Re: Inadvertent media malevolence

Anders Sandberg (asa@nada.kth.se)
14 Jun 1999 17:39:30 +0200

"Raymond G. Van De Walker" <rgvandewalker@juno.com> writes:

> On 08 Jun 1999 13:39:17 +0200 Anders Sandberg <asa@nada.kth.se> writes:
> >"Raymond G. Van De Walker" <rgvandewalker@juno.com> writes:
>
> >>>> . . . Do any of you have specific ideas for preventing an
> interspecific war of
> >>>> extermination . . . [?]
>
> >Suppose tomorrow a transhuman species with a green skin (but no other
> >differences) appeared. Would this cause a war of extermination . . . [?]
>
> Green skin probably isn't enough of a difference. However, let's say
> transhumans all have the IQs of Marilyn Vos Savant, and see what happens

Hmm, maybe she is not a good example since her high IQ doesn't seem to help her that much given the factual errors and mistakes I have heard about her columns. Just look at Mensa - lots of people with high IQ but they do not seem behave that smartly. And I do not seem any evidence that Mensa is hated by the rest of society or in any risk of taking it over.

> Ecological research seems to indicate an extinction. Let me quote from
> "The Science of Ecology", Paul Ehrlich & Jonathon Roughgarden, 1987 (pg
> 351),
> the discussion of "resource limited guilds"
>
> "MacArthur and Levins [Am. Nat. 101:377-385, 1967] were the first to
> investigate limiting similarity. They showed that the limiting
> similarity between the two species depends on the ratio of the carrying
> capacities. Specifically, let species 2 be the species with a lower K
> [Ray: the fitness constant] The limiting similarity depends on K2/K1.
> If this ratio is much less than 1, indicating a severe disadvantage to
> species 2, then species 2 must be quite different from species 1 in order
> to coexist with it. However, as K2/K1 approaches 1, the limiting
> similarity approaches 0."

So what about the difference in fitness between the developed world and the underdeveloped world? The carrying capacity of technological advanced people can be extremely high, while the carrying capacity of less advanced people is much less. Are we seeing a situation where the first world is outbreeding the third world? No, quite the reverse, and mainly because first world people have fewer children to get a higher standard of living.

Another example would be people with an academic background compared to people without; the former can obviously get not just specialized jobs but also the nonspecialized jobs the latter are dependent on. Still, you don't see that academics try to crowd out unskilled labor or unskilled laborers trying to wipe out academics.

> The book then gives two quantitatively studied examples: freshwater fish
> in lakes [Werner, Am. Nat. 111:553-578, 1977], and Anolis lizards on
> Caribbean islands [Roughgarden Am. Nat 108:429-442, with companion piece
> Theor. Pop. Biol. 5:163-186] The above logic and studies were
> originiated to explain structured guilds, groups of similar species that
> partitioned resources by apparently evolving to different sizes, or some
> other accomodation. The book has too many of these to excerpt.

>

> By definition, transhumans will be high-fitness groups, causing
> theoretically certain extinction of ordinary humans in the absence of
> other effects. This gives low-fitness humans strong incentives to
> exterminate transhumans early-on.

No, you are not applying the above ecology correctly here. The guild phenomenon instead suggest that it will be advantageous to specialize, so that the high-fitness transhumans will deal with certain parts of the economy, and the humans with others (likely with further subdivisions between different levels of transhumanisation). This is exactly what the law of comparative advantage in economy suggests. Instead of extermination, specialisation.

> > There is no
> >economical reason for them to try to remove all individuals of the
> >other species. In fact, there might be strong economical incentives
> >for specialization instead (such as the law of comparative
> >advantage).
>
> Gosh, it sounds so plausible. We are so used to intra-specific
> responses, so conditioned to think that anything that talks is the same
> species, that we forget how different inter-specific responses are.
>
> The science is against it. Most species completely ignore other species'
> systems of territory and resource-allocation. I believe that transhumans
> eventually will ignore human systems, too, when they become able to brush
> them aside, in the same way that humans ignore the territories of dogs
> and cats.

