--- Log opened Mon Dec 11 00:00:55 2023
01:33 < hprmbridge> kanzure> we are NERV
01:33 < hprmbridge> kanzure> .m https://twitter.com/elonmusk/status/1734063349313269955
01:55 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has joined #hplusroadmap
02:27 < jrayhawk> more like NERD
02:40 -!- SDr [~SDr@user/sdr] has quit [Quit: Ping timeout (120 seconds)]
02:40 -!- SDr [~SDr@li1189-192.members.linode.com] has joined #hplusroadmap
02:43 -!- muurkha_ [~kragen@adjuvant.canonical.org] has joined #hplusroadmap
02:43 -!- smiles_alot [~smiles@68-249-181-248.lightspeed.chrlnc.sbcglobal.net] has quit [Ping timeout: 264 seconds]
02:43 -!- smiles_alot [~smiles@68-249-181-248.lightspeed.chrlnc.sbcglobal.net] has joined #hplusroadmap
02:47 -!- muurkha [~kragen@adjuvant.canonical.org] has quit [Ping timeout: 276 seconds]
02:48 < fenn> so is gendo ilya or elon or someone else or all of us?
03:41 < hprmbridge> kanzure> https://diyhpl.us/wiki/transcripts/2023-03-12-twitter-eacc-grimes/ "I have never seen anyone in a position of power that doesn't feel immense pain a lot of the time. Rather than assuming that is the byproduct of anti-accelerationism or anti-optimism, I would just default to saying that's the nature of power. When one is in a position of power, like anyone working on future tech, then they should
03:41 < hprmbridge> kanzure> take that seriously and any pain is the pain of you imposing yourself on humanity."
03:48 < hprmbridge> kanzure> from the department of applied science fiction
04:00 < docl> one thing I'm still wondering is whether fiber optics need to be glassy or if there are crystalline ones. hmm, I should google that
04:02 < docl> .wik Photonic-crystal_fiber
04:02 < EmmyNoether> "Photonic-crystal fiber (PCF) is a class of optical fiber based on the properties of photonic crystals. It was first explored in 1996 at University of Bath, UK." - https://en.wikipedia.org/wiki/Photonic-crystal_fiber
04:14 < docl> .wik Photonic_crystal
04:14 < EmmyNoether> "A photonic crystal is an optical nanostructure in which the refractive index changes periodically. This affects the propagation of light in the same way that the structure of natural crystals  gives rise to X-ray diffraction and that the atomic lattices (crystal structure) [...]" - https://en.wikipedia.org/wiki/Photonic_crystal
05:58 < kanzure> fenn: just get in the damn robot shinji
06:30 -!- Jay_Dugger [~jwd@47-185-240-109.dlls.tx.frontiernet.net] has joined #hplusroadmap
06:32 < hprmbridge> kanzure>  https://cdn.discordapp.com/attachments/1064664282450628710/1183778283679142048/image0.jpg?ex=65899270&is=65771d70&hm=344f1241db89fc47f6f47567fb297921e077e83776e2d6f245632bce9f2d52f7&
07:02 -!- Gooberpatrol66 [~Gooberpat@user/gooberpatrol66] has quit [Ping timeout: 260 seconds]
07:04 -!- TMM_ [hp@amanda.tmm.cx] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
07:05 -!- TMM_ [hp@amanda.tmm.cx] has joined #hplusroadmap
07:29 < hprmbridge> svl4791> Hi, What do you think of using biological age instead of chronological age as the formal age of individuals in society?
07:30 < hprmbridge> kanzure> Government should not track ages. It's none of their business.
07:31 < L29Ah> svl4791: what's formal age?
07:32 < nsh> for formal matters
07:32 < L29Ah> like what?
07:32 < nsh> as in, what matters when you fill in a form :)
07:32 < nsh> age of majority, etc.
