![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
herewiss13More for the sake of its historical significance than anything else, I've started reading 'Triplanetary', the first novel in E.E. Doc Smith's 'Lensman' series. The writing, as one might expect, is equal parts fairly decent and bloody awful. Smith should never allow his men to speak with his women. Still, I'm pretty much able to hold my nose and wade on through without much trouble. Parts of it are tremendously impressive, especially given that it was written prior to WWII. Of course, the flip side of that is he's attempting to portray an essentially computer-less future, which is archaic and continually jars.
The most amusing thing is how technical language has changed over the intervening decades. One of the chief weapons in the Good Guys' arsenal is called a "dirigible atomic torpedo." The first few times I ran across that phrase I was stymied as to why you'd combine blimps with rockets, ungainly to say the least. Then it struck me that dirigible originally meant steer-able, before it was applied to lighter-than-air craft with a rigid skeleton. So Smith is essentially describing a guided nuclear missle...impressive feat for 1934.
Maybe there's a reason he's known as the Father of Space Opera after all. ;-)
Just started reading "Tinker" by Wen Spencer, an "Elfin Borderland" type story which is a wonderful blend of magic and science, with more emphasis on the technology than most works in the genre. One idea got me thinking:
Spencer's Elves are immortal. It's theorized that they wove a spell into their DNA at some point in the distant past which continually rejuvenates their cells. As a side-effect, this also allows them to heal very quickly.
It occurs to me that this isn't really all that different than nano-technology. The name of the game is redundancy and fidelity. DNA is double stranded so that information isn't corrupted if one string of code is damaged. Senescence occurs when enough point-errors accumulate that continued cell growth stops. Tissues succumb to environment wear-and-tear and we generally grind to a halt. The cell has its own error-checking functions, but they aren't perfect in the long-run (obviously).
A simple ("simple," he says!) system of nanites ought to be able to pick up the slack, combing an individual's many genome copies for unwanted mid-life mutations. The one problem here is that nanites ought to be just as vulnerable to point-errors as the cells they're checking. You've just pushed the senescence problem back a level and introducted the potential for nanotech cancer. Fun!
The thing you can do with nanites that you can't do with the cells of your body is to get them to check _each_other_. Every time three nanites meet, they check each other. If one isn't identical, the other two disassemble it. If none of them match they all self destruct (they're self-assembling so "pure" nanites will take up the slack and reproduce replacements).
Get millions of these circulating through the body, spot-checking nucleuses, and no coding error, mechanical or biological, should be able to survive for long. The one concern would be the "evolution" of the nanites as they exist within the body. Point errors might create a faulty nanite that resisted disassembly, or lacked a self-destruct protocol. So there'd need to be a collective "immune" system to police for such deviants, and it'd need to be pretty clever (and probably adaptive too), but if controls are strict enough from the beginning, the problem shouldn't run rampant. You might design the nanites with some built-in hard limitation, like an physical inability for the nanite to exist more than N days or penetrate more than N to the Nth cells. Some time-delay mechanism like telomeres. It's an irony that to prevent death in us, we'd have to create it in our saviors.
The other serious question, which I've just alluded to, is "how long should individual nanites last?" Long generations leave a lot of time for errors within the nanite to accumulate, while short generations increase the number of copies and the risk of transcription errors. Given that there are other safeguards against accumulated error, but no way to ensure perfect facsimiles 100% of the time, we'd probably want to go with relatively longer generations. Perhaps even some sub-program so that reproduction is only triggered after the nanite has been in a spot-check and disassembled a faulty peer. Which brings up the specter of natural selection again.
It's really quite interesting to figure out how to _prevent_ evolution from taking place and very hard. Every road block you put up gives an added advantage to any individual who manages to slip by it...and no road block is perfect. At the same time, you don't want the system to stagnate either. What you've essentially done with the system is merge two "people" together so their "dna" can keep tabs on each other.
In the end, I think you'd have to trust to the uniformity and massive redundancy of the original nanite population to ensure continuity. For added immortality, you could create anti-toxin and wound-repair subroutines, but every piece of added complexity increases the nanite's flexibility and gives it more mutational options for circumventing its safeguards. A completely separate system would probably be better...a strictly anti-injury one _monitored_ by the anti-aging one. Being less rooted in the "infrastructure", as it were, it can do less damage if things go haywire.
You could still die from sudden, traumatic injury, but I'm not sure there are any really _trustworthy_ ways to protect yourself from falling pianos without opening up risks in your other systems. The more technology you integrate into yourself, the more that can fail.