That assumes a power gap between humans and transhumans, something I regard as highly unlikely. An enhancing technology doesn't just turn a human into a much much powerful transhuman (with no emotional, social and economic ties to their former friends) - enhancing technologies improve humans, broadening the normal distribution and pushing it more in the positive direction. They occur in a social and economic context. Breakthroughs in technology appear to be rare and are usually more qualitative at first than quantitative (the first airplanes were less useful than baloons and zeppelins), gradually developing quantitatively. This means that I consider it unlikely that there will ever be a situation of mere humans versus super-powerful transhumans, instead we will have a society where there will be everything from ordinary humans to superhumans with people of all levels in between.

This means that the superhuman who brushes humans aside will have to deal with a lot of posthumans and transhumans who dislike it (for reasons ranging from ethics to pure economy - "you wiped out the cast of my planetary soap opera?!"). There may even be powerful entities acting on the behalf of lesser entities, paid by them to act as their protector. In short, the situation is not clear cut.

> >The main reason for the different behavior is that humans
> >are not just trying to maximize the number of offspring they have, but
> >rather have many other memetic goals.
>
> I'm not sure I believe this. It looks _to me_ like these are secondary,
> lower-priority goals that have floated to the top because
> species-survival is a solved problem at this time. Here's why: memes
> necessarily evolve to maintain their growth-media. Thus growth-media
> survival will have priority over other memtic goals.

This does not follow. Memes do not need to evolve growth-media maintenance, just look at monasticism. The meme of birth control is doing great right now, despite limiting the eventual number of hosts. To a meme human population is not the growth media, it is the space of information and attention - the Internet is closer to a meme-driven growth medium.

If people were driven to maximize the number of offspring, the fastest population growth would be in the rich countries where people can afford many children, not in the poorest countries. People seem instead follow a K strategy when resources are rich, investing much in the children.

> Also, I have grievous doubts that transhuman and human memes will remain
> infective across species. Theory is against that, too, provided that
> memes are any of parasitic, commensal or symbiotic with hosts. All of
> these relations cause strong coevolution, and thus should cause strong
> memetic specialization.
>
> Memetic specialization would prevent shared goals and values from being
> stably-shared between the species, over evolutionary time.

This again assumes strong speciation, with a huge gap in between. But if transhumans and humans form an economic community, even one where specialization is advantageous, communication between them is necessary and would mean there are ways of memes to leak over. And if you still believe that memes create their own growth-medium, they would really exploit the possibility of opening communications with the other side.

> >. . . it looks more like increased diversity . . . makes a win-win
> situation . . .
>
> But the scientific predictions do not depend upon, nor should we expect
> them to be changed by, divergent abilities. The science depends on
> speciation combined with convergent resource use, and asymmetric
> competition. These preconditions are the _hypothesis_ of transmumanity.

Are they? The resource use part becomes doubtful since transhumans could exploit new resources humans cannot exploit (such as space), and the transhumanist idea of development doesn't necessarily imply a one-dimensional form of progress but rather many forms of improvements that does not have to be along competitive dimensions or lead to highly nontrivial competitive networks (transhuman A and transhuman B compete for a resource X, but cooperate in supplying each other with resources Y and Z).

> The _only_ way out that _I_ see is to explicitly prevent human
> speciation, which really would not be too limiting. Think about it. A
> clade of humans, spanning numerous levels of intelligence, or other
> ability, but all able to interbreed, would be both powerful and
> plausible.

And likely more economically viable than distinct species - that ways good strategies can spread memetically, genetically and technologically within the clade.


Anders Sandberg                                      Towards Ascension!
asa@nada.kth.se                            http://www.nada.kth.se/~asa/
GCS/M/S/O d++ -p+ c++++ !l u+ e++ m++ s+/+ n--- h+/* f+ g+ w++ t+ r+ !y