Thoughtful discussions (rather than the usual doom and gloom predictions) regarding the consequences of genetic engineering and technological progress in computers, AI, etc. may be found in the Extropy Institute'sMailing List. There are many years of discussions in the Archives.
The people involved in these organizations actively discuss and investigate the many issues and concerns related to our future evolution as a species.
Nanomedicine and nanotechnology can be safe
on
Nanomedicine
·
· Score: 4
I was a reviewer for Nanomedicine and I speak with Robert Freitas frequently. He is very serious about designing nanobot medical devices so they are non-replicating, have numerous failsafes, and do not create the possible problems most people envision. One reason writing all three volumes will take 6 years is the depth of analysis that has to be done to meet this standard. While it is doubtful that a single individual can think of everything, Nanomedicine clearly will lay the foundation for safe and very useful nanobots such as Respirocytes.
The problems mentioned by Bill Joy in his interview point out how poorly informed he is. Anyone who has been in the computer industry as long as he has, should know enough to "read the manual(s)" before offering uninformed opinions. The problems regarding nanotechnology run amok have been discussed for many years in the sci.nanotech newsgroups as well as at conferences for the Foresight Institute'sSenior Associates. The basic solutions involve making "safe" (e.g. reviewed, open source) designs available while at the same time developing defenses against nanotech run amok. The Extropy Institute'sMailing ListArchives, for example, contains recent discussions about encouraging the availability of "almost anything" manufacturing boxes (similar to Star Trek "replicators"), while discouraging the availability of "everything" boxes.
Diamondoid or saphire based molecularly assembled nanobots used in medical applications will greatly exceed the capabilities in of "biobots" built on existing genetic machines (DNA, enzymes, bacteria, cells, etc.) because they are stronger, can pack the "code" more densely, and can have more complex programs than the rather "ad hoc" designs that nature has provided us with. Most of the first volume of Nanomedicine is devoted to determining exactly what the physical limits will be on power, communication, mobility, etc. Most of the applications will be discussed in Volumes II and III.
Joy may be right that the technology poses a threat to the "human species", but that begs the question of "Why would you want to run on obsolete hardware?". Anyone who understands even a little astronomy knows that galactic hazards doom biological human forms to death at some point. Only those humans who choose to upload have any hope of living the trillion or so years that seems quite feasible. So while the hopes for biochemical humans are rather dismal even with Nanomedicine, the long term prospects for humanity, based on what nanotechnology allows are quite good indeed.
As far as nanotechnology background material goes, the best (nontechnical) source is Engines of Creation. Other references can be found in Eric Drexler's CV.
Updating O'Neill's vision with nanotechnology
on
The High Frontier
·
· Score: 1
The real problem with O'Neill's vision was that it was based on the idea of using macro-scale technology to build the colonies. That was what made it expensive and is why we don't have such colonies today. NASA did a study in the early 1980's (at the request of Jimmy Carter, one of the few presidents who had an understanding of technology), on how to produce self-replicating factories that would have lowered the costs. The study is online here and here. Robert Freitas was one of the authors of this study, and has indicated to me that one of the problems was the long doubling times (decades?) that the lunar factories required. I strongly suspect the reason for this was because the technologies they envisioned using were macro-scale technologies that did not allow significant amounts of parallization. We know that bacteria have doubling times as low as 20 minutes, and Josh Storrs Hall has estimated that properly designed nanoscale assembly lines may have doubling times as low as 2 msec (see here). Large objects such as O'Neill's colonies can be built rapidly and cheaply if you make your workers small enough, e.g. nanobots.
While commenting on some problems regarding SETI searches, I provide a discussion of how O'Neill's colonies might be updated using biotechnology and nanotechnology. Steel and aluminium are terrible structural materials compared with diamond, buckytubes and sapphire.
The combination of the short replicating times allowed by nanoscale self-replicating systems and the material properties of the strongest materials will allow us to rapidly go far beyond O'Neill's vision -- to the point of dismantling entire planets.
