--- Log opened Wed Jul 12 00:00:56 2023
00:01 < fenn> Andrew: any continuously variable trait (not discrete values) will be governed by many genes, and probably also the environment
00:03 < fenn> i don't agree that a large number of eggs means you don't have to do engineering. there's still only so much variation, and most of it would be deleterious. you won't know ahead of time whether the random mutation is good or bad, but it's probably bad
00:04 < fenn> there's an impossibly huge gap between "a very large number" and "infinite"
00:05 < fenn> the idea behind embryo selection is to pick from known good genes that are already present
00:09 < fenn> without developing a much more comprehensive understanding of human development, i don't see how to introduce new variants without someone getting hurt. maybe finding out what does and doesn't work is just the price of progress
00:10 < fenn> sure we can do tests on mice and chimps, but then you end up with super-mice and super-chimps which are moral patients just as much as humans, but then the results don't necessarily translate to enhancing humans
00:52 -!- TMM_ [hp@amanda.tmm.cx] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
00:53 -!- TMM_ [hp@amanda.tmm.cx] has joined #hplusroadmap
01:54 -!- test_ [flooded@gateway/vpn/protonvpn/flood/x-43489060] has joined #hplusroadmap
01:56 -!- test__ [flooded@gateway/vpn/protonvpn/flood/x-43489060] has joined #hplusroadmap
01:58 -!- flooded [flooded@gateway/vpn/protonvpn/flood/x-43489060] has quit [Ping timeout: 260 seconds]
01:59 -!- flooded [flooded@gateway/vpn/protonvpn/flood/x-43489060] has joined #hplusroadmap
02:00 -!- test_ [flooded@gateway/vpn/protonvpn/flood/x-43489060] has quit [Ping timeout: 245 seconds]
02:02 -!- test__ [flooded@gateway/vpn/protonvpn/flood/x-43489060] has quit [Ping timeout: 240 seconds]
05:44 -!- yashgaroth [~ffffffff@2601:5c4:c780:6aa0:31a9:955b:cd6d:647b] has joined #hplusroadmap
06:07 < kanzure> "Longevity factor klotho enhances cognition in aged nonhuman primates" https://www.nature.com/articles/s43587-023-00441-x
06:10 < kanzure> lesswrong tries to figure out if the word "consciousness" actually has any utility https://www.lesswrong.com/posts/KpD2fJa6zo8o2MBxg/consciousness-as-a-conflationary-alliance-term
06:55 < jrayhawk> whatever they replace it with will also be gerrymandered by the same motivated thinking. a similar semantic treadmill effect as with banning the word "retarded".
06:57 < jrayhawk> something equivalent to "The IAU nonsensically and arbitrarily gerrymandered the concept of planet. I guess the term 'planet' is useless now."
06:58 < kanzure> jrayhawk: what is the appropriate response to "but then you end up with super-mice and super-chimps which are moral patients just as much as humans" 
06:58 < jrayhawk> "good"
06:59 < nsh> first you have to butcher an animal to the point at which you consider it sufficiently sentient to be ethical towards, then you ask it for retrospective informed consent to the aforementioned
06:59 < kanzure> jrayhawk: large scale factory farming is actually good though, even if you consider chickens to have high moral value
07:01 < nsh> oh it's actually good. glad that turned out to be such a trivial calculation 
07:01 < nsh> now to colonising
07:01 < nsh> *do
07:02 < nsh> god save us from the confidence of men 
07:02 < kanzure> space colonization seems to be a trivially obvious thing we should be doing
07:04 < nsh> i meant the colonisation that created the conditions whereby y'all associate colonisation with space
07:04 < nsh> :)
07:04 < nsh> but it doesn't matter. i'm speaking to the wings
07:04 < kanzure> like "everything is just a giant complex trolley problem and therefore we shouldn't do anything until someone stops the trolleys"?
07:04 < jrayhawk> Factory farmed animals are bred for docility; literally the opposite of the three components of "consciousness" people care about (induction, valuation, EV-maximizing action)
07:06 < nsh> thousands of years of philosophy to bring you the superlative conclusion that "well it doesn't really matter because you know, it's like, kinda wuzzy isn't it?"
