--- Log opened Thu Jan 24 00:00:29 2019 00:05 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has joined ##hplusroadmap 01:23 -!- TC [~talinck@cpe-174-97-113-184.cinci.res.rr.com] has joined ##hplusroadmap 01:27 -!- hehelleshin [~talinck@cpe-174-97-113-184.cinci.res.rr.com] has quit [Ping timeout: 250 seconds] 01:40 -!- spaceangel [~spaceange@ip-86-49-16-65.net.upcbroadband.cz] has joined ##hplusroadmap 01:59 -!- ebowden_ [~ebowden@unaffiliated/ebowden] has quit [Ping timeout: 244 seconds] 02:17 -!- Gurkenglas [~Gurkengla@unaffiliated/gurkenglas] has joined ##hplusroadmap 02:44 -!- pasky_ [~pasky@nikam.ms.mff.cuni.cz] has quit [Ping timeout: 252 seconds] 02:44 -!- pasky_ [~pasky@nikam.ms.mff.cuni.cz] has joined ##hplusroadmap 03:25 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has quit [Ping timeout: 240 seconds] 03:31 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has joined ##hplusroadmap 03:33 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has quit [Client Quit] 03:34 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has joined ##hplusroadmap 03:38 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has quit [Client Quit] 03:40 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has joined ##hplusroadmap 03:44 -!- sachy [~sachy@78.108.102.220] has joined ##hplusroadmap 05:08 -!- aeiousom1thing [~aeiousome@unaffiliated/aeiousomething] has joined ##hplusroadmap 05:10 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has quit [Ping timeout: 240 seconds] 05:28 -!- aeiousom1thing [~aeiousome@unaffiliated/aeiousomething] has quit [Ping timeout: 268 seconds] 05:30 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has joined ##hplusroadmap 05:33 -!- Gurkenglas [~Gurkengla@unaffiliated/gurkenglas] has quit [Ping timeout: 245 seconds] 06:03 < kanzure> .tw https://twitter.com/Be_seixas01/status/1088304715790639105 06:03 < yoleaux> @antonioregalado "Hey, I created babies who are immune to AIDS!" Jennifer Doudna: *feels sick* (@Be_seixas01, in reply to tw:1088255352213307394) 06:28 -!- Urchin [~urchin@unaffiliated/urchin] has quit [Ping timeout: 246 seconds] 07:19 -!- aeiousom1thing [~aeiousome@unaffiliated/aeiousomething] has joined ##hplusroadmap 07:22 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has quit [Ping timeout: 245 seconds] 07:30 -!- ebowden_ [~ebowden@unaffiliated/ebowden] has joined ##hplusroadmap 07:56 < kanzure> .title https://futurism.com/gene-edited-baby-due 07:56 < yoleaux> Next Gene-Edited Baby Due in Six Months, Scientist Says 08:06 -!- strages [uid11297@gateway/web/irccloud.com/x-zugolbqzhtchwdbk] has joined ##hplusroadmap 08:11 -!- aeiousom1thing [~aeiousome@unaffiliated/aeiousomething] has quit [Ping timeout: 240 seconds] 08:15 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has joined ##hplusroadmap 08:17 -!- aeiousom1thing [~aeiousome@unaffiliated/aeiousomething] has joined ##hplusroadmap 08:20 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has quit [Ping timeout: 250 seconds] 08:44 -!- l_wl [~l_wl@pool-173-66-183-216.washdc.fios.verizon.net] has joined ##hplusroadmap 08:58 < kanzure> https://www.technologyreview.com/s/612774/we-wont-use-crispr-to-make-super-smart-babiesbut-only-because-we-cant/ 09:00 < kanzure> "Gene experts speculate that our worst gene-editing fears won’t come true because they are too complex for us to pull them off." 09:00 < kanzure> what a terrible lack of imagination 09:01 -!- yashgaroth [~yashgarot@2606:6000:c308:f700:201a:dc66:3dae:4474] has joined ##hplusroadmap 09:57 < kanzure> they think intelligence modification *requires* changing those "polygenic traits". 09:57 < kanzure> imagine being that dumb 10:02 -!- aeiousom1thing [~aeiousome@unaffiliated/aeiousomething] has quit [Ping timeout: 268 seconds] 10:07 < nsh> what is the better way of looking at it, kanzure? 10:13 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has joined ##hplusroadmap 10:14 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has quit [Client Quit] 10:16 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has joined ##hplusroadmap 10:20 < kanzure> nsh: there are many genetic methods for targeting synapses, axons, neurons, receptors, receptor modifications, receptor density, receptor binding affinity coefficient, reuptake tweaks, hippocampus size anomalies, cell density. 10:20 < kanzure> nsh: they think you have to look across the whole population and find a magic silver bullet mutation that makes some smarter, this is the church of the polygenic scoring method 10:20 < nsh> a search space which we will explore sequentially or in parallel? 