Government support or programs is not required to do this. Molecular Nanotechnology of the type being developed by Zyvex is required. In addition, we need the designs for the nanobots to take apart the asteroids or planets, construct the mass drivers and solar arrays, etc. The lack of molecular designs, is discussed in the Nano@Home proposal. Because we will be able to do the designs at home, a small dedicated group will eventually be able to bootstrap the development of space and achieve the vision O'Neill described. Because of the rapid increase in the available resources (matter and energy) per person, the large number of people living in poverty should disappear as well. The only potential problem I see is if Mind Uploading becomes feasible (or real AIs are developed) and unlimited copying of such entitites is allowed. This has been explored in more detail by Robin Hanson in If Uploads Come First.
I took the time to try and see what format of "The Saga Begins" was better. These were tested on a dual processor 200 MHz Pentium Pro, 64MB ram, running Windows 2000 Beta 3, connected to the net via a 256K ADSL modem. The sound card was an old ISA C. L. SoundBlaster output to a Bose Acoustimass Multimedia speakers.
I rated the players in 4 catagories: Video: The subjective video quality; Synch: The ability of the video & voice to stay synchronized. CPU: The CPU time consumed by each player Mem: The memory required by each player
The results are: Video: RP > QT > MP; mainly because the QT & MP videos appear very dark. Synch: QT > MP >> RP; Quicktime seemed to keep things in sync well, while MediaPlayer seemed to lose sync periodically and RealPlayer consistantly stopped at various points in the video (even when loading from a network file on an unloaded 10Mbps net!) CPU: MP > QT >> RP; ~50% vs. 33% vs 12% of the dual CPU (100%) system) Mem: RP > QT > MP; ~ 11.2 vs. 8.5 vs. 5.5 MB
I ran all three players simultaneously to get the CPU & Memory figures. Since the line is rated at 256K and I was using the 128K ISDN for MP/QT data sources and the RP video was on the local net this should not be network bandwidth constrained. Running in 3 part harmony appears to be difficult to do because you can't seem to balance the sound source volumes.
I'm rather stunned that W2K Beta 3 actually didn't crash during this test!
It seems to me that MP might be the best choice on a faster system. QT wins in terms of maintaining presentation quality. RP doesn't offer much but a good looking video that fails to stay in sync. The MP & QT sound quality was richer than that of RP, but the QT sound seemed to have occasional clicks in it. I don't know if that is a network problem or a software/hardware glitch. I don't believe any packets were lost during these tests.
When you can run this test on Linux let me know (:-))
Disclaimer: I have a background in both computer science and micro/molecular biology and have been studying the problem of aging for ~8 years and the problem of the limits of intelligence for ~2 years.
Minor complaint -- I can't believe the amount of commentary on this topic by people who clearly are ill-informed or do not understand what they are talking about.
I'll address 2 issues:
1) Why We Age?
This problem is well understood by biologists and dates back to the 1950's with theories proposed by Medawar and Williams. There are 2 principles:
(a) The declining force of natural selection. Due to preditors and accidents nature has fewer and fewer individuals from which it can "select" genes which promote health in the elderly. When you have "no" individuals from whom to select natural selection fails. So to evolve longer and longer lived species takes greater #'s (running into environmental limits) and increasing periods of time. Under this axiom, we age because "the program is incomplete".
(b) Antagonistic plieotropy. Under this theory, nature selects for genes which promote survival in the youth and reproductive success, but which in the long run are detrimental. These could include fundamental aspects of biology -- iron is a great carrier for oxygen, but promotes free radical damage to DNA causing program corruption. Under this axiom we age because nature balances personal survival with reproductive fitness.
As the declining force of natural selection and the effects of antagonistic plieotropy vary with the environment it is likely that both of these axioms operate in different species to different degrees.
A consequence of the these theories is that the human genome does not contain the programs necessary to maintain and repair neurons indefinately. So neurons eventually die and the brain gradually loses capacity.
Regarding Merkle's article, Landauer's research and the capacity of the brain --
Many people misunderstood (or didn't bother to read carefully!) the experimental situation that derived the estimate of ~120 MB of storage. This is the "stored" and "retreived" long term memory capacity! It has nothing to do with the bandwidth of our senses (such as the eyes) *or* the processing capacity of the brain (when differentiating between stored tones).
It is certainly true that the neuronal synaptic capacity of the brain is much greater than ~120MB. However, it is not clear how much of that is simply redundant information and how much of it is *generated internally*! (without having to go though our normal short-term to long-term memory processing algorithms). The human brain is very good at "filling in" very bad video or audio information (because its survival depended on it)! A brain that was good at generating "hypothetical" situations that could enhance survival strategies (and storing them for future use) would presumably be quite successful as well.
