[jasim]

My feelings as well. But this one sentence made the book worth it - it is a commonplace thought that should've occurred to most others, but anyway was a clearer articulation than whatever I'd had:

"I'm suspicious of any plan to fix unfairness that starts with 'step one, dismantle the entire system and replace it with a better one,' especially if you can't do anything else until step one is done. Of all the ways that people kid themselves into doing nothing, that one is the most self-serving."

[donw]

Love that quote.

I think why so many people want to "dismantle the system" is exactly the same reason that less-experienced engineers push for rewrites: it seems easier to build a good system from scratch, than it does to build the skills required to refactor what you already have into what you need.

(Not to mention, the skills to understand what you actually need!)

Because, more often than not, it turns out that the old system had evolved to deal with a crazy number of edge-cases that were, in fact, really, really important.

Also, and connected: there seems to be this odd zeitgeist of "make other people solve things".

Like, most of us think homelessness is bad, and most of us want to help the homeless, but how few people actually directly invest their own money -- which I would argue is more important than donating time -- towards actually solving that problem?

If I lived in the US, I'd be half-tempted to do a YouTube series where all I do is run around, ask people what they think the biggest problem is in society, and then ask them how much money they've spent to help fix it. Might be interesting, especially if I could do so with a curated list of charities (e.g., ones that don't just pocket the money for nefarious purposes)...

[AnthonyMouse]

> Because, more often than not, it turns out that the old system had evolved to deal with a crazy number of edge-cases that were, in fact, really, really important.

The problem is that a lot of crufty old systems evolved to deal with a crazy number of edge cases that no longer exist.

Or they started out with a bad assumption, and then had some hacks applied to deal with the problems that caused, and then had some other hacks applied to deal with the problems those hacks caused, until nobody can see from one end to the other.

It's important to be able to tell the difference between something which is complicated because it's dealing with a complicated problem and something which is complicated because it's hot garbage.

> Like, most of us think homelessness is bad, and most of us want to help the homeless, but how few people actually directly invest their own money -- which I would argue is more important than donating time -- towards actually solving that problem?

The trouble with many of these issues is that they're results rather than causes.

Why is someone homeless? For one person it's mental illness, for another it's unemployment, for another it's drugs, for another it's housing costs.

So if you want to solve homelessness, all you have to do is solve mental health, unemployment, drug policy and zoning. And then six other problems that caused six other people to be homeless.

Which, it turns out, somebody needed to solve anyway, but now you've got to pick something to focus on. And I think that's where people have trouble.

https://en.wikipedia.org/wiki/Buridan%27s_ass

[donw]

> The problem is that a lot of crufty old systems evolved to deal with a crazy number of edge cases that no longer exist.

Totally true! But it's usually a lot easier to factor those cases out than to rebuild the entire thing from zero.

> It's important to be able to tell the difference between something which is complicated because it's dealing with a complicated problem and something which is complicated because it's hot garbage.

My rule-of-thumb here is "If you have to force people to use it over alternatives, you can safely throw it out. And possibly not even bother replacing it."

> So if you want to solve homelessness, all you have to do is solve mental health, unemployment, drug policy and zoning. And then six other problems that caused six other people to be homeless.

If you want to solve all homelessness, sure.

But there is something to be said for helping fix real problems, for real people, now. To make the world that tiny bit better. And in doing so, you gain a more nuanced view of the problem you are solving, which makes you better at proposing and evaluating proposed solutions.

I think too many people end up in this weird place where you have to solve either all of the problem, or none of the problem -- evolutionary change just doesn't seem to be on the table.