10:20 < nsh> and to what extent that latter and not the former 10:20 < nsh> because otherwise you're still imagining a silver bullet 10:21 < kanzure> nsh: it's very similar to height. they think you need to know 500 different genes that influence height. but instead you just need an upregulator or promoter to make certain bone cells grow more, which is admittedly non-natural but nobody cares. 10:21 < nsh> where the silver bullet is fumbing around and finding the complex combination sooner rather than later 10:21 < nsh> i would venture 10:21 < nsh> that it is not very similar to height :) 10:21 < nsh> or rather it is perhaps in the way you are imagining and not in other ways, that i might be imagining 10:22 < nsh> height is far more limited and continent on other desiderata 10:22 < kanzure> height is assumed by geneticists to be controlled by >500 genes and polymorphisms, and in a certain sense it is..... but it also isn't. 10:22 < nsh> such as, does your organism require motility? 10:22 < nsh> then it will not have roots and grow as tall as a giant redwood 10:22 < nsh> these constrains for intelligence are far more subtle 10:22 < nsh> *constraints 10:23 < nsh> intelligence is not even necessarily inherent in the elaboration of the genome in the developmental and homeostatic dynamics of the organism 10:23 < kanzure> we also know about a human mutation that confers 19% improvement in working memory capacity (granted, this was an association study not a direct cause-effect known thing) 10:23 < nsh> intelligence is something the universe does. 10:23 < nsh> i do not know how to parse such a sentence 10:23 < nsh> what is the sample size? what was the variation in the "percentage improvement in working memory capacity" 10:23 < nsh> this seems like a reductive model :) 10:24 < nsh> statements like this should be taken as unliterally as possible while retaining from them some nonzero utility 10:24 < nsh> but you have to sail very close to the wind 10:25 < nsh> 'working memory' here is just a euphemism for success defined relative to some test or task into which was projected huge basketloads of human interpretations 10:25 < nsh> the rats or monkeys or fruitflies might not have shared those interpretations 10:25 < kanzure> WWC1 rs17070145 10:25 < kanzure> see "working memory" http://diyhpl.us/wiki/genetic-modifications/ 10:25 < nsh> at least height is relatively objectively measureable 10:25 < nsh> there is no SI unit of working memory 10:25 < nsh> there will never be an SI unit of intelligence 10:26 < kanzure> i don't care. i'm much more interested in memory capabilities i think. 10:26 < nsh> well 10:27 < nsh> if we want to select for traits we had better at least convince ourselves that we're measuring them competently 10:27 < nsh> and have some way of measuring the non-selected-for traits affected by the selection too 10:27 < nsh> i'm not trying to be a party pooper, just to instil a more appropriate attitude of relative nescience :) 10:28 < nsh> i think that within my model in which i have assumed many things without examination that i have produced a measureable and repeatable difference in a trait as i have defined it, provincially, and imperfectly 10:28 < nsh> and i have failed to capture more than a tiny fraction of the consequences of that variation on the rest of the organism 10:28 < nsh> this is good starting point for such experimenters 10:28 < nsh> and even that is leaving a lot to be desired 10:29 < kanzure> okay well you go make love to "something the universe does" and i'll get back to work over here.. 10:30 < kanzure> also i'm under the belief that better dopamine receptors are a useful and good thing to deploy 10:30 < kanzure> biology doesn't know how to magically solve for binding affinity, which is why labs are so easily able to make receptors and binding assays that have 1000x improved binding 10:30 * nsh smiles 10:31 < nsh> biology might know but have elected not to maximise it's investment in that optimisation as we define it to be optimal 10:31 < nsh> because biology is playing the game on many more levels than we are, currently 10:31 < kanzure> biology doesn't elect. i think you are ignorant. 