...and this is what happens after I read a single throw-away line while riding the bus. It's a wonder I finish any books at all!
Next up: a meditation on emergent properties and levels of explanation...unless I get distracted by something else.
The most amusing thing is how technical language has changed over the intervening decades. One of the chief weapons in the Good Guys' arsenal is called a "dirigible atomic torpedo." The first few times I ran across that phrase I was stymied as to why you'd combine blimps with rockets, ungainly to say the least. Then it struck me that dirigible originally meant steer-able, before it was applied to lighter-than-air craft with a rigid skeleton. So Smith is essentially describing a guided nuclear missle...impressive feat for 1934.
Maybe there's a reason he's known as the Father of Space Opera after all. ;-)
Just started reading "Tinker" by Wen Spencer, an "Elfin Borderland" type story which is a wonderful blend of magic and science, with more emphasis on the technology than most works in the genre. One idea got me thinking:
Spencer's Elves are immortal. It's theorized that they wove a spell into their DNA at some point in the distant past which continually rejuvenates their cells. As a side-effect, this also allows them to heal very quickly.
It occurs to me that this isn't really all that different than nano-technology. The name of the game is redundancy and fidelity. DNA is double stranded so that information isn't corrupted if one string of code is damaged. Senescence occurs when enough point-errors accumulate that continued cell growth stops. Tissues succumb to environment wear-and-tear and we generally grind to a halt. The cell has its own error-checking functions, but they aren't perfect in the long-run (obviously).
A simple ("simple," he says!) system of nanites ought to be able to pick up the slack, combing an individual's many genome copies for unwanted mid-life mutations. The one problem here is that nanites ought to be just as vulnerable to point-errors as the cells they're checking. You've just pushed the senescence problem back a level and introducted the potential for nanotech cancer. Fun!
The thing you can do with nanites that you can't do with the cells of your body is to get them to check _each_other_. Every time three nanites meet, they check each other. If one isn't identical, the other two disassemble it. If none of them match they all self destruct (they're self-assembling so "pure" nanites will take up the slack and reproduce replacements).
Get millions of these circulating through the body, spot-checking nucleuses, and no coding error, mechanical or biological, should be able to survive for long. The one concern would be the "evolution" of the nanites as they exist within the body. Point errors might create a faulty nanite that resisted disassembly, or lacked a self-destruct protocol. So there'd need to be a collective "immune" system to police for such deviants, and it'd need to be pretty clever (and probably adaptive too), but if controls are strict enough from the beginning, the problem shouldn't run rampant. You might design the nanites with some built-in hard limitation, like an physical inability for the nanite to exist more than N days or penetrate more than N to the Nth cells. Some time-delay mechanism like telomeres. It's an irony that to prevent death in us, we'd have to create it in our saviors.
The other serious question, which I've just alluded to, is "how long should individual nanites last?" Long generations leave a lot of time for errors within the nanite to accumulate, while short generations increase the number of copies and the risk of transcription errors. Given that there are other safeguards against accumulated error, but no way to ensure perfect facsimiles 100% of the time, we'd probably want to go with relatively longer generations. Perhaps even some sub-program so that reproduction is only triggered after the nanite has been in a spot-check and disassembled a faulty peer. Which brings up the specter of natural selection again.
It's really quite interesting to figure out how to _prevent_ evolution from taking place and very hard. Every road block you put up gives an added advantage to any individual who manages to slip by it...and no road block is perfect. At the same time, you don't want the system to stagnate either. What you've essentially done with the system is merge two "people" together so their "dna" can keep tabs on each other.
In the end, I think you'd have to trust to the uniformity and massive redundancy of the original nanite population to ensure continuity. For added immortality, you could create anti-toxin and wound-repair subroutines, but every piece of added complexity increases the nanite's flexibility and gives it more mutational options for circumventing its safeguards. A completely separate system would probably be better...a strictly anti-injury one _monitored_ by the anti-aging one. Being less rooted in the "infrastructure", as it were, it can do less damage if things go haywire.
You could still die from sudden, traumatic injury, but I'm not sure there are any really _trustworthy_ ways to protect yourself from falling pianos without opening up risks in your other systems. The more technology you integrate into yourself, the more that can fail.
...and this is what happens after I read a single throw-away line while riding the bus. It's a wonder I finish any books at all!
Next up: a meditation on emergent properties and levels of explanation...unless I get distracted by something else.
