hi.
well, its been a couple of days since my last attempted parry-thrust at eliezer's points over the questions concerning whether we should actively try to make transcendent ai. i blame this delay on a complex series of events involving all kinds of mundane things... but anyway, i think our motavational/ethical differences might very well stem from 2 premise differences. i feel longwinded today. stand clear, no step, you have been warned. anyway, here goes.
perceived premise difference 1:
sayke: nanowar* is _less_ likely to occur then human pseudo-uploading*
before the creation* of a transcendent ai*.
eliezer: nanowar is _more_ likely to occur then human pseudo-uploading before the creation of a transcendent ai.
where:
nanowar = red goo; not caused by a transcendent ai. will kill me/us
almost certainly.
human pseudo-uploading = nondestructive, probably nanotech-assisted, personal transcendence. method? whatever method turns out to be doable, if any... will increase the odds that i/we live* for a very long time.
creation = this is accidental creation; that is to say, without our (mine or eliezer's or other people reading this) help, and even in spite of our efforts to slow it down and do it in a sanitary ai playground first. this means somebody will do it with their basement beowulf cluster, or that commercial ai will quietly transcend without anybody's help. unofficial creation; beyond-our-control creation; whatever...
transcendent ai = big, selfevolving, badass, ai. will probably (please, other plausible options?) either eat the solar system, including me/us, or leave me/us alone and take off into space, or transcendify me/us into badassness, in that order of plausibility, with several orders of magnitude between each likelihood...
live = to continue personal existence; involves discussion of functional solipsism, enlightened self-interest and other ethical/motavational concerns. ill get back to this; why trying to live is good is what my second premise difference is about. shit, this setup feels clumsy...
alrighty. i would first (first?) like to say that, quite frankly, i made the mistake of being on a trend towards equating the odds concerning nanowar with the odds concerning more conventional wars of mass destruction. i thought nanowar is fairly unlikely, because of fairly conventional mutually-assured-destruction type arguments, and because i was underestimating the potential that comes with putting accidents, suicidal people, and nanotech all into your bong, at once, and lighting up. i hereby suspend that opinion, because i really think i need to get the some input on this.
however, my main argument in support of my premise would be that, by the time nanotech is advanced and common enough to make in a nanowar likely, nanotech would be advanced enough to transcendify yours truly (aka me/us). it seems to me that nanowar is unlikely to be started by the intentional act of a Large Power, for MAD-ish reasons. it basically seems to me that nanowar would be probably be started either by accident, or by Evil Genius, and personal ascension could very well happen before this. comments? does this hold?
alright... on to the second premise.
perceived premise difference 2:
sayke: functionally*, significance* _ceases_ to exist if i* dont exist*
to label it as such.
eliezer: functionally, significance _continues_ to exist if i dont exist to label it as such.
where:
functionally = for all practical purposes, and from my individual point
of view, and for use in an ethical/motavational system.
significance = motivation points. value... etc... i dont think i need to try to define this with terrible rigor, because our definitions of significance are probably pretty similar... threads here have probably been based around discussion of significance... if you wanna give me pointers to a relevant past thread, or if you wanna start another one about this, go for it...
i = the point of view that i am... that which bestows significance. there have been threads about this. ive seen em. again, if you wanna give me pointers to a relevant past thread, or if you wanna start another one about this, go for it...
dont exist: death is presumed to be the end of personal existence...
by definition im right ;) but, i dont think ill be let off that easy. still... it seems to me that, well, by definition, im right. all significance i can interact with goes away when i do, right? is that not exactly the same, functionally, as saying "all significance, period, goes away when i do"? if not, whats the difference? or am i just missing something obvious that i really should know about...? is it just my turn to do the "im wrong" dance? hehe...
next episode, on extropy-l: does sayke do the "im wrong" dance? or will he go postal on a herd of innocent churchgoers with a brace of flare guns? catch extropy-l, same time, next week, only on channel 23; 'your source for nifty shit (tm).' brought to you by JOE'S EATS. 'hungry? eat at joe's! (tm)'...
alright, im tired. may you live in interesting times!
sayke, v2.3.05