Re: The wirehead problem

Tim Tyler <seemysig@xxxxxxxxxxxxxx> wrote:
Curt Welch wrote:

So the question before us, is whether the meme is the map, or the
territory.

If we build an automated machine tool which can build machines from
scratch, and it's driven by software which specifies what robot to
build, and it builds a PC, is the CD which holds the PC blueprint the
meme, or is the behavior of "building a PC" the real meme here? If the
first machine sends the CD to a second machine, and the second machine
starts to build PCs, then we can say the meme of "PC building"
replicated from one machine to the other.

I would argue that meme is the territory, not the map. The meme is
really the behavior that gets produced. How the behavior moves from
one machine to the next, is not important or relevant, to the existence
of the meme.

It's the gene/phenotype split in biology.

Memes are like genes - and not much like phenotypes.

Yes, except genes are _unique_ physical objects that have an ability to
survive. No matter what phenotypes they help to influence or create, it's
the survival of the gene itself that is important.

When we talk about memes, we have switched our focus from the genotype to
the phenotype. An idea, like the meme of suicide, exists not as a unique
physical object, but as a large class of different types of physical
objects which share a common attribute - a pattern of behavior.

So when we talk about the spread and survival of memes, we are actually
talking about the spread and survival of phenotypes. It spreads by the
transmission of information, but unlike the gene, the transmission doesn't
take a single form, it can take an infinite number of different forms. So
when we talk about a meme, there is no single genotype which exists. The
commonality only exists in the phenotype.

Genes don't have the power to duplicate phenotypes. If one person has
green eyes, another person can't adapt and change to match their eye color.
Genes don't change in order to duplicate a phenotype that exists in another
person.

But brains can change in order to duplicate patterns of behavior in other
people. And as such, when we talk about memes, we are talking about the
movement of that pattern of behavior from one person to the next. We are
talking about the movement of phenotypes because with the help of machines
like brains, behavior phenotypes can move from person to person.

When we talk about software moving from one computer to the next, we have
yet another mechanism of replication at work that is sort of half way
between how genes work and how human brain memes work. Like genes, there's
is information being encoded and passed from one system to another which
represents the genotype, and like genes, the same genotypes are expected to
produce the same phenotypes (aka the same source code produces the same
behavior on different computers). So the copying of the software copies
the genotype and the phenotype at the same time - just like with genes, not
not like memes - which don't copy their genotype data. But unlike DNA
genes, the software isn't stored in a single fixed configuration (amino
acid pairs), they can be stored in many different types of memory, but the
information is fixed.

In the brain, we can pretend the encoding of a phenotype (aka behavior, aka
meme) might be done by chaining the weights of connections in the network.
If we had the ability to encode those weights as strict binary data, and
copy that to another brain, it would work like software in moving the
behavior form system to system. With AI systems, we can do that, but with
human brains, there is no practical way to do it.

So all these systems of copying have features in common, but there is no
strong dividing line between them - and as more systems are created in
time, the lines will become even more blurred. When we talk about genes,
we are focused more on the genotype, when we talk about software, we are
focused sort of half and half on the genotype and phenotype, (we talk about
both the code and the behavior as being the "software"), and when we talk
about memes, the focus is on the phenotype.

--
Curt Welch http://CurtWelch.Com/
curt@xxxxxxxx http://NewsReader.Com/
.

Relevant Pages

  • Re: The root of all evil? - Dawkins Documentary
    ... selection not being a very common phenomenon, ... Dawkins himself refers to genes are 'selfish cooperators'. ... what the biochemical nature of life itself and how it might have begun. ... parasitical is one manner in which a meme can operate, ...
    (talk.origins)
  • Re: Genetic or Epigenetic: The Causal Basis of Evolutionary Change
    ... genes that has led to an evolutionary morphological change in the ... is the transmission of these genetic changes (not your imaginary CNS ... So here we have *three* heritable phenotypes (dwarf, ... ultimately due to specific genes (and their DNA-based regulation). ...
    (talk.origins)
  • Re: Baldwin Effect-Patent Nonsense
    ... Phenotypes can only be selected and not the genes which code for them ... Phenotypes only can be selected and not the genes which code for them... ... of the species-memberthat first learnt these behaviors and NOT the ... explaing/preaching the principle (and existence) of patterning ...
    (sci.bio.evolution)
  • Re: What do geneticists think of memes?
    ... Genes are inherited almost exclusively vertically (parent to ... >>the meme does. ... >>genetics math that can be transferred from genes to memes. ... >>memetics, it is anyone's guess as to whether individuation matters ...
    (sci.bio.evolution)
  • Re: OT: Local Optimization?
    ... The people who make the laws, usually instigate laws for their own ... > a local stability point with the genes most humans have had selected. ... Its a transient random meme that doesn't maximise numbers of all ... > talking as though it's a given assumption that religion is the source ...
    (sci.electronics.design)
Loading