From: Robert J. Bradbury (bradbury@aeiveos.com)
Date: Tue Apr 30 2002 - 02:22:04 MDT
On Tue, 30 Apr 2002, Emlyn O'regan wrote:
> [snip], but it seems clear to
> me that biological enhancement and modification of humans can happen in many
> more interesting and rapid ways than by modifying the genome of embryos.
Yes, but Greg didn't believe that a few years ago. Though I think he is
slowly moving in my direction.
> Firstly, my understanding is that with gene therapy, we will use modified
> bacteria to modify the genes in somatic cells, thereby allowing us to alter
> fully adult humans.
All current gene therapies, that I'm aware of, and almost all,
except those that I've proposed, use either viruses or naked
DNA to deliver the code. To my knowledge I'm the only individual
willing to stick his neck so far out into the guillotine of history
and suggest gene therapies (actually genome patches) based on bacteria
could be feasible.
> The logical extent of this is the ability to perform some rather
> radical modifications on adults, purely biological.
True, but they aren't anywhere near as far as you can go with nanotech.
> Next, of course, is nanotechnology. The accepted path to full blown nanotech
> is via taking control of cellular mechanisms of protein production, and
> harnessing them to build new machines (proteins) of our own design, which
> can eventually bootstrap us into non-biological nanotechnology.
There are some big hurdles here involving the immune system treating
non-"self" proteins as foreign and attacking cells producing them.
> With this level of technology [soft/wet nanotech], and
> wholy aside from non-biological nanotech, we will have the power to
> understand human biology to a far more fine-grained degree (because sensors
> can travel within us to gather information);
You still have some rather significant I/O problems. I'd place
my bets on things like neuron patch clamps, microelectrodes,
pill-sized cameras, ultrahigh resolution NMR or ultrasound to
be our "eyes" and "ears" into the body, not engineered cells
for quite a number of years.
> [snip] to begin to simulate human biology at a fine-grained level
> in software, and thus study potential manipulations of that biology in
> software; we will have the power to implement those manipulations in living
> humans. We will also have the beginnings of mind upload technology.
With the genome, the proteome and the crystal structure of most/all
of the proteins, you have all the information you need for the simulations.
You will still need a *lot* of computer power.
> There's a lot more besides this to talk about with molecular biological
> technology - Anders or Robert could do it far greater justice. Certainly,
> its a lot more than merely modifying the genome in the germline.
Yes, that was my notorious objection at Extro 4. I think Greg is
slowly starting to lean my way.
> I'm not writing off the computer technology->AI stream of development, by
> any means; my impulse is to say that machine AI will come before we can do
> significant modification of the human body/brain at the molecular level.
> However, I wouldn't say it with 100% certainty.
I think the order may be:
a) We have the capability for radical genome/body mods.
b) We get very interesting levels of machine AI
c) We get approval from regulatory authorities to actually use (a)
on a widespread basis.
To me right now those look to be about 2005-2010, 2010-2015 and
2015-2020 respectively.
> This is because the same factor drives both technological domains, and that
> factor is computers. The same hyper-exponential increase in computing power
> which is used as proof of eventual machine dominance will also increase our
> molecular biological abilities, and at the same rate, because the key
> advancement is the ability to simulate cellular mechanisms.
I disagree. I think you can do a *lot* of radical genome patching
without having to have cellular simulations. If indeed we do need
cellular simulations to apply patches, then the date of (c) will
be much delayed because Moore's Law isn't moving fast enough
to accomplish this for the dates I've indicated (go figure out
the number of atoms in proteins that have been simulated vs.
the number of atoms in the volume of a eukaryotic cell!).
Robert
This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 09:13:43 MST