07:33 < nsh> as a matter of form, alcohol is not sold to people below a certain age
07:33 < nsh> that is the age, formally, when they are suitable to be sold alcohol
07:33 < L29Ah> using chronological age as rites of passage is a shortcut that doesn't work well in a diverse society due to different development in various spheres of interests in every human; i don't see why would biological age be any different in that regard
07:34 < nsh> indeed where discretion is available, standardisation to some tyranny of average should be avoided
07:34 < nsh> discretion is the capacity to decide about a margin on the basis of individual cases
07:35 < nsh> it is not as easy to enforce as a rigid rule by definition
07:35 < nsh> and so... lapsed... cultures tend towards the latter
07:35 < nsh> s/enforce/cultivate to effective sophistication/
07:36 < nsh> rules are for automatons, agents lacking in discernment and judgement
07:36 < nsh> the more a culture tends to produce the latter, the more it will tend to require the former
07:37 < nsh> discipline is very effect for an entity that requires regimentation, such as a military
07:37 < nsh> *rigid discipline, *effective
08:02 < fenn> mental "age" would be closer to what they are trying to measure for most things
08:06 < fenn> we're gonna need a more principled moral framework based on capabilities once the AI start doing things
08:34 < hprmbridge> kanzure> https://www.richardhanania.com/p/anti-surrogacy-as-a-totalitarian
08:38 < fenn> .t
08:38 < EmmyNoether> Anti-Surrogacy as a Totalitarian Dysgenics Movement
08:39 < Ashstar> a wha
08:41 < hprmbridge> kanzure> https://mistral.ai/news/mixtral-of-experts/
08:49 < jrayhawk> from the perspective of a four-year-old, a parent enforcing the rule "don't walk into the road" is a hypocrite for doing so themselves. from the perspective of the parent, the four year old is an idiot. deontology and virtues ethicism are for people who are bad at inductively predicting consequences of conditionals and maintaining internal alignment and alignment-over-time; we use age as a cheap 
08:49 < jrayhawk> legible proxy for inductive capacity and alignment, and professional licensure as an expensive legible proxy.
08:58 < fenn> i think capability metrics need to become more orthogonal and fine grained
08:59 < fenn> humans are mostly all alike; "g" correlates well with pretty much everything. robots however, won't have this property
09:00 < jrayhawk> that would be nice. it is worth pointing out, though, that biological age and onerous professional licensure might still represent pareto-optimal options on a forgability/auditability-vs-expense-vs-signal-to-noise tradeoff.
09:01 < jrayhawk> most other options involve the vaguery of social manipulation, including lawyers and accountability dispersion
09:01 < jrayhawk> which deeply underprivilege e.g. the autistic
09:04 < jrayhawk> IQ testing having become deontologically forbidden and unvirtuous in various contexts is a horrible place to find ourselves in
09:11 < jrayhawk> i am confused as to why e.g. the SATs managed to be relatively uncorrupt in comparison to, say, medical licensure.
09:13 < L29Ah> IQ testing is easy to train for, so it shouldn't be used in places where an agent can be motivated to prepare for the test
09:15 < L29Ah> (everywhere but scientific studies?)
09:16 < jrayhawk> just because it's not perfect doesn't mean it's not pareto-optimal for this application
09:17  * nsh puts all the dollar words back on the high shelf out of reach
09:19 < L29Ah> i hope you provide an example where it is optimal
09:20 < jrayhawk> At the risk of being U.S.-centric, the SATs are, in essence, a traditional IQ test that have actually served fairly well for their purpose.
09:20 < fenn> IQ testing is highly valued by militaries
09:21 < fenn> of course they don't call it that
09:22 < fenn> https://www.officialasvab.com/applicants/sample-questions/
09:26 < fenn> there are other aptitudes not listed there like "Coding Speed" which is your ability to press the right button quickly
09:34 < jrayhawk> USPS postal exams are also constantly skirting the edge of being an illegal IQ test
09:34 < L29Ah> alright, a test that evaluates how good/industriously you studied is indeed adequate to verify if you are a good candidate for further study
10:05 < jrayhawk> no, measures of conscientiousness are done by other means.