10:31 < nsh> (despite the fact that we are biology) 10:31 < nsh> you're missing the point perhaps 10:31 < nsh> this can be very rigorously defined 10:32 < nsh> for example the multiplicative weights update mechanism 10:32 < nsh> as an evolutionary metaalgorithm 10:32 < nsh> will not ever try to singularly optimise a binding affinity 10:32 < nsh> because that is not the game it is playing 10:32 < nsh> if i call it elective, i'm just using language 10:32 < nsh> that is humanistic 10:32 < nsh> to discuss a mathematical property 10:32 < kanzure> http://diyhpl.us/~bryan/papers2/neuro/human_chimpanzee_brain_differences.png 10:32 < nsh> which is fine as long as people understand :) 10:35 < nsh> nice 10:36 < nsh> but the gaps between humans are almost certainly much larger than the gaps between the average human and the average chimp 10:36 < nsh> the vista expands 10:37 < kanzure> what exactly is immeasurable about memory in your view? 10:37 < nsh> well, it all starts with how you define a memory 10:37 < nsh> which we don't know how to do with 100 confidence that it's the right definition 10:38 < nsh> you can elect, again, to define it as something represented, contained, engrammed, inside a brain or neural network, or CNS or CNS and parasympathetic nervouse system 10:38 < nsh> or all of those plus musculature and skeletal system 10:38 < nsh> or all of that plus environmental interactions 10:38 < nsh> or you can elect, again, to say that all information that has, does, or ever will exist is 'out there' 10:38 < nsh> and the brain is just riding along it 10:39 < nsh> how you proceed in theoretics and experiment will depend quite heavily on how you define memory 10:39 < nsh> i gtg for a run 10:39 < nsh> ttyl 10:46 < nmz787> isn't memory becoming more and more to be known as self-reinforcing, and that over time memories which a person holds as truths can actually shift based on their own biases and other associations which help form/hold that memory? 11:02 < nmz787> hmm, veritas doesn't seem to have a "buy it now" link on their products page for whole genome seq 11:04 < nmz787> oh, it's just buried in some clickbaity looking link 11:43 -!- juri_ [~juri@205.166.94.162] has quit [Ping timeout: 250 seconds] 11:48 -!- juri_ [~juri@205.166.94.162] has joined ##hplusroadmap 11:59 -!- Urchin [~urchin@unaffiliated/urchin] has joined ##hplusroadmap 12:08 -!- TMA [tma@twin.jikos.cz] has quit [Ping timeout: 240 seconds] 12:08 -!- TMA [tma@twin.jikos.cz] has joined ##hplusroadmap 12:54 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has quit [Ping timeout: 240 seconds] 12:58 < kanzure> lab removes wall decoration to punish old man https://twitter.com/erlichya/status/1088456015715024897 13:00 < kanzure> .tw https://twitter.com/kuriousmind93/status/1074315652821909505 13:00 < yoleaux> @FSIStanford @StanfordHP @HankGreelyLSJU The patent holders must restrict #CRISPR use only under supervision of established academic PIs. The governments and @UN @UNIDIR should classify germline #geneediting as potential biological weapon of mass destruction. @DHScwmd @doddtra https://www.technologyreview.com/s/600774/top-us-intelligence-official-calls-gene-editing-a-wmd-threat/ (@kuriousmind93, in reply to tw:1074110366802763777) 13:13 < nsh> .title https://www.youtube.com/watch?v=upXsle3yB6A 13:13 < yoleaux> Jeff Hawkins - Human Brain Project Keynote [Screencast] - YouTube 13:13 < nsh> some good [imho] ideas about the neocortex, and how information or representations of information organised by location might work 13:14 < nsh> and how this might relate to learning and intelligence 13:14 < nsh> (apropos of earlier on memory) 13:16 < kanzure> do you have any productive ideas or just mystic comments about complexity? 13:51 < kanzure> guess not. 14:17 -!- spaceangel [~spaceange@ip-86-49-16-65.net.upcbroadband.cz] has quit [Remote host closed the connection] 14:26 < kanzure> austein heinz had a designer baby startup around the time of cambrian genomics 14:36 < kanzure> craig venter's wife is his pr agent? 15:07 < fenn> nsh do a dual n-back task to the best of your ability. a number will consistently appear, regardless whether you try "hard" or "really hard" and this is what we claim to represent working memory 15:08 < fenn> it's the same as any measurement. if you want to question the validity of this measurement, you might as well question the concept of measurement in general, of anything 15:08 < nsh> what's a dual and back task sorry? 15:08 < nsh> you should question the concept of measurement 15:08 < fenn> N-back is where you recall the Nth last item in a list 15:08 < nsh> that's basically all physicists have done since 1920 or so 15:09 < nsh> (productively) 15:09 < nsh> ah 15:09 < fenn> dual N-back is where you recall the N-th last item in two lists 15:09 < nsh> what if how hard i try includes being born as someone with supranatural abilities at recollecting lists of numbers? 