Re: Uploaded memories

Robert J. Bradbury (bradbury@www.aeiveos.com)
Wed, 8 Dec 1999 12:34:37 -0800 (PST)

On 8 Dec 1999, Anders Sandberg wrote:

> Damien Broderick <d.broderick@english.unimelb.edu.au> writes:
>
> > But maybe I've misunderstood him as well. (I also suspect that the kind of
> > mad-dog ravening exponential runaway that *could* happen under, say,
> > Robert's scenarios changes everything so wildly that *anything* we say
> > makes no difference - Eliezer's long-held position, of course.)

Well, I do say we can exponentially build the hardware, I haven't commented very much on what we run on it. If its anything like today's software development process, its going to take us the lifetime of the universe that remains to do anything with it (which is I suppose what Eliezer is trying to solve), though probably not for the same reasons.

>
> This explosion would at the same time be competitive for
> all the resources, and likely post-economical in the sense that the
> huge amount of resources becoming accessible would likely make each
> individual richer and richer unless they multiplied faster than the
> econosystem could grow.

You should read Science News Nov. 27 1999, pg 340-341 citing an article in Nov. 25, Nature by Jef Huisman and Franz J. Weissing who appear to have a mathematical rationale for why sub-fit species can survive and prosper. Its fundamentally wrapped up in the number of essential nutrients there are. The more nutrients there are the more species an environment can support, precisely because of the exponential growth problem. A species grows until it exhausts the supply of the essential nutrient it is best at harvesting, then another species that depends on different nutrients grows until that nutrient becomes limiting, etc.

So in switching from bio-life (with a few essential elements, C/H2O/N/Fe) to computronium you get dozens of elements that can be used in interesting ways (thats in the MBrain architecture). Then at an entirely different level, you have uploaded, evolving copies where "memes" become the nutrients. Pity poor Damien or Gina who have to sit around waiting for people to come up with new information so they can organize and redistribute it. Pity poor Anders waiting for new ideas for character traits for his role-players to occur to him. Pity poor Greg waiting for new laws to be made...

While, these individuals are spinning their electrons, entities that aren't dependent on those "nutrients" are multiplying away on some underutilized meme-set getting ready to become the dominant species.

It may be that one of the reasons that "everyone" is so "busy" in our society is that the "nutrients" in our society have gone "virtual". We have graduated from "hard-memes" to "soft-memes" which are much easier to copy and distribute and intermix in different proportions. We are facing a nutrient deluge.

>
> [snip] but it could also (and I believe this is more likely)
> diversify into an evolutionary radiation into the new resources, where
> the distribution of resources would be less competitive.

The discussion & paper I've pointed out above seems to suggest that. The problem now is that nature, and to a much lesser degree, humanity, are dealing with a limited nutrient sets. Think of something like the fashion industry as supplying nutrients for the masses and what would happen if someone new came up with something really different. The people who consumed that new nutrient would be viewed as chic, cool, hip, etc. In the meantime the old-nutrient users would be saying, "I don't need that". Who dominates? I don't know. It probably depends on the nutrient and the environment and the degree to which substitute nutrients are available. But the theory seems to say that the more nutrients there are, the more species there will be.

Robert