10:07 -!- Guest95 [~Guest95@2.59.157.237] has joined #hplusroadmap
10:12 < jrayhawk> https://gwern.net/doc/iq/2018-scharfen.pdf
10:12 < jrayhawk> "Retest effects in cognitive ability tests: A meta-analysis"
10:13 < jrayhawk> the training effect varies wildly depending on how you define and measure it, but it is not even intrinsically positive
10:43 -!- test_ is now known as _flood
11:50 -!- muurkha_ is now known as muurkha
12:00 -!- Gooberpatrol66 [~Gooberpat@user/gooberpatrol66] has joined #hplusroadmap
12:42 -!- flooded [flooded@gateway/vpn/protonvpn/flood/x-43489060] has joined #hplusroadmap
12:46 -!- _flood [flooded@gateway/vpn/protonvpn/flood/x-43489060] has quit [Ping timeout: 260 seconds]
13:04 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has quit [Ping timeout: 240 seconds]
13:06 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has joined #hplusroadmap
13:08 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has quit [Remote host closed the connection]
13:09 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has joined #hplusroadmap
13:13 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has quit [Remote host closed the connection]
13:15 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has joined #hplusroadmap
13:22 -!- Guest95 [~Guest95@2.59.157.237] has quit [Quit: Client closed]
13:38 -!- pongo [~pongo@108.223.46.57] has joined #hplusroadmap
13:47 -!- pongo [~pongo@108.223.46.57] has quit [Quit: Client closed]
14:06 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has quit [Remote host closed the connection]
14:06 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has joined #hplusroadmap
15:05 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has quit [Ping timeout: 240 seconds]
15:06 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has joined #hplusroadmap
15:17 -!- flooded [flooded@gateway/vpn/protonvpn/flood/x-43489060] has quit [Ping timeout: 245 seconds]
15:32 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has quit [Ping timeout: 240 seconds]
16:12 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has joined #hplusroadmap
16:36 < hprmbridge> Eli> There are 30,000 - 60,000 sperm donation births a year. If we paid people like Terrence Tao to essentially be a sperm donation clinic, it would likely have a higher research ROI than paying people like him to remain in STEM. Imagine 30,000 kids who got DNA from someone who got a STEM PhD by age 21. Ghengis Khan but voluntarily. And since the DNA is likely to be well adapted to our current laptop
16:36 < hprmbridge> Eli> class environment, it will likely stick around and we will see benefits compounded over time.
16:37 -!- juri__ [~juri@79.140.117.100] has joined #hplusroadmap
16:39 < hprmbridge> kanzure> or you can just copy the business plan for the nobel prize sperm bank and call it a day
16:40 -!- juri_ [~juri@79.140.117.38] has quit [Ping timeout: 268 seconds]
16:52 < hprmbridge> Eli> huh, I was not aware of that idea. However, I'm not sure Nobel Prize is a great metric. There are many smart Nobel Prize winners, but we can't really compare an Alexander Flemming to a Terrence Tao. In my model, a billionaire looking to get the greatest charitable ROI would pay Terrence Tao a million dollars a year to donate sperm and then sell it at a competitive price to the market rate. In
16:52 < hprmbridge> Eli> terms of poverty reduction, this seems like it would be one of the best possible investments that could be made.
16:54 < L29Ah> what does it have to do with poverty reduction?
16:54 < Ashstar> lol, I know a few brilliant professionals that have been asked for their sperm donation
16:54 < L29Ah> Tao's father, Billy Tao,[a] was a Chinese paediatrician who was born in Shanghai and earned his medical degree (MBBS) from the University of Hong Kong in 1969.[12] Tao's mother, Grace Leong,[b] was born in Hong Kong; she received a first-class honours degree in mathematics and physics at the University of Hong Kong.[10]
16:55 < L29Ah> doesn't seem like his parents were poor
16:55 < Ashstar> by women who want the best'
16:56 < hprmbridge> kanzure> is that the best plan for intelligence
16:56 < hprmbridge> Eli> @L29Ah Increasing research productivity will improve and lower costs of everything in a free market.
16:57 < hprmbridge> Eli> I'm open to being Red Teamed
16:57 < Ashstar> there is so much that makes up "intelligence" not all of them make what is term a fenancial success
16:57 < Ashstar> financial
16:58 < Ashstar> some have astounding abilities in other areas
16:58 < Ashstar> music, art
17:00 < Ashstar> also, it takes a resources to sometimes suport the best learning environment, open oportunities of learning, good nutrition to support brain developement
17:01 < Ashstar> Im sure there are many, who if they had had the resources while developing, might have done great, recognizable things
17:02 < hprmbridge> Eli> My a priori bias is that people receiving sperm donations are likely more considerate and ready to raise a child than many people who have children.