07:07 < nsh> at least civilisation still fall
07:07 < nsh> +s
07:07 < nsh> having it seems inexorably reached the point of decline
07:11 < kanzure> jrayhawk: i mean to ask for a conceptual framework that lets people do stuff even in the presence of the accelerating expansion of the moral value universe
07:13 < nsh> what is the space in which the moral value universe is expanding?
07:13 < nsh> or is the metric itself expanding. strange way to arrange words :)
07:15 < nsh> what is the morality of an intelligent species that emerges from crystal lattice arrangements, or exists within the riemann zeta function?
07:15 < kanzure> there is a general increase in quality of life afforded by technology, likewise a general increase and expansion for our general moral concern or moral quality of universal life, but that capacity doesn't seem to be infinite- we have to actually work on technology and build things and exert energy and use our available tools
07:16  * nsh nods
07:16 < nsh> indeed the scope within which to act so as to result in better or worse quality of experience is a function of various things that give us agency in the world
07:17 < nsh> though i'd not be so flattering as to say we've resulted in a net increase
07:17 < nsh> quality of human life over the last few hundred years has improved at a terrible cost for the rest of the ecosystem which nutures us
07:17 < nsh> *nurtures
07:18 < nsh> and quality of the lives of certain humans is a direct result of the exploitation of others 
07:18 < kanzure> i forget, are you arguing against large-scale chicken farming and/or against humans presently making genetically modified superintelligent chimpanzees?
07:18 < nsh> just dullness and conceit :)
07:18 < nsh> symptoms are secondary to the syndromes
07:18 < kanzure> you as much as others should know the navigation of the landscape is exploration vs exploitation search algorithm tradeoff. i don't recall there being a third choice.
07:19 < nsh> i think algorithms are about algorithms and analogies can be extended beyond their useful remit :)
07:20 < nsh> fish have algorithms about as much as jelly has a carburetor
07:20 < nsh> humans used to have codes
07:20 < jrayhawk> nsh: "expansion" here is the expansion of both the scope of the concept of moral agency beyond humans due to advanced in bioengineering (brain organoids, uplift) or electrical engineering (AI), and the expansion scope of the stakes involved due to advances in resource management.
07:21 < nsh> but that's a notion that predates and is not especially gainfully employed by later concepts involving digital electronics
07:21 < nsh> *enriched
07:21 < nsh> i'm aware that you've dined extensively on a hodgepodge of waffle but i'm just inclined to humour it :)
07:21 < nsh> it won't be discussed in a hundred years 
07:21 < nsh> *not 
07:22 < kanzure> nsh: your fish don't seem to have a reason to have neurons; i would argue they do have neurons because the neurons are more useful than not having neurons. otherwise it would be metabolically deadweight shedded a long time ago in favor of other processes.
07:22 < nsh> and a rock tumbles in one direction down a mountain because was more useful than the others, i suppose
07:23 < nsh> the dice found better use for falling on a three and four
07:23 < kanzure> i thought the rock tumbling down a hill is precisely a good example of kinetic energy landscape and gravity..? or am i mistaken?
07:23 < nsh> it can be modelled thus
07:24 < nsh> but one is rarely teleological about it
07:24 < nsh> and we can imagine different starting conditions and a different gravitational landscape
07:24 < nsh> it's hard (as yet) to imagine difference scenarios of evolution 
07:24 < nsh> *different
07:24 < kanzure> you've lost me.
07:24 < nsh> except where we can observe different relatively isolated theatres
07:25 < nsh> it doesn't matter
07:25 < nsh> it's just an understanding evolution that's archaic at this point
07:25 < nsh> +of
07:25 < kanzure> jrayhawk: in fact, i would argue the expansion even happens without bioengineering and AI; in general it seems that as human quality of life increased, we could expand our scope of moral concern beyond immediate shelter-level survival.