So while, R. Merkle's/Landauer's discussion may be very useful in determining how much information we may be able to "memorize" for future recall, it tells us little about what the internal memory capacity (that we use for discrimination calculations) really is.
This is important because humans have to deal with very poor quality data feeds. Computers/AIs/etc. may on the other hand be operating on very high quality data (or objects with very well defined meanings) and so may require much less capacity to perform operations of the complexity that humans do (when considering higher thought operations vs. simply separating the data from the noise). The processing done by the human auditory system can already be done by most computers and the human visual system will be compressed into a few chips with ~1 Teraops computational capacity.
For further information see recent books by Hans Moravec (on Robots) and Steve Austad (on Why We Age).
Even if these folks had the compilers that would allow you to take large chunks of code, convert it into a hardware representation and program the FPGA to execute it you still have to get have some DATA to feed the instruction stream! The only people that seem to understand true parallel programming models seem to be the people at Tera Computer). They have the only architecture that can do a context switch on each instruction to allow the processors to execute those instructions that happen to be executable because the operand data fetches are complete. Everyone else (Compaq(DEC), Intel, AMD, Sun, SGI, etc.) consume huge amounts of chip real estate with primary & secondary caches rather than really solving the problem of memory latency. The old CPU/Cache IS DEAD in the long run (the chips get too hot). What will work are architectures like Tera's and/or approaches like " Processor in Memory"/" Intelligent RAM"/" Embedded DRAM" that are innovative ways of dealing with the problem of operand latency and memory bandwidth.
If anyone has read Hans Moravec's "Mind Children" or more recent "Robots" books, in which he discusses transfering mental processes into machines, then they might wonder how close we are to "The Matrix". If you allow that Petaflops computing (1-100x human brain capacity) arrives ~2005-2010 and according to Drexler, nanoassembly arrives sometime between 2010-2020, then constructing a Matrix like supercomputer into which our minds can be uploaded seems feasible with ~25 years (within my lifetime). Why would anyone want to become pure "software"? It seems much more interesting than remaining pure "hardware"!
Many years ago (~30), when I was a teenager, I read a sci-fi novel where the people were engaged in hunting dinosaurs. During their adventure one of them gets killed. Then his life begins anew. After a while, it becomes apparent that the people can't really die, they have some subconscious control over the personalities (friends/enemies)end up in their neverending relalities. At the end of the story they discover they are nothing more than brains kept in vats underground by "machines" that have taken over the world. In order to protect humans from their own limitations but keep them entertained it was decided to provide them with the ultimate "safe" habitats but allow them to continue living (and dying).
The question is Who wrote this book and What was the title?
Slashdot Mirror
Thoughtful discussions (rather than the usual doom and gloom predictions) regarding the consequences of genetic engineering and technological progress in computers, AI, etc. may be found in the Extropy Institute's Mailing List. There are many years of discussions in the Archives.
Some additional sources of useful information include the The Transhumanist FAQ and the Journal of Transhumanism . The World Transhumanist Association is an umbrella organization for many regional transhumanist groups.
The people involved in these organizations actively discuss and investigate the many issues and concerns related to our future evolution as a species.
The problems mentioned by Bill Joy in his interview point out how poorly informed he is. Anyone who has been in the computer industry as long as he has, should know enough to "read the manual(s)" before offering uninformed opinions. The problems regarding nanotechnology run amok have been discussed for many years in the sci.nanotech newsgroups as well as at conferences for the Foresight Institute's Senior Associates. The basic solutions involve making "safe" (e.g. reviewed, open source) designs available while at the same time developing defenses against nanotech run amok. The Extropy Institute's Mailing List Archives, for example, contains recent discussions about encouraging the availability of "almost anything" manufacturing boxes (similar to Star Trek "replicators"), while discouraging the availability of "everything" boxes.
Diamondoid or saphire based molecularly assembled nanobots used in medical applications will greatly exceed the capabilities in of "biobots" built on existing genetic machines (DNA, enzymes, bacteria, cells, etc.) because they are stronger, can pack the "code" more densely, and can have more complex programs than the rather "ad hoc" designs that nature has provided us with. Most of the first volume of Nanomedicine is devoted to determining exactly what the physical limits will be on power, communication, mobility, etc. Most of the applications will be discussed in Volumes II and III.
