I partly had in mind personal communications, but some public examples (and very brief summaries of my reactions, not fleshed out counterarguments):
In âSequence thinking vs. cluster thinkingâ, Holden says, âFor example, obeying common-sense morality (âends donât justify the meansâ) heuristics seems often to lead to unexpected good outcomes, and contradicting such morality seems often to lead to unexpected bad outcomes.â
I guess the argument is supposed to be that we have empirical evidence of heuristics working well in this sense. But on its face, this just pushes the question back to why we should expect âhow well a strategy works under unknown unknownsâ to generalize so cleanly from local scales to longtermist scales. (Related discussion here.)
âHeuristics for clueless agentsâ claims that âheuristics produce effective decisions without demanding too much of ordinary decision-makers.â
Their arguments seem to be some combination of âin some decision situations, itâs pretheoretically clear which decision procedures are more or less âeffectiveââ (Sec. 5) and âheuristics have theoretical justification based on the bias-variance tradeoffâ (Sec. 7). But pretheoretic judgments about effectiveness from a longtermist perspective seem extremely unreliable, and appeals to bias-variance tradeoffs are irrelevant when the problem (under UUs) is model misspecification.
Appreciate the examples, especially the Holden essay which was the main reason I started doing more cluster reasoning to form decision-oriented views. And thanks for the pointer to your writeup, youâve given me food for thought.
I partly had in mind personal communications, but some public examples (and very brief summaries of my reactions, not fleshed out counterarguments):
In âSequence thinking vs. cluster thinkingâ, Holden says, âFor example, obeying common-sense morality (âends donât justify the meansâ) heuristics seems often to lead to unexpected good outcomes, and contradicting such morality seems often to lead to unexpected bad outcomes.â
I guess the argument is supposed to be that we have empirical evidence of heuristics working well in this sense. But on its face, this just pushes the question back to why we should expect âhow well a strategy works under unknown unknownsâ to generalize so cleanly from local scales to longtermist scales. (Related discussion here.)
âHeuristics for clueless agentsâ claims that âheuristics produce effective decisions without demanding too much of ordinary decision-makers.â
Their arguments seem to be some combination of âin some decision situations, itâs pretheoretically clear which decision procedures are more or less âeffectiveââ (Sec. 5) and âheuristics have theoretical justification based on the bias-variance tradeoffâ (Sec. 7). But pretheoretic judgments about effectiveness from a longtermist perspective seem extremely unreliable, and appeals to bias-variance tradeoffs are irrelevant when the problem (under UUs) is model misspecification.
Appreciate the examples, especially the Holden essay which was the main reason I started doing more cluster reasoning to form decision-oriented views. And thanks for the pointer to your writeup, youâve given me food for thought.