07:25 < nsh> and was a caricature at best
07:26 < nsh> people who lived in caves hundreds of thousands of years ago were engage in art
07:26 < nsh> *engaging 
07:26 < nsh> and song and story
07:27 < kanzure> caves are shelte ryes
07:27 < kanzure> caves are shelter yes
07:27 < kanzure> oh look at you with the double wide cave. i bet you have more than the average 2.3 rocks too.
07:37 < jrayhawk> kanzure: morality having no nash equilibria/being anti-inductive means moral structures tend to get more complicated over time, yes. one of the games being played is letting always-cooperators espouse and constrain themselves by principled systems of morality which privileges always-cooperating. Fundamentally such systems are incompatible with biology (exponential growth, competition over shared 
07:37 < jrayhawk> resources) and exists only insofar as always-cooperators remain useful to agents capable of playing the prisoner's dilemmas at the highest levels.
07:38 < jrayhawk> The binary classification of "consciousness" and gerrymandering is part of this game.
07:39 < nsh> sentience has degrees and kinds
07:39 < kanzure> when you say "and exists only insofar as always-cooperators remain useful to [other] agents" what does that mean, what if they stop being useful to those agents- what happens
07:40 < jrayhawk> agreed; making it a binary classification is very, very political and not based meaningful taxonometrics
07:44 < jrayhawk> kanzure: Any utility reallocation you care to name. Resource allocation, cost allocation, information withholding, deception, communication denial, violent coercion, murder...
07:46 < kanzure> in the exploration/exploitation dichotomy, actors in both sets have moral frameworks, and it is sort of annoying that actors in both sets claim the actors in the other set are unprincipled
07:48 < kanzure> well, i am probably mixing too many different systems there. exploration/exploitation is separate from what kind of game the actors think they are playing (you can get always-cooperators in exploitation like exploit-the-status-quo-and-keep-things-here but also in exploration)
07:59 < L29Ah> omg these philosophers
07:59 < jrayhawk> we should all be banned
08:00 < kanzure> probably
08:08 < jrayhawk> one of the funny ways people try to make binary classification seem reasonable is by putting the boundary on some nonlinear jump in capabilities, like self-attention ("introspection", "perception of perception", "awareness of awareness", "memory of memory")
08:08 < jrayhawk> but there are a lot of such jumps and picking any one is just a matter of political convenience
08:09 < jrayhawk> pain is another one
08:13 < jrayhawk> i wonder if there's some trollish relatively-high-iq jump i can advocate for classifying around as a parody
08:14 < jrayhawk> counterfactual induction would be ~ 80 IQ, but that doesn't seem high enough
09:32 < docl> seems like language is the main basis used for biological sapients. we don't treat nonverbal humans as outside the circle of concern, but there aren't very many of them (other than children who will later learn it) so the political pressure is not aligned against that (if you treat this as political for parody purposes, I mean)
09:51 < docl> breadth-depth tradeoffs are an interesting lens for understanding intelligence. the striking thing about smart people is that they tend to have a lot of breadth *and* depth. they use breadth to discover novel specializations to pursue, and then use generalization techniques to extend breadth in them once discovered. there are tradeoffs still, and I suspect g does mean something like more available 
09:51 < docl> compute (an indicator that evolution found a way to do something more efficiently -- although repurposing irrelevant/replaceable specialized stuff can account for this too). but I guess transitioning back and forth introduces something nonlinear, which would explain how humans can do some things a dog just couldn't learn to do.
09:57 < docl> but a lot of human vs dog is explained as humans have a well developed symbol recognition package directly in our hardware, vs dogs devote vast amounts of brain tissue to processing smells. thing is, symbol recognition at our level creates room for generalization capacity to grant an evolutionary advantage, whereas smell recognition does not
10:05 < nsh> (people still take IQ seriously?)
10:09 < docl> not for political purposes, of course
11:03 -!- ANACHRON [~Malvolio@idlerpg/player/Malvolio] has quit [Quit: i hate myself and i want to die]
12:59 -!- popeye [~jaydu@115.89.23.44] has joined #hplusroadmap
12:59 < popeye> Hello, everyone.
