From: Robert J. Bradbury (bradbury@aeiveos.com)
Date: Sun May 20 2001 - 09:45:55 MDT
CurtAdams (my own favorite neoluddite wrote):
> Organisms with complex macrostructural construction abilities
> have rather slow doubling times. If you want to look at a house-
> builder, the appropriate model is a tree, with doubling times in
> years or decades.
Bamboo actually grows *quite* fast. I think it can reach 3-4
meter heights in a month or two. The question you have to ask
is *why* are the doubling times that long? Is it an energy
limitation? Is it an essential nutrient limitation? Is it
a water limitation? etc. One of the things that humans are
good at is mass-transport. Figure out what the limiting elements
are in the growth equation and supply them in sufficient abundance
to work around the problem. You could even connect your natural
gas pipeline to a conversion unit that pumped out ATP in abundance
for the house-seeds so they can grow in the dark! Yes, I know it
makes it more expensive, but presumably after your house is
finished growing, it can manufacture natural gas and pump
it back into the pipeline, so you only need a short term loan.
> Further, we're a very long way from enigneering trees to grow
> particular shapes. And think how long it will take to test it! :-)
The shape problem isn't solved yet because its historically
been mostly of scientific interest. I suspect a lot is
known about how Drosophila gets its shape for example.
The Zebrafish genome will lay the groundwork for research in
vertebrate shape development. If the variety of dogs I see
people walking around Greenlake in Seattle are any indication
there is going to be a *big* interest in applying that
informormation. Contrary to Eugene's comments there might
be an application for 6-legged grayhounds if it turns out
they can run faster. And of course the Arabadopsis genome
begins to lay the foundation for understanding plant
structure. You can bet that the seed companies are going
to *love* growing the corn *without* the stalks.
I'm fairly confident we will get fairly fine structural
control.
> I said:
> > I'm willing to make $100 bets with people that the
> > manufacturing cost of designer bacteria with 1 million+
> > base genomes will be under $10.00 by 2010.
> You have to define "designer". I can make a bacteria
> to do certain things for far less than that, but only
> a few particular limited things. If you mean a bacteria
> which can build a house, you're on.
I doubt you could fit a house-seed genome in a megabase. Its probably
at least the complexity of a bamboo genome which even in compressed
form is probably between a 100 Mbp and 1 Gbp. I was however *careful*
about how I phrased the bet -- I said "manufacturing cost" and not
"design cost". Whether we have the designs will depend a lot on
whether some hot shot entrepreneur can convince a bunch of VC's
that there is a market for "houses that grow themselves". I suspect
that is going to be less sellable than "designer organs".
> We don't have anything complex that grows in a fermenter.
> The fact that they're simple is essential to fast growth in
> a fermenter - you can just mix them around.
Actually the nutrient delivery problem is *why* you mix them around.
Bacteria locally deplete nutrients far faster than diffusion can
supply them. Plants and eukaryotic cells however have transporters
that could be adapted to importing nutrients on the abundant side
of the cell and exporting them on the depleted side of the cell.
Then Eugene (just call me a pessimist) Leitl came along with:
> We can structurally modify plants, a bit, but nothing like the
> precision required. For house-growing, deviations of a fraction of
> an inch would have to be recognized and communicated across the
> organism and we don't have anything like that in plants.
Ah Eugene, you have to work *across* technologies as well as in depth!
You use external site-management lasers driven by the blueprint that
interact with highly specific photoantenna complexes within the cells
that in turn are tied to the signal transduction pathways for "divide"
or "commit apoptosis".
That should allow you to control the shape of your new house to
within 10-20 micron accuracy -- much better than human workers!
> What do you do if your house seedling gets a bad case of mycoplasma?
> Do you want your construction people to daily look out for nitrogen
> and microelements in the soil, and see that the bark doesn't get
> infested with wood bugs?
and
> Parasites sap vigor, and hence feedstock ROI. They also
> tie up a scarce resource (people's attention) thus sapping
> elsewhere by proxy.
You only have to tinker with 20 transfer RNA's and proteins
to give your house an entirely different genetic code. I
suspect making stop codons out of the most common amino
acid triplets will put a real dent into Nature trying
to turn your house into lunch. Its also going to have to
get pretty inventive as well if the non-growing surfaces are
completely coated with abalone shell.
Presumably you handle the nitrogen problem in ways similar
to those used by the nitrogen fixing bacteria. Probably
several layers under the surface cells in a relatively
anaerobic environment with transporters for exporting fixed
nitrogen to the aerobic cells. I'm sure you can do the
arithmetic on the mass of the house and the mass of the
essential elements needed. I suspect its probably in the
vicinity of a pickup truck load or less. You can probably
gain a lot by programming in the nutrient export methods
trees use in the fall. It would be interesting to see if
those cells completely reverse the micronutrient flow from
importing to exporting. If you can restrict the micronutrient
needs to the growing surfaces only, I suspect the micronutrient
mass requirements will be very low.
I do agree however that the medical applications will probably
capture the imagination of researchers long before the construction
industry is revamped.
Robert
This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 08:07:44 MST