Update readme.md

This commit is contained in:
Nuño Sempere 2019-05-28 11:59:53 +02:00 committed by GitHub
parent 4b32f3c013
commit 198ba2e2b1
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23

View File

@ -629,7 +629,7 @@ The point being that there are a lot of mental health ressources and information
In personal converations, a person in the outer orbit of the EA community has pointed out to me that providing mental health is creepy, and that they feel cringe when thinking about the idea. The word "cult" is mentioned. I am uncertain about their epistemic level, but it is not implausible that providing mental health may put prospective EAs off.
[1] Technical note: Let a be a variable which stands for an individual eas, and consider a mapping of O: A-> |N, such that O(a) falls in {1,...,n}, and consider a function like f(x) = c\*x^(-j)\*(1 + 1/sqrt(2\*pi\*9)\*exp(-x^2 / 2\*9}\*sin(x)/BB(6)), where BB is the busy beaver function. It may be that the counterfactual impact of eas follows such a distribution; j and c would be arbitrary constants, with j preferably greater than 3, because otherwise the variance is not well defined, and consider the relationship which the integral from 1 to k of f(x)dx and the integral from k+1 to n of f(x) dx have. It wouldn't be unsurprising if O(a) were not inversely correlated with conscientiousnes and initiative, and correlated, perhaps causally, with more mental health problems, as these variables often are. Now consider the first k such that the integral from 1 to k of f(x)dx > the integral from k+1 to n of f(x) dx. The question is now whether for high O(a), offering mental health is worth it, given that O(a) is a priori unknown, and that computing the exact value of f(O(a)) is arduous / subject to Goodhart's law or to moral hazards.
[1] [Technical note](https://concepts.effectivealtruism.org/concepts/information-hazards/): Let a be a variable which stands for an individual eas, and consider a mapping of O: A-> |N, such that O(a) falls in {1,...,n}, and consider a function like f(x) = c\*x^(-j)\*(1 + 1/sqrt(2\*pi\*9)\*exp(-x^2 / 2\*9}\*sin(x)/BB(6)), where BB is the busy beaver function. It may be that the counterfactual impact of eas follows such a distribution; j and c would be arbitrary constants, with j preferably greater than 3, because otherwise the variance is not well defined, and consider the relationship which the integral from 1 to k of f(x)dx and the integral from k+1 to n of f(x) dx have. It wouldn't be unsurprising if O(a) were not inversely correlated with conscientiousnes and initiative, and correlated, perhaps causally, with more mental health problems, as these variables often are. Now consider the first k such that the integral from 1 to k of f(x)dx > the integral from k+1 to n of f(x) dx. The question is now whether for high O(a), offering mental health is worth it, given that O(a) is a priori unknown, and that computing the exact value of f(O(a)) is arduous / subject to Goodhart's law or to moral hazards.
## Inf. Survey questions
1. How involved are you in the Effective Altruism Community?