= 27d567bc2d48d9f40e9ccc20ea370b15)
JGWeissman comments on Procedural Knowledge Gaps - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (1477)
A big red flag is that where I expected to see an analysis of what mixture of the various ancestral types a person is, I see a declaration that the person fits into one of 6 little boxes. This despite the fact these types describe multiple features, controlled by different genes, in a sexually reproducing population that represents all types. Now, not all genes are selected for independantly, genes nearby on the same chromosone can be correletated. But a model of which dietary traits correspond to which easily measurable traits should be more complicated than assigning a cluster of dietary traits to a cluster of easily measured traits.
FYI, my wife's actual profile showed this; more precisely, IIRC it rated her as 44% Explorer, based on the traits given. It did not show what percentages the other 5 boxes broke down to, and I don't know whether those factors were also taken into account in the analysis. (I also don't know what precisely the percentage represents; i.e. is it a probability, a "percentage of your traits"...?)
The sample profile I linked appears to date from 2008; so perhaps the percentage report was added to the software later. But in both cases, if I understand correctly, the software simply presents the highest-scoring of the six boxes, rather than saying, "this is you".
Still, compared to most ways of nutritionally grouping people, six is actually a LOT of boxes.
From what I read in his book, he describes the types in terms of basic strategies for responding to the environment, where there are only a few good choices to make. IOW, the stereotypes are supposed to represent stable strategies for responding to infections, shortages, and other stressors. That is, there are not an unlimited number of ways to do things in those areas, so you end up with large clusters.
I have not studied this in any detail, mind you; I confess my primary interest in the book was more to look at the food lists for my type, to compare against my personal dietary history.
As I said, I'm less interested in the plausibility or sensibility of a theory per se, than with the correlation of its advice with the obtaining of results... especially results for myself in particular. (And I remain cautiously optimistic on that front where his advice is concerned.)
Better, but still not good enough. If it is a mixture, what about the other 56%? Which 44% of the explorer traits? If it is a confidence level, the model doesn't seem to rate itself very highly, so why should I be impressed with it?
Suppose there are 2 stable strategies that each say how to deal with several specific problems, such that either of the 2 strategies work but a mixture somehow fails. These strategies, being complicated, are coded for by multiple genes. Suppose a man with one strategy and a women with the other strategy mate and have offspring. Those offspring are going to inherit some mixture of the two strategies, even discounting complications such as being heterzygous where the parents are homozygous, and therefore will employ an unfavorable mixture. You can not have multiple non mixable complicated traits in a sexually reproducing population, without tricks like having each member have the complete code for all possible traits, and a varying gene that switches on one of them, which we observe in sexual dimorphism at not much otherwise. Eliezer has written of this.
But we do observe epigenetic traits - variations in gene expression based on environmental conditions, such as genes that act differently depending on how much exercise you get, or the level of testosterone in the fetal environment, or various other things.. D'Adamo's claim here is that his typing groups are a combination of gene inheritance and gene expression, and his notion of "strategies" isn't really the same as say, having a completely different way of digesting foods.
It's more like identifying which places to store fat in first -- something that (IIUC) we already know is heritable. The fact that fat gets stored isn't changed, just how much, where, and how quickly. Something like that can make a big difference on a practical level to a person's life, without being a particularly complex adaptation in itself.