« Eliezer's Meta-Level Determinism | Main | Should Bad Boys Win? »

June 24, 2008


Given that we're heading towards morality and the Singularity, I imagine this is here as a dependency for a future post arguing that there is a more-or-less objective notion of morality/humaneness/Friendliness for humans. However, even though all humans are virtually alike compared to the vastness of possible mind design space, there is still, from a human perspective, an enormous amount of variation amongst people. The things that everyone has in common are ipso facto irrelevant: it would border on contradiction to tell a human that her moral stance is wrong because it is contrary to human nature, for if it really were, a human couldn't hold it.

Something to address in a future post, perhaps.

Is this true of ideas/memes also? Because there are a lot of wacky groups out there, more than just froth on the surface.

Z. M. Davis, that's not what this is a dependency for.

Ian, this principle does not hold for human artifacts and human memes; we can make large jumps through design space with many coordinated simultaneous changes.

Ah yes, you covered that in _Optimization and the Singularity_.


hm. there's a lot to this...but let's just say that when it comes to *psychology* the residual of non-universality seems really important to focus on in the area of individual differences. specifically, it seems entirely possible that frequency dependent personality morphs abound; e.g., all the stuff about drd4 that i've been posting on my weblog. of course, this is single-locus, but i think it's just the locus of biggest effect. i suspect that what evolutionary psychology misses is that after you account for the substrate of human universals you've got personality morphs which are playing "games" with each other, and in particular there are a host of low frequency morphs running around. these morphs are "complex" to my way of thinking, though perhaps not complex in the way you're implying (i know the evo psych argument against a lot of variation on traits). i happen to think epistasis might also be pretty significant in the transient.

second, i'm also getting curious about variation in traits we perceive as on-off, where those who are "off" are purely pathological. e.g., it turns out that 2% of the population might be "face blind" but have been cryptic because they develop techniques to mask this problem and don't talk about it. i don't think it is just a 2% vs. 98%, i think there are a few other steps "in between", though there's a skewness toward the "normal" facial recognition ability side. on the basis of this i'm willing to dig a little deeper into "human universals" to see what might crop up on the margins.

You can have real X-Men, check out a discovery special about "real superhumans". There was one guy who could withstand cold so well that the doctors thought it shouldn't be possible. A single mutation sometimes does create significant changes (and in this case advantages).


razib, great post. Overall, if there's a particular weakness among enthusiasts (and "experts") in rational thinking, it's a tendency towards overreductionism. Way to often the discussion here goes from Most X are Y to let's discuss why X is Y. It can be revealing to analyze the subset of X that aren't Y, as well as the overall distribution of X according to how much they are or aren't Y. But these other analyses are done or discussed much here at OB, or at most similar sites.

In general, I definitely agree with this post. Though I put little trust in Donald E Brown, my concerns aren't relevant here. Slightly relevant is the fact that it seems conceivable that the complex functional adaptations of humankind are so numerous and so complex that we exist at an equilibrium where more critical adaptations appear in almost everyone but significant adaptations exist which are less critical. Genes necessary to those adaptations might break regularly and not be weeded out by selection quickly. If such adaptations are numerous, most people might be missing several. Other adaptations might require, as vision does, environmental conditions that are reliably present in the ancestral environment but which may not be so reliably present today in order to emerge.


1) A situation with AIs whose intelligence is between village idiot and Einstein -- assuming there is a scale to make "between" a non-poetic concept -- is not very likely and probably short-lived if it does occur (unless perhaps it is engineered that way on purpose).

2) Aspects of human cognition -- our particular emotions, our language forms, perhaps even pervasive mental tricks like reasoning by analogy -- may be irrelevant to Optimization Processes in general, making their focus for AI research possibly "voodoo doll" methodology. AI may only deal with such things as part of communicating with humans, though mastering them well enough to participate effectively in human culture may be as difficult as inventing new technologies.

3) Optimization Processes built by Intelligent Designers can develop in ways that those built by evolution cannot because of multiple coordinated changes (this point has been beaten to death by now I think).

4) Sex is interesting.

For once, I have no complaints. I assume the path is being cleared for a discussion of what actually IS required for an optimization process to do what we need it to do (model the world, improve itself, etc), which seems only marginally related to what our brains do. If that's where this is headed, I'm looking forward to it.

I agree with the basic points about humans. But if we agree that intelligence is basically a guided search algorithm through design-space, then the interesting part is what guides the algorithm. And it seems like at least some of our emotions are an intrinsic part of this process, e.g. perception of beauty, laziness, patience or lack thereof, etc. In fact, I think that many of the biases discussed on this site are not really bugs but features that ordinarily work so well for the task that we don't notice them unless they give the wrong result (just like optical illusions). In short, I think any guided optimization process will resemble human intelligence in some ways (don't know which ones), for reasons that I explained in my response to the last post.

Which actually makes me think of something interesting: possibly, there is no optimal guided search strategy. The reason why humans appear to succeed at it is because there are many of us thinking about the same thing at any given time, and each of us has a slightly differently tuned algorithm. So, one of is likely to end up converging on the solution even though nobody has an algorithm that can find every solution. And people self-select for types of problems that they're good at.

The theme of this post isn't very accurate, as large phenotypic polymorphisms in various other species demonstrate - e.g.:

From a strategic point of view we are especially interested in those species that have two kinds of males. It's almost like having a third sex. In fact the winged males look far more like females than they look like wingless males. Both females and winged males are almost believable as wasps, although they are tiny. But the wingless males are nothing like wasps to look at. Many have savage pincer jaws which make them look a bit like miniature earwigs going backwards. They seem to use these jaws only for fighting — lacerating and slicing to death other males that they encounter as they stalk the length and breadth of the dark, moist, silent garden that is their only world. - Climbing Mount Improbable, Richard Dawkins.

Where did the reasoning go off the rails?

It isn't just the case of males and females where different alleles can form a truce. There's the whole phenomenon of frequent-dependent selection. Most people are familiar with this from blood types, and sickle-cell anaemia. Alleles with phenotypic effects involving disease resistance can be advantageous when rare and disadvantageous when common - resulting in them never going near extinction or fixation.

Also, this premise is inaccurate:

If gene B depends on gene A to produce its effect, then gene A has to become nearly universal in the gene pool before there's a substantial selection pressure in favor of gene B.

Gene B can spread if gene A is present at a frequency of 20% in the population - provided it is not deleterious in the absence of gene A. Sure, then the selection pressure maintaining it is reduced by a factor of five, but that's not necessarily enough to kill it off.

Finally, phenotypic variation does not necessarily depend on genetic variation. There's also the influence of the environment to consider. In general, the environment is quite capable of sending some organisms down different developmental paths depending on the circumstances in which they find themselves. This is known as phenotypic plasticity.

There are plenty of examples of phenotypic plasticity in humans - e.g. the effect is an important part of the reason why a Sumo wrestler and a racing jockey have different phenotypes.

I argue that the idea that complex adaptations are necessarily universal is fundamentally not correct - in my "Species Unity" essay:


The comments to this entry are closed.

Less Wrong (sister site)

May 2009

Sun Mon Tue Wed Thu Fri Sat
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30