13:00 < hprmbridge> kanzure> greetings
13:00 -!- popeye [~jaydu@115.89.23.44] has left #hplusroadmap []
13:02 -!- Jay_Dugger [~jaydu@115.89.23.44] has joined #hplusroadmap
14:31 -!- test_ [flooded@gateway/vpn/protonvpn/flood/x-43489060] has joined #hplusroadmap
14:31 -!- stipa_ [~stipa@user/stipa] has joined #hplusroadmap
14:32 -!- stipa [~stipa@user/stipa] has quit [Read error: Connection reset by peer]
14:32 -!- stipa_ is now known as stipa
14:33 -!- Jay_Dugger [~jaydu@115.89.23.44] has quit [Ping timeout: 240 seconds]
14:34 -!- flooded [flooded@gateway/vpn/protonvpn/flood/x-43489060] has quit [Ping timeout: 245 seconds]
14:44 -!- o-90 [~o-90@gateway/tor-sasl/o-90] has joined #hplusroadmap
14:47 < hprmbridge> jaisel> hey
14:49 -!- o-90 [~o-90@gateway/tor-sasl/o-90] has quit [Remote host closed the connection]
15:29 -!- superkuh [~superkuh@user/superkuh] has quit [Read error: Connection reset by peer]
15:33 -!- superkuh [~superkuh@user/superkuh] has joined #hplusroadmap
16:25 -!- potatope_ [sid139423@id-139423.lymington.irccloud.com] has joined #hplusroadmap
16:28 -!- berndj-blackout [~berndj@197.189.254.139] has joined #hplusroadmap
16:32 -!- Netsplit *.net <-> *.split quits: berndj, heath2, potatope
16:32 -!- potatope_ is now known as potatope
16:33 -!- berndj-blackout is now known as berndj
16:37 -!- TMM_ [hp@amanda.tmm.cx] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
16:38 -!- TMM_ [hp@amanda.tmm.cx] has joined #hplusroadmap
17:12 < kanzure> https://claude.ai/ seems to be working now.
17:25 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has quit [Ping timeout: 246 seconds]
17:27 -!- Netsplit over, joins: heath2
18:54 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has left #hplusroadmap []
19:11 -!- yashgaroth [~ffffffff@2601:5c4:c780:6aa0:31a9:955b:cd6d:647b] has quit [Quit: Leaving]
19:12 -!- gptpaste [~x@yoke.ch0wn.org] has quit [Ping timeout: 245 seconds]
20:47 < jrayhawk> A modern example of "always-cooperators" getting well and truly screwed once high-level players run out of uses for them is China's "One Child" policy.
20:48 < jrayhawk> Mind you this turned out to be a huge mistake in several regards, but the principle remains.
20:49 -!- flooded [flooded@gateway/vpn/protonvpn/flood/x-43489060] has joined #hplusroadmap
20:52 -!- test_ [flooded@gateway/vpn/protonvpn/flood/x-43489060] has quit [Ping timeout: 240 seconds]
20:57 < jrayhawk> Closer to home, the pharmaco-regulatory-academia-media complex comes to mind.
21:01 < jrayhawk> plus medical-
21:12 < muurkha> what do any of these have to do with "always-cooperators"?
21:13 < muurkha> just because you don't devote your life to smuggling illegal drugs are you an "always-cooperator"?
21:31 < hprmbridge> nmz787> This message brought to you by Tesla Auto steer
22:07 -!- Gooberpatrol66 [~Gooberpat@user/gooberpatrol66] has joined #hplusroadmap
22:18 -!- TMM_ [hp@amanda.tmm.cx] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
22:18 -!- TMM_ [hp@amanda.tmm.cx] has joined #hplusroadmap
22:49 -!- deltab [~deltab@user/deltab] has quit [Ping timeout: 245 seconds]
22:53 -!- deltab [~deltab@user/deltab] has joined #hplusroadmap
23:32 -!- Croran_ [~Croran@user/Croran] has joined #hplusroadmap
23:32 -!- Croran [~Croran@user/Croran] has quit [Ping timeout: 252 seconds]
23:48 -!- ANACHRON [~Malvolio@idlerpg/player/Malvolio] has joined #hplusroadmap
--- Log closed Thu Jul 13 00:00:57 2023