Joy may be right that the technology poses a threat to the "human species", but that begs the question of "Why would you want to run on obsolete hardware?". Anyone who understands even a little astronomy knows that galactic hazards doom biological human forms to death at some point. Only those humans who choose to upload have any hope of living the trillion or so years that seems quite feasible. So while the hopes for biochemical humans are rather dismal even with Nanomedicine, the long term prospects for humanity, based on what nanotechnology allows are quite good indeed.
As far as nanotechnology background material goes, the best (nontechnical) source is Engines of Creation. Other references can be found in Eric Drexler's CV.
While commenting on some problems regarding SETI searches, I provide a discussion of how O'Neill's colonies might be updated using biotechnology and nanotechnology. Steel and aluminium are terrible structural materials compared with diamond, buckytubes and sapphire. The combination of the short replicating times allowed by nanoscale self-replicating systems and the material properties of the strongest materials will allow us to rapidly go far beyond O'Neill's vision -- to the point of dismantling entire planets.
Government support or programs is not required to do this. Molecular Nanotechnology of the type being developed by Zyvex is required. In addition, we need the designs for the nanobots to take apart the asteroids or planets, construct the mass drivers and solar arrays, etc. The lack of molecular designs, is discussed in the Nano@Home proposal. Because we will be able to do the designs at home, a small dedicated group will eventually be able to bootstrap the development of space and achieve the vision O'Neill described. Because of the rapid increase in the available resources (matter and energy) per person, the large number of people living in poverty should disappear as well. The only potential problem I see is if Mind Uploading becomes feasible (or real AIs are developed) and unlimited copying of such entitites is allowed. This has been explored in more detail by Robin Hanson in If Uploads Come First.
I took the time to try and see what format of "The Saga Begins" was better. These were tested on a dual processor 200 MHz Pentium Pro, 64MB ram, running Windows 2000 Beta 3, connected to the net via a 256K ADSL modem. The sound card was an old ISA C. L. SoundBlaster output to a Bose Acoustimass Multimedia speakers.
I rated the players in 4 catagories:
Video: The subjective video quality;
Synch: The ability of the video & voice to stay synchronized.
CPU: The CPU time consumed by each player
Mem: The memory required by each player
The results are:
Video: RP > QT > MP; mainly because the QT & MP videos appear very dark.
Synch: QT > MP >> RP; Quicktime seemed to keep things in sync well, while MediaPlayer seemed to lose sync periodically and RealPlayer consistantly stopped at various points in the video (even when loading from a network file on an unloaded 10Mbps net!)
CPU: MP > QT >> RP; ~50% vs. 33% vs 12% of the dual CPU (100%) system)
Mem: RP > QT > MP; ~ 11.2 vs. 8.5 vs. 5.5 MB
I ran all three players simultaneously to get the CPU & Memory figures. Since the line is rated at 256K and I was using the 128K ISDN for MP/QT data sources and the RP video was on the local net this should not be network bandwidth constrained. Running in 3 part harmony appears to be difficult to do because you can't seem to balance the sound source volumes.
I'm rather stunned that W2K Beta 3 actually didn't crash during this test!
It seems to me that MP might be the best choice on a faster system. QT wins in terms of maintaining presentation quality. RP doesn't offer much but a good looking video that fails to stay in sync. The MP & QT sound quality was richer than that of RP, but the QT sound seemed to have occasional clicks in it. I don't know if that is a network problem or a software/hardware glitch. I don't believe any packets were lost during these tests.
When you can run this test on Linux let me know (:-))
Disclaimer: I have a background in both computer science and micro/molecular biology and have been studying the problem of aging for ~8 years and the problem of the limits of intelligence for ~2 years.
Minor complaint -- I can't believe the amount of commentary on this topic by people who clearly are ill-informed or do not understand what they are talking about.
I'll address 2 issues:
1) Why We Age?
This problem is well understood by biologists and dates back to the 1950's with theories proposed by Medawar and Williams. There are 2 principles:
(a) The declining force of natural selection.
Due to preditors and accidents nature has fewer and fewer individuals from which it can "select" genes which promote health in the elderly. When you have "no" individuals from whom to select natural selection fails. So to evolve
longer and longer lived species takes greater
#'s (running into environmental limits) and
increasing periods of time. Under this axiom,
we age because "the program is incomplete".