17:02 < Ashstar> Faraday was bright but uneducated, if it wasn't for Dvey's eye accident, Faraday might have bound books all his life
17:02 < L29Ah> Eli: that's very long-term, i would argue for more immediate things like lobbying for investment climate improvement and dismantling kleptocracy, but that's not my cup of tea overall
17:04 < hprmbridge> Eli> Those are super expensive tasks with low ROI, imo
17:05 < kanzure> https://scholarlykitchen.sspnet.org/2023/12/07/where-did-the-open-access-movement-go-wrong-an-interview-with-richard-poynder/
17:06 < kanzure> i think for intelligence we should rapidly breed animal populations and then port over the learnings
17:06 < kanzure> you can also start by importing human brain genes if you want things to go faster
17:06 < kanzure> and then you can take those lessons and put them in human
17:06 < kanzure> ... plus or minus seizure issues.
17:07 < hprmbridge> Eli> Which animals would be a good model?
17:07 < kanzure> most of them, the mammalian brains are fairly similar in architecture
17:08 < kanzure> also, there's a few things already known https://diyhpl.us/wiki/genetic-modifications like the CONV2 copy number variation for IQ
17:09 < hprmbridge> Eli> so basically look at the genome of human super geniuses and then gene mod monkeys and see results?
17:10 < hprmbridge> Eli> or even rats I suppose
17:10 < kanzure> well they have already looked at genius genomes, it just turns out to be ~difficult and most of the studies are lame GWAS<->polygenic correlation studies
17:13 < kanzure> also, i  think one area that has been overlooked is peripheral traits that help with intelligence like lack of mental exhaustion, resistance to sleep deprivation, higher agency, better memory, etc
17:13 < kanzure> vocabulary size is a trivial one to test for
17:13 < hprmbridge> Eli> I think I want to see super genius genomes. There is a massive difference between Lebron James and the worst athlete in the NBA. Even though the NBA contains the best 500 players in the world, the pay difference over lifetime can be more than two orders of magnitude. I suspect there are similar things going on in intelligence.
17:13 < kanzure> wordcel maxxing is a poor measure but maybe there's still some payoff
17:14 < kanzure> working memory size is also helpful even if you have low IQ
17:16 < kanzure> https://old.reddit.com/r/SteamDeck/comments/18bf9op/i_ran_llms_on_steam_decks_gpu_tutorial_in_comments/
17:17 < hprmbridge> Eli> This is definitely true. I used to be in an incubator for startups and I don't think anyone was neurotypical. The number of founders who have the sleep four hours a night gene is mostly likely higher than the normal population. There are also traits like synesthesia and high functioning aspergers that might be beneficial in that world in certain cases. But that's my a priori bias. It's similar to
17:17 < hprmbridge> Eli> how there are a greater percentage of professional musicians with perfect pitch than in the normal population.
