I don’t recommend this post for the Best-of-2018 Review.
It’s an exploration of a fascinating idea, but it’skind of messy and unusually difficult to understand (in the later sections). Moreover, the author isn’t even sure whether it’s a good concept or one that will be abused, and in addition worries about it becoming a popularized/bastardized concept in a wider circle. (Compare what happened to “virtue signaling”.)
one that will be abused, and in addition worries about it becoming a popularized/bastardized concept in a wider circle. (Compare what happened to “virtue signaling”.)
This is a terrible rationale! Our charter is to advance the art of human rationality—to discover the correct concepts for understanding reality. I just don’t think you can optimize for “not abusable/bastardizable if marketed in the wrong way to the wrong people” without compromising on correctness.
And likewise “virtue signaling.” Signaling is a really important topic in economics and evolution and game theory more generally. If we were doing a Best-of-2014 review and someone had written a good post titled “Virtue Signaling”, I would want that post to be judged for its contribution to our collective understanding, not on whatever misuse or confusion someone, somewhere might subsequently have attached to the same two-word phrase. (Indeed, I myself have “Virtue Signaling Is Costly and Honest” as an entry in my long list of not-yet-written blog post ideas.)
Some say: if you don’t pre-emptively take into account the ways in which your work might be misused, you’re shooting yourself in the foot. I say this is a decision theory problem and that the “shooting oneself in the foot” metaphor is choosing the wrong counterfactual. Rather, we should think of the people bastardizing and abusing concepts as the ones shooting us in the feet! Rather than passively accepting the inevitability of the cultural forcesmaking language less precise, we need to be figuring out how to fight back: how to challenge misrepresentations (“Well, actually—unironically well-actually—‘signaling’ doesn’t mean that”), rather than allowing ourselves to be cowed into not saying anything that could be misrepresented.
(Oh, separately, I agree that this post should not be included in the Best-of-2018 collection. I had drafted a long reply post (working title: “Firming Up Not-Lying Around Its Edge-Cases Is Surprisingly Useless”), but pre-readers pointed out that my reply had its own serious problems; I’ll probably be able to edit it down into a brief review that I’m comfortable with, but it’s not my highest priority writing task right now.)
say this is a decision theory problem and that the “shooting oneself in the foot” metaphor is choosing the wrong counterfactual. Rather, we should think of the people bastardizing and abusing concepts as the ones shooting us in the feet!
You still end up with a shot foot. People tend to confuse solving problems with apportioning blame—I call it the “guns don’t kill people” fallacy.
I don’t recommend this post for the Best-of-2018 Review.
It’s an exploration of a fascinating idea, but it’skind of messy and unusually difficult to understand (in the later sections). Moreover, the author isn’t even sure whether it’s a good concept or one that will be abused, and in addition worries about it becoming a popularized/bastardized concept in a wider circle. (Compare what happened to “virtue signaling”.)
This is a terrible rationale! Our charter is to advance the art of human rationality—to discover the correct concepts for understanding reality. I just don’t think you can optimize for “not abusable/bastardizable if marketed in the wrong way to the wrong people” without compromising on correctness.
Concepts like “the intelligence explosion” or “acausal negotiation” are absolutely rife for abuse (as we have seen), but we don’t, and shouldn’t, let that have any impact on our work understanding AI takeoff scenarios or how to write computer programs that reason about each other’s source code.
And likewise “virtue signaling.” Signaling is a really important topic in economics and evolution and game theory more generally. If we were doing a Best-of-2014 review and someone had written a good post titled “Virtue Signaling”, I would want that post to be judged for its contribution to our collective understanding, not on whatever misuse or confusion someone, somewhere might subsequently have attached to the same two-word phrase. (Indeed, I myself have “Virtue Signaling Is Costly and Honest” as an entry in my long list of not-yet-written blog post ideas.)
Some say: if you don’t pre-emptively take into account the ways in which your work might be misused, you’re shooting yourself in the foot. I say this is a decision theory problem and that the “shooting oneself in the foot” metaphor is choosing the wrong counterfactual. Rather, we should think of the people bastardizing and abusing concepts as the ones shooting us in the feet! Rather than passively accepting the inevitability of the cultural forces making language less precise, we need to be figuring out how to fight back: how to challenge misrepresentations (“Well, actually—unironically well-actually—‘signaling’ doesn’t mean that”), rather than allowing ourselves to be cowed into not saying anything that could be misrepresented.
(Oh, separately, I agree that this post should not be included in the Best-of-2018 collection. I had drafted a long reply post (working title: “Firming Up Not-Lying Around Its Edge-Cases Is Surprisingly Useless”), but pre-readers pointed out that my reply had its own serious problems; I’ll probably be able to edit it down into a brief review that I’m comfortable with, but it’s not my highest priority writing task right now.)
You still end up with a shot foot. People tend to confuse solving problems with apportioning blame—I call it the “guns don’t kill people” fallacy.