15:09 < nsh> which exist 15:10 < nsh> so we're talking about average working memory or something related to numerical temporal subitization 15:10 < nsh> but average isn't the limitation of what has been achieved 15:10 < nsh> however it's notable that these people with these supranatural abilities tend to have difficulties too 15:10 < nsh> probably relating to the pleotropy of traits 15:11 < nsh> (ie same genetic variation that gives amazing memory/recall can result in impairments through mechanisms we have yet to elucidate) 15:11 < nsh> breed foxes until they're super timid and they tend to have ears more like puppy dogs 15:11 < nsh> etc 15:12 < fenn> yes there is the hypothesis that humans are already maxed out on intelligence, and any further meddling will only cause bad things 15:12 < fenn> but i don't believe this, because there are genetic differences in human intelligence 15:12 < nsh> there's an anecdote about Henry Ford that he summoned all his top engineers and tasked them to go to junkyards and recall which components had the least wear 15:12 < nsh> and then returned expecting there to be some analysis of these great lasting components to make everything else as good 15:12 < nsh> but Ford just said "make these to a lower tolerance" 15:13 < nsh> same with the anecdote about fighter jets returning with bulletholes but pilots intact 15:13 < nsh> the places where surviving planes never had bulletholes were the places that fucked the planes 15:13 < nsh> so ageing and intelligence might maxed out by process like this 15:13 < nsh> but we see variation still in both 15:14 < fenn> i guess you're trying to say that intelligence doesn't necessarily improve fitness for reproduction beyond a certain point, and may be detrimental 15:14 < nsh> possibly 15:14 < nsh> but that's still assuming a certain metaalgorithm that depends on a variable roughly "how likely to reproduce, as singular organism" 15:14 < nsh> where things like multiplicative weights update are operating over populations 15:14 < kanzure> let's get back to the part where nsh doesn't believe in measurement 15:14 < nsh> and are relatively apathetic about individuals except in their contributions 15:14 < fenn> what weights are you talking about? 15:15 < nsh> one sec lemme get out of bath and find reference 15:15 < fenn> i'm not interested in quantum woo 15:15 < nsh> .t https://arxiv.org/abs/1502.05056 15:15 < yoleaux> nsh: Sorry, I don't know what timezone that is. If in doubt, see https://en.wikipedia.org/wiki/List_of_tz_database_time_zones for a list of options. 15:15 < nsh> .title https://arxiv.org/abs/1502.05056 15:15 < yoleaux> [1502.05056] On Sex, Evolution, and the Multiplicative Weights Update Algorithm 15:15 < nsh> .title https://jeremykun.com/2017/02/27/the-reasonable-effectiveness-of-the-multiplicative-weights-update-algorithm/ 15:15 < yoleaux> The Reasonable Effectiveness of the Multiplicative Weights Update Algorithm – Math ∩ Programming 15:16 < nsh> basically everything gets multiplied by everything else and this results in reweighting 15:16 < nsh> iirc 15:17 < nsh> or rather an abstracted environment multiplies/combines the weights in some complicated way and there's a measure of success and that results in the reweighting 15:17 < nsh> for evo it would be something like frequency of alleles but obviously that's reductive 15:20 < fenn> so the weight for a population is the frequency of the allele in that population? 15:22 < nsh> yeah maybe, i don't recall precisely sorry 17:26 -!- justan0theruser [~justanoth@unaffiliated/justanotheruser] has joined ##hplusroadmap 17:28 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has quit [Ping timeout: 245 seconds] 17:28 -!- justan0theruser is now known as justanotheruser 17:32 -!- Jay_Dugger [~jwdugger@47.185.249.138] has joined ##hplusroadmap 17:33 -!- Jay_Dugger [~jwdugger@47.185.249.138] has quit [Client Quit] 17:41 -!- sachy [~sachy@78.108.102.220] has quit [Quit: Leaving.] 17:50 -!- Gurkenglas [~Gurkengla@unaffiliated/gurkenglas] has joined ##hplusroadmap 18:25 -!- strages [uid11297@gateway/web/irccloud.com/x-zugolbqzhtchwdbk] has quit [Quit: Connection closed for inactivity] 18:31 < kanzure> https://www.theinformation.com/articles/Custom-Babies-Ethical-Issues-Loom-as-DNA-Tech-Advances 18:32 < kanzure> "Austen Heinz has seen the future of human reproduction, and it’s not one that everyone will like." 19:20 -!- Gurkenglas [~Gurkengla@unaffiliated/gurkenglas] has quit [Ping timeout: 245 seconds] 20:07 -!- rancyd_ [stryfe@gateway/shell/firrre/x-enpknwngnyhtidpj] has quit [Excess Flood] 20:08 -!- rancyd_ [firrre@gateway/shell/firrre/x-dsipwbsiolukiqrj] has joined ##hplusroadmap 22:28 -!- yashgaroth [~yashgarot@2606:6000:c308:f700:201a:dc66:3dae:4474] has quit [Quit: Leaving] 22:38 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has joined ##hplusroadmap 22:51 -!- l_wl [~l_wl@pool-173-66-183-216.washdc.fios.verizon.net] has quit [] --- Log closed Fri Jan 25 00:00:30 2019