You have also misinterpreted what I said.
Some of the old DNA and related survival mechanisms from way back when
life was a blob of cells in a hostile environment can be a serious
threat to the well being of a complex differentiated organism.
In general, populations survive because each individual survives long
enough, on average to transmit their genome to a descendant. In
mammals - and other diploid species that reproduce sexually - this
involves two descendants per individual, since each descendant get
half of each parents genome.
Once you have survived long enough to do this, your continuing health
provides a progessively decreasing benefit on your reproductive
success, and detracts from the potential reproductive success of your
off-spring.
Actually in humans you need the adults to live at least long enough to
look after their children up to almost sexual maturity.
Humans are odd in that they cooperate effectively enough that a
surviving grandmother contributes to the reproductive success of her
daughters by increasing the survival rate of her daugther's off-
spring, and this does seem to have pushed human evolution to select
for relatively long life compared with the other anthropoid apes.
I thought familial protective behaviour did occur in other mammals too
(sometimes with only the highest ranking females breeding).
We can't know that DNA can't plan. We can know that it doesn't have
anywhere safe to put it's plans nor any mechanism to check out how
they are working out, beyond the simple survival of the indvidual
embodying the latest modification, whose genome doesn't seem to
contain anything that looks like a version control system. None of
that is dogma.
The crucial point here is that unless the DNA and what it codes for is
challenged from time to time you have no guarantee that it still works.
Or that it still codes for useful behaviour in the environment of an
animal with complex organs as opposed to a blob of slime mould.
It is certainly complex, but nobody who has thought about the
complexities of the jobs it can do is going to be astounded by the
complexity, though they may be astonished that it works at all.
And that according to the local chemical environment stem cells can be
coaxed to develop into very different specific organs.
That's the immune system. But that's not the kind of intelligent
design system that you are talking about. Finding the right antibody
to mass-produce to swamp a virus of a bacterium is rather simpler job
than the one's you seem to have in mind.
The immune system does store patterns of previously recognised
infections as antibodies, some are even passed on to offspring. Though
this can also be dangerous where parents have opposite Rhesus factors.
Sure. Look how well we've done, now that we can transmit information
from generation to generation by something more flexible than DNA
sequences. And since we aren't obliged to have the cultural
transmitted information subjected to mutation to maintain a pool of
variation to cope with a changing environment, we can use error
detection and corredction to make sure that the information we
transmit is good.
Sure. But we are the only example that we know about.
And we are doing it by harnessing a subset of the viral RNA toolkit and
some very clever chemistry. Someone is presently close to swapping the
stop tag "TAG" for "TAA" in E. Coli with the intention of freeing "TAG"
up to code for a novel amino acid. Some details at New Scientist:
http://www.newscientist.com/article/dn20694-e-colis-genetic-code-has-been-hacked.html
(unsure if you need to be subscriber to view)
They are literally harnessing evolutionary mechanisms to do the work for
them by making controlled crosses of tweaked bacteria.
Your have been perfectly sound rational reasons why the scheme that
you proposed won't work. This isn't disapproval, it's rational
scepticism.
His ideas are always "Larkin correct" and we all know what that means.
True. But evolution got to the place you think it is already at by
evolving us, rather than by putting a version control system into the
DNA. I'm not saying that you couldn't design a version control system
that could run in a genome-like processor - in fact I imagine that
genetic engineering will do exactly that in a generation or two - but
such a processor would have to have much better error-detection and
correction than anything we've yet found inside the genome.
The genome's error correction is actually surprisingly good and the cell
death mechanism for uncorrectable errors does work most of the time. But
if it fails then there is a really bad problem if the resulting dodgy
cell is effectively immortal and reverts to the most primitive survive
and reproduce at any cost cellular strategy.
Such a processor would presumably have to cope with the need to
generate random variation to cope with the eventual changes in the
environment (such as better software in rival processors) but with any
lack we could give it a simulator in which the variants could fight it
out, rather than by testing them by raising lots of children with
genetic defects aka sub-optimal variations.
The main thing that our mastery of DNA manipulation will give us in the
medium term is the ability to eliminate certain very nasty genetic
diseases like haemophilia and cystic fibrosis. See for example:
http://www.sciencedaily.com/releases/2009/09/090914172332.htm
We will have to decide at some point whether to allow GM style
corrective gene therapy on embryos and small children. I am not sure
where I stand on this from a purely ethical point of view. But I think I
could be persuaded those that want this treatment should have it.
It would be a very bad idea to use the technology to correct all
"mistakes" in the human genome. Maintaining genetic diversity is very
important as we will one day learn to our cost when industrial scale
high yield agricultural monoculture goes pearshaped.
Regards,
Martin Brown
