siIver comments on The Proper Use of Humility - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (42)
I actually don't quite agree (this is the first time I found something new to criticize on one of the sequence posts).
To me, it seems like humility as discussed here is inherently a distortion, that when applied, shifts a conclusion in some way. The reason why it can be a good thing is simply that, if a conclusion is flawed, it can shift it into a better place, sort of a counter-measure to existing biases. it is as if I do a bunch of physical measurements and realize that the value I observe is usually a bit too small, so I just add a certain value to my number every time, hoping to move it closer to the correct one.
However, once I fix my measurement tools, that distortion then becomes negative. Similarly, once I actually get my rationality correct, humility will become negative. In this case, there also seems to be a general tool to get your conclusion fixed, which is to use the outside view rather than the inside view. Applying that to the engineer example:
If the engineer used the outside view, he should know that humans are fallible and already conclude that he should spend an appropriate amount of time on fail-safe mechanics. If he then applied humility on top of it, thus downplaying his efforts despite having used the outside-view, it should lead him to worry/work on it more than necessary.
Of course, you could reason that in my example, applying the outside view is itself a form of applying humility. My point is simply that even proper humility doesn't seem to cover any new ground. It's not "part of rationality," so to speak. It's simply a useful tool, practically speaking, to apply when you haven't conquered your biases yet. In that sense, I would argue that, ultimately, the correct way to use humility is not at all / automatically without doing anything.