From Twitter:
PSA: I'm not agreeing to dismantle sh*t until you can tell me:
1) What we're building instead
2) By what means
3) On what time-frame
4) How it will function, &
5) How it will *better* guard against misanthropes' impulses to exploit, oppress, & divide masses for self-serving ends
The implicit assumption here is that humans are trying to make bad things happen, and destroying a demon by default empowers an even stronger, nastier demon (I think this is at the root of my (and Robin Hanson's) disagreement with Zvi on blackmail.) This sort of conservative bias fails to get the right answer to questions like "should I sabotage the train tracks to Auschwitz?". (That was a live question for the US during WWII, and Roosevelt got the wrong answer.)
Very often, the correct answer to "what will we replace it with?" is "nothing." It's often hard to dismantle stuff, and we often sublimate violence instead of deescalating, but that's not really the same problem.
Related: Raise the Crime Rate, Reflections from the Halfway Point
Fair enough, but how is Auschwitz an example? In terms of the tweet, Roosevelt was already destroying. He was happy to replace many things with rubble in the short term. He was happy to replace all the train tracks with nothing, but he had limited capacity to destroy train tracks.
Roosevelt was fighting to win, rather than trying to save people. That's an error, but is it conservative? Maybe, but it seems pretty subtle to me.
Are you saying there is an outside view in support of the idea that dismantling institutions/infrastructure *in general* reduces oppression and improves humanity's average/aggregate well-being? If so, I'd be interested in hearing more examples in support of that POV. I agree re: the example of "train tracks, and infrastructure in general, within Nazi Germany."
The prior should be, with respect to institutions that are, if anything, optimized for destruction, that there shouldn't be the same kind of conservative "Chesterton's fence" presumption against action to destroy them, regardless of the apparent beneficial side effects. The argument should be on the level of "is this institution at its core trying to hurt or trying to help?", not directly on the level of "can I directly make a CDT-style utilitarian case for this particular action with respect to this institution, considered in isolation?". To do otherwise rewards extortionary tactics such as hostage taking, and the full decision-theoretic utilitarian calculus is uncomputable.
It seems to me like there's a pretty obvious, strong case for "the police force, as an institution, is making war on a substantial subset of the population," i.e. "trying to hurt." The presumption against abolition is an endorsement or at least acceptance of that war, and should be responded to accordingly. Sometimes we should make war on someone or fail to interfere with a war between two parties, but then we're not living in a society with those parties, they're just foreign powers and our relations with them need to be modeled as such.