[ExI] the formerly rich and their larvae, was: Impressive book: Farewell to Alms

Rafal Smigrodzki rafal.smigrodzki at gmail.com
Fri Feb 8 18:08:17 UTC 2008

On Feb 8, 2008 2:32 AM, Emlyn <emlynoregan at gmail.com> wrote:

> responsible for that decision. But I bet there is no significant
> correlation between birth control choices and genetic makeup. The
> result of that? No matter how many generations people make birth
> control choices over, genes will be entirely unaffected.
### I strongly suspect this is a losing bet here. The variance of the
vast majority, if not all, of human behaviors is at least in part
explained by genetic influences. Most trivially, low IQ individuals
(mostly a genetically distinct group) do make different birth control
choices than high IQ ones. There are genetic differences in
religiosity, which implies that after exposure to religious
indoctrination regarding birth control humans of the same IQ will
still differ in their decisions. The strength of paternal and maternal
instincts is most definitely influenced by genetic factors, so again
genes could influence birth control by modulating the degree of
reinforcement received after the first child.


> But... if you know of any genes that have been found for not passing
> on one's genes, let me know.
### There is an enormous catalog of genetic defects with profound
impact on fertility.

>There is a generally accepted idea that, slight tweaks
> not withstanding, we evolved for the pleistocene. That ended 10,000
> years ago.

### This is no longer believed to be the case. Now there is sufficient
evidence to believe that human evolution actually accelerated over the
last 5,000 years, by one estimate by a factor of 100 compared to
prehistoric times.


 Perhaps proposed genetic effects of reproductive choices
> would be quicker to rear their heads than this, but you're still
> dealing with many generations, at least maybe a couple of hundred
> years?

### Not really. If a gene under new environmental conditions (e.g.
availability of contraceptive pills and vasectomy) leads to a fitness
of 0 in both men and women, then in about 50 years it will be
eliminated from the gene pool, no matter how high the initial
frequency. If there is a difference in fitness between two groups of
only a factor of 2, then with every 25 years or so the relative sizes
of the populations will change by a factor of 2.

> Now it's true that a revulsion to the form of these contraceptives
> (let's say the pill) could be selected for, in that the meme for
> reproductive control might be more strongly resisted by human
> phenotypes with that revulsion already. But well before that began to
> take hold, you'd see the companies that make the pill changing the
> form to no longer match the nascent yuck instinct (or being replaced
> by companies that changed it). Then, that instinct, minor as it must
> be, would no longer be supported by selection pressure (which would
> now start its slow push at the new form(s)), and would disperse.

### Only if the revulsion was to the form of the pill, and not to the
effect. The reason for the commercial success of contraception is not
marketing but the fact that it fills a need. Most humans like to have
sex but many don't like children, the work, responsibility, and loss
of freedom associated with having them. Contraception lets them get
what they want but this reduces their fitness. What if some men and
women evolved who just love to have children - not sex, children?
These would not be interested in contraception.


> Memetic selection is faaaaster than genetic selection by orders of
> magnitude. The only new selection such as you describe will ever
> influence us again is if civilisation falls (because civilisation can
> be viewed as being only about memes slugging it out ever more
> efficiently).

### Here we come to the interesting part. In the not-too-distant
future the distinction between genes and memes will disappear.
Uploading and/or the AI singularity will result in a population
explosion of entities that will be capable of rewriting their own
minds as completely as if new genes were used to remake a brain.
Trillions of minds will be generated and (self-) deleted. I find it
fascinating to speculate about the outcome. I have previously written
here about the possible impact on the average per mind wealth - I
disagreed with Anders who defended the notion that future humans will
be much wealthier per capita than we are. I would not be at all
surprised if uploads devoted so much resources to replication that
they would end up extremely impoverished per capita (i.e. owning
nothing but the software they are written on, and renting all
hardware), although mind-bogglingly numerous.

On the other hand, things could turn out differently. Let me first
discuss "wealth". There are certain material resources that improve
survival of those who control them. The survival of Masai herdsmen
depends to a large extent on the number of cows they own, and thus
cows are how a Masai would measure wealth. Owning money always did,
and to a lesser extent, still does improve survival by an indirect
effect on the behavior of humans around the owner. By and large, by
"wealth" we mean "resources whose accumulation improves survival" (the
story is more complicated but we can leave that aside).

The poor-upload scenario assumes that accumulation of resources would
not greatly improve the relative fitness of an upload. Let's say that
most resources in the upload world are derived from the work of the
minds themselves - new ideas, new information are important, while
acquisition of matter (carbon, silicon, space) represent a small
fraction of resources produced, in direct analogy to farming today. In
this situation the amount of resources produced would scale with the
number of minds. A mind would not be able to significantly
out-accumulate other minds of similar power. Copying oneself (or
reproducing in other ways, perhaps analogous to sexual exchange of
information) would not decrease the resources at your disposal and the
most prolific minds (in the replicative sense) would come to dominate
the society by their sheer numbers. Here the only resource you control
is your own mind, and this is sufficient to survive and replicate.

But what if important resources could be better produced by extensive
methods, e.g. taking over a cubic mile of computronium and running
calculations not involving the presence of other minds, such as
million-dimensional simulations of false-vacuum decay paths? Then
there would be a viable strategy consisting of acquisition and
accumulation of computronium without making many copies of yourself.
As long as you have the means of defending such resources, your
economic position in the upload world would be good, and your survival
would be supported by the resources you control. There would not be
much evolutionary pressure towards massive replication, and per capita
wealth differences could be enormous.

In practice I would think that there would be an equilibrium of
various strategies, depending on the details of application. Some
tasks, such as dealing with quickly mutating viral entities, would be
dealt with by swarms of small, independent and quickly replicating
minds while large scale integration of data could be the domain of
very long-lived entities controlling large resources.


> > That's a good point, but still, just how long can a continuously
> > decreasing fraction of the population continue to hold on to a
> > continuously increasing fraction of the wealth?  Something breaks
> > sooner or later.
> Well, there's capitalism in a nutshell. I wonder what will happen? But
> on topic, there's no reason to think that the fraction of people
> holding these wealth-attracting ideas would decrease over time.

### As I discussed above, it all depends.

And, BTW, capitalism actually leads under most human conditions to a
continuously increasing fraction of workers controlling continuously
increasing wealth.... but this is a whole different story.


More information about the extropy-chat mailing list