From: Anders Sandberg (asa@nada.kth.se)
Date: Fri Feb 18 2000 - 09:54:25 MST
Some interesting papers I found in my piles this week:
* RNA and DNA computing
* Asteroid defense
* Mesicopters
* Information theory limits control
* Self-assembling computations
* Future psychology
* Irish coffee against brain damage?
* Ageing in yeast, worms and humans - same system?
* Do we live long because we are smart?
Molecular computation: RNA solutions to chess problems by Dirk
Faulhammer, Anthony R. Cukras, Richard J. Lipton and Laura F. Landweber
in PNAS, 97:4 1385-1389 2000
We have already seen DNA computers solve problems, but this is the
first RNA computer I have seen. They solve the problem of placing
knights on a chessboard so that none are threatening each other by
creating a mixture of RNA strings encoding all possible states,
digesting all the illegal ones using ribonukleases and then reading
out the winning sequences. Still a fairly small problem, but nice
nevertheless.
DNA computing on surfaces by Quingha Liu, Liman Wang, Anthony
G. Frutos, Anne E. Condon, Robert M. Corn and Lloyd M. Smith, Nature
403 175--179 2000
Another DNA computing paper, this time using DNA strings attached to a
support to solve an instance of the satisfiability problem. This is
likely a more scaleable method than the usual wet approach.
SHIELD: A Comprehensive Earth Protection System by Robert E. Gold, A
Phase I Report to NASA Institute for Advanced Concepts
http://peaches.niac.usra.edu/files/9801/9801Final/GoldFinal.pdf
A report about what would be needed for an Earth defence system
against asteroids and comets. Contains a nice review of the threats,
ways of averting threats and estimates of what such a system would
need to be able to do. Gold suggests a system containing sentry
satelites in solar orbit at the distance of Venus to catalogue NEOs,
and an early warning system at Jupiter orbit for detecting long
periodic comets. As for the active defense, the "soldiers", he discuss
kinetic impact, nuclear detonation and various propulsion methods to
redirect the threat in time. Overall, a cautious and readable paper.
Mesicopter: A Meso-Scale Flight Vehicle, NIAC Phase I Final Report
http://peaches.niac.usra.edu/files/9801/9801Final/KrooFinal.pdf
The results of a program researching flying vehicles on the centimeter
scale. They have developed and built an externally powered
"mesicopter" with 4 rotors with 1.5 cm diameters, which was able to
lift over 700 mg (more than enough to support the 325 mg motor-rotor
assembly). Current work involve free flight tests and onboard
lightweight batteries (for missions lasting tens of minutes). A lot of
aerodynamics I don't understand well, but a cool little device which
could likely be a great basis for swarm systems.
Information theoretic limits of control, by Hugo Touchette and Seth
Lloyd, Physical Rev. Letters vol 84:6 2000 1156--1159
A nice little analysis of controllability of physical systems based on
information theory. The second law of thermodynamics places limits to
how little dissipation we can get when using open-loop control. Also,
feedback control is a kind of zero sum game, where each bit of
information a controller gets from a system can at most be used to
decrease the entropy of the system with one bit.
Using lateral capillary forces to compute by self-assembly, by Paul
W. K. Rothemund, PNAS 97:3 984--989
It is possible to create sets of tiles that are computationally
universal; if they are used to tile a plane the edges force the tiling
to act as a computing device. Rothemund has explored self-assembling
computation using acrylic tiles with hydrophobic or hydrophilic edges,
coding tiles generating checkerboard patterns, penrose tiles or the
XOR cellular automaton. It turns out that they gladly self assemble,
but you get defects of various kinds; the amount of defects seem to
hinge on the tiling rule. Still far from any useful computation, but
self assembly is a very young field.
Looking back: The History of Psychiatry in the 21st Century, by Robert
Michels, Arch Gen Psychiatry Vol 56 Dec 1999
A fictional paper sugegsting one way psychiatry might change in this
century. The basic idea is that widespread use of antenatal screening
with DNA chips makes parents avoid selecting children with genes
causing or predisposing psychatric disorders. The result is a world
where they become very rare, and the only remaining mental illnesses
are those caused by experiences or upbringing. Clinical psychiatry
turns increasingly more towards the adjustment of personality, with
personally customised treatments based on known genetic potentials and
risks. Overall a very positive and transhumanist article.
Combination of low dose ethanol and caffeine protects brain from
damage produced by focal ischemia in rats, by R. Strong, J.C. Grotta
and J. Aronowski. Neuropharmacology 39 (2000) 515-522
This study studied the effects of the combination caffeine and alcohol
on brain damage in rats. They found that caffeine alone, when given
before ischemia, had no effect, and that alcohol alone actually
worsened the damage. But the combination, given orally a few hours
before the damage or injected within 120 minutes afterwards, decreased
the brain damage significantly! Unfortunately, chronic daily caffine
and ethanol consumption for a few weeks before the experiment removed
these effects, so Irish coffee should only be drunk when you are close
to ischemia in time. Hmm... almost sounds like a nice sendoff ceremony
for would-be cryonauts :-)
Brains, Maturation times and Parenting, by John Allman and Andrea
Hasenstaub, Neurobiology of Aging 20 (1999) 447-454
Why does it take so long for humans to grow up, and why do we age so
slowly compared to other primates? One explanation could be that we
need a long maturation period to develop our complex brains (since
they need to interact with the environment to develop, and this cannot
be speeded), and this in turn has favored the development of genes
that slow brain senescence. This article analyses the time and costs
for brain development, and compares relative brain size to development
time for various species. They also look at other parenting factors
for longevity, such as the relative male-female survival as a function
of how much parenting they do in different species. All in all, this
article suggests that the reason we live so long even without
technology is that we need the cultural transmission and memory of
parents and grandparents, whose skills help their children and
grandchildren survive.
Mutations in signal transduction proteins increase the stress
resistance and longevity in yeast, nematodes, fruit flies and
mammalian neuronal cells, by Valter D. Longo in Neurobiology of Aging
20 (1999) 479-486.
A review of proteins involved in extending the lifespan of yeast
cells, and how they relate to proteins known to affect longevity in
C. elegans and drosophilia. The article discuss how these proteins
relate to stress resistance, antioxidant activity and ageing, and
suggests that the same mechanisms go for many eucaryotes.
-- ----------------------------------------------------------------------- Anders Sandberg Towards Ascension! asa@nada.kth.se http://www.nada.kth.se/~asa/ GCS/M/S/O d++ -p+ c++++ !l u+ e++ m++ s+/+ n--- h+/* f+ g+ w++ t+ r+ !y
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 15:26:54 MST