17:22 < kanzure> short sleep is a good one yeah
17:56 -!- Malvolio is now known as IsoLogic
18:24 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has quit [Ping timeout: 252 seconds]
18:26 -!- Jay_Dugger [~jwd@47-185-240-109.dlls.tx.frontiernet.net] has quit [Ping timeout: 256 seconds]
18:52 < muurkha> it's hilarious to read someone talking about the importance of IQ and high-IQ genes who misspells both "Terence Tao" and "Genghis Khan"
18:53 < Ashstar> I think IQ's can increase
18:54 < Ashstar> with training
18:54 < muurkha> no doubt next we'll hear about how important it would have been for Einstien, Hawkins, and Feynmann to participate: https://math.ucr.edu/home/baez/crackpot.html
18:54 < Ashstar> also, we have some inaccurate assesment of what intelligence is
18:55 < Ashstar> there are various very capable geniuses, in specific fields
18:55 < Ashstar> shitty at other stuff
18:56 < muurkha> sure, like, maybe they're dyslexic, or blind
18:56 < Hooloovoo> to be fair there's a lot of ways to spell the Khan's name. an 80s disco band comes to mind
18:57 < Ashstar> yeah, or issues with certain brain regions, sparkle in others
18:58 < Ashstar> Shakespeare obviously had a well developed Brocas n Wernickes region (Auditory/speech area)
18:59 < Ashstar> prbably a shitty carpenter or mathematician
19:01 < muurkha> Hooloovoo: yes, but that's not one of them
19:02 < muurkha> Ashstar: that's a fallacy; in general the correlation between abilities in different areas is positive, not negative
19:02 < muurkha> fairly strongly so
19:02 < Hooloovoo> interesting https://en.wiktionary.org/wiki/Genghis_Khan#English
19:03 < Hooloovoo> oh wait crap that's the right spelling nevermind
19:03 < Ashstar> well, sure he was better at a lot of things as a result of his elevated capability
19:03 < Ashstar> from writing
19:03 < muurkha> maybe, we don't know
19:04 < muurkha> but he's more likely to have been an above-average carpenter or mathematician than a below-average one
19:04 < muurkha> it's not fair, but that's the way it is
19:04 < Ashstar> its just such a mystery, his life
19:04 < Ashstar> his vast knowledge of curt
19:04 < Ashstar> court
19:05 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has quit [Ping timeout: 240 seconds]
19:05 < Ashstar> things that make us wonder how a guy with elementary education could write such so[histicated stuff
19:06 < Ashstar> From Stratton upon Avon
19:06 < muurkha> *Stratford
19:06 < muurkha> he didn't die until he was 52
19:07 < Ashstar> oh well, people will be arguing about this centuries forward
19:07 < Ashstar> yeah
19:07 < Ashstar> he was young
19:07 < Ashstar> 1616
19:07 < muurkha> if you spend 52 years learning stuff you can learn a fuckload of stuff
19:07 < Ashstar> there is this missing period
19:07 < Ashstar> we know nothing of
19:08 < muurkha> most people spend their lives being mad at or worried about their family members, getting to know fictional characters, plowing fields, caring for babies and the aged
19:08 < muurkha> and only learn the minimum amount they have to learn to get by, once they reach adolescence anyway
19:09 < muurkha> a "highly educated person" is someone who spend four to ten years studying (if we don't count K-12, which we probably shouldn't)
19:09 < Ashstar> sure there were a lot of things to deal with in Elizabethan England
19:09 < Ashstar> just surviving
19:09 < muurkha> think about what happens if you spend twenty or thirty years studying
19:09 < muurkha> even with a just average capacity
19:09 < Ashstar> in high mortality world
19:11 < Ashstar> he had a lot of catching up to do, to have all the language skills to create those beautiful Iambic Pentameter
19:11 < muurkha> nowadays you can do a lot more: opencourseware, youtube, wikipedia, library genesis, arxiv, etc., mean you have access to this immense wealth of information
19:11 < Ashstar> writints
19:11 < Ashstar> writings
19:11 < Ashstar> his erudition was great
19:11 < muurkha> but there were already more books around than you could read in a lifetime even in Shakespeare's time
19:12 < Ashstar> he ha to be autodidactive
19:13 < Ashstar> unless, as y mother, who was a English prof, Shakespeaare scholar, is right n it was the 17th Earl of Oxford
19:13 < Ashstar> DeVere
19:13 < Ashstar> who knew Shakepeare
19:15 < Ashstar> was very knowledgeable about courtly life
19:16 < Ashstar> yes, after Guttenburg's 1452 printing
19:16 < Ashstar> it took off
19:16 < Ashstar> lots of publications came out
19:17 < Ashstar> no longer was a person able to only learn, unless he was part of the church
19:19 < Ashstar> that was a revolution, like industrial, nuclear, information, a paradign shift
19:32 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has joined #hplusroadmap
20:55 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has quit [Ping timeout: 240 seconds]
21:08 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has joined #hplusroadmap
21:59 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has quit [Write error: Broken pipe]
21:59 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has joined #hplusroadmap
--- Log closed Tue Dec 12 00:00:56 2023