therefore, when possible small and simple changes should be made, and we should think twice before proposing complex changes, and ponder three times again before implementing them. It's the only possible way to avoid getting buried under a relentless onslaught of unexpected consequences.
Not enough people are even talking about this idea. AFAICT it's neither conservative nor liberal nor libertarian; it's literally conservative in the sense of "we don't understand the problem space very well outside the status quo, so let's take care not to get too far away from the familiar in case we fall into a hole we can't climb back out of." Imagine a decades-old project which gets dozens to hundreds of commits a year, but whose authors always reject any commits smaller than 1000 lines of code. What's surprising isn't that it contains bugs, what's surprising is that the end result does anything resembling what it's supposed to.
> The best, surest way to get a complex system that works is to begin with a simple system that works and make a series of many incremental changes.
Shades of Gall's Law from Systemantics:
A complex system that works is invariably found to have evolved from a simple system that worked. A complex system designed from scratch never works and cannot be patched up to make it work. You have to start over with a working simple system. – John Gall (1975, p.71)
The parent's sentiment may be a tired old theme in the context of a game company that produces a long-awaited sequel to a much-beloved classic, meddles too much with the winning formula, and ends up killing the goose that lays the golden eggs, but I feel that this idea needs more airtime in politics.
The best, surest way to get a complex system that works is to begin with a simple system that works and make a series of many incremental changes.
This should be a fundamental axiom in politics -- that anything you do, from creating a top-secret surveillance agency, to sweeping healthcare reform, to the war on drugs, to the legalization of gay marriage, is not guaranteed to have the effect its creators intended, at the magnitude they intended, without any other potentially larger effects, some of which may range from undesirable to catastrophic.
Therefore, when possible small and simple changes should be made, and we should think twice before proposing complex changes, and ponder three times again before implementing them. It's the only possible way to avoid getting buried under a relentless onslaught of unexpected consequences.
Not enough people are even talking about this idea. AFAICT it's neither conservative nor liberal nor libertarian; it's literally conservative in the sense of "we don't understand the problem space very well outside the status quo, so let's take care not to get too far away from the familiar in case we fall into a hole we can't climb back out of." Imagine a decades-old project which gets dozens to hundreds of commits a year, but whose authors always reject any commits smaller than 1000 lines of code. What's surprising isn't that it contains bugs, what's surprising is that the end result does anything resembling what it's supposed to.