(b) Antagonistic plieotropy.
Under this theory, nature selects for genes
which promote survival in the youth and reproductive success, but which in the long run
are detrimental. These could include fundamental aspects of biology -- iron is a great carrier for oxygen, but promotes free radical damage to DNA causing program corruption. Under this axiom we age because nature balances personal survival with reproductive fitness.
As the declining force of natural selection and the effects of antagonistic plieotropy vary with the environment it is likely that both of these axioms operate in different species to different degrees.
A consequence of the these theories is that the human genome does not contain the programs necessary to maintain and repair neurons indefinately. So neurons eventually die and
the brain gradually loses capacity.
Regarding Merkle's article, Landauer's research and the capacity of the brain --
Many people misunderstood (or didn't bother
to read carefully!) the experimental situation
that derived the estimate of ~120 MB of storage.
This is the "stored" and "retreived" long term
memory capacity! It has nothing to do with the
bandwidth of our senses (such as the eyes) *or*
the processing capacity of the brain (when
differentiating between stored tones).
It is certainly true that the neuronal synaptic
capacity of the brain is much greater than ~120MB.
However, it is not clear how much of that is simply redundant information and how much of it
is *generated internally*! (without having to
go though our normal short-term to long-term
memory processing algorithms). The human brain
is very good at "filling in" very bad video
or audio information (because its survival
depended on it)! A brain that was good at
generating "hypothetical" situations that
could enhance survival strategies (and storing
them for future use) would presumably be quite successful as well.
So while, R. Merkle's/Landauer's discussion
may be very useful in determining how much information we may be able to "memorize" for future recall, it tells us little about what the internal memory capacity (that we use for discrimination calculations) really is.
This is important because humans have to deal with very poor quality data feeds. Computers/AIs/etc. may on the other hand be operating on very high quality data (or objects with very well defined meanings) and so may require much less capacity to perform operations of the complexity that humans do (when considering higher thought operations vs. simply separating the data from the noise). The processing done by the human auditory system can already be done by most computers and the human visual system will be compressed into a few chips with ~1 Teraops computational capacity.
For further information see recent books by Hans Moravec (on Robots) and Steve Austad (on Why We Age).
Even if these folks had the compilers that would allow you to take large chunks of code, convert it into a hardware representation and program the FPGA to execute it you still have to get have some DATA to feed the instruction stream! The only people that seem to understand true parallel programming models seem to be the people at Tera Computer). They have the only architecture that can do a context switch on each instruction to allow the processors to execute those instructions that happen to be executable because the operand data fetches are complete. Everyone else (Compaq(DEC), Intel, AMD, Sun, SGI, etc.) consume huge amounts of chip real estate with primary & secondary caches rather than really solving the problem of memory latency. The old CPU/Cache IS DEAD in the long run (the chips get too hot). What will work are architectures like Tera's and/or approaches like " Processor in Memory"/" Intelligent RAM"/" Embedded DRAM" that are innovative ways of dealing with the problem of operand latency and memory bandwidth.
If anyone has read Hans Moravec's "Mind Children" or more recent "Robots" books, in which he discusses transfering mental processes into machines, then they might wonder how close we are to "The Matrix". If you allow that Petaflops computing (1-100x
human brain capacity) arrives ~2005-2010 and according to Drexler, nanoassembly arrives sometime between 2010-2020, then constructing a Matrix like supercomputer into which our minds can be uploaded seems feasible with ~25 years (within my lifetime). Why would anyone want to become pure "software"? It seems much more interesting than remaining pure "hardware"!
Many years ago (~30), when I was a teenager, I
read a sci-fi novel where the people were
engaged in hunting dinosaurs. During their
adventure one of them gets killed. Then his
life begins anew. After a while, it becomes
apparent that the people can't really die, they
have some subconscious control over the
personalities (friends/enemies)end up in their
neverending relalities. At the end of the story
they discover they are nothing more than brains
kept in vats underground by "machines" that have
taken over the world. In order to protect humans
from their own limitations but keep them entertained it was decided to provide them
with the ultimate "safe" habitats but allow
them to continue living (and dying).
The question is Who wrote this book and What
was the title?
If you know, please respond to my email address.