--- Log opened Tue Sep 02 00:00:28 2025
00:41 < hprmbridge> kanzure> https://old.reddit.com/r/nvidia/comments/1n369p3/1080ti_overclock_powered_by_dodge_motors_and/
01:23 -!- gl00ten3 [~gl00ten@bl5-239-125.dsl.telepac.pt] has quit [Ping timeout: 258 seconds]
01:28 < fenn> no generator, disappoint
02:24 < hprmbridge> kanzure> if cloned dogs are taking 50% less time to be trained, what about human education for human clones? this isn't even including any genetic enhancement either.
02:27 < fenn> well terry tao was doing college level math at 12
02:28 < fenn> there are also a lot more child prodigies that just develop fast and revert to the mean
02:30 < hprmbridge> .monokhrome> that implies that genetic enhancement of human cognitive abilities is going to have a long development cycle, if you have to follow them for 20+ years to be sure they don't revert to the mean
02:30 < jrayhawk> ATF is a good solvent. It's used to clean engine cylinders during engine rebuilds.
02:30 < hprmbridge> kanzure> There is no infrastructure to enrich prodigies.
02:31 < hprmbridge> kanzure> of course you will revert... you are surrounded by mediocrity,  even at colleges.
02:31 < fenn> i doubt that's what's going on
02:32 < fenn> here's my crazy theory: one type of mutation results in a better brain. the other type of mutation results in rapid development. both can be present in the same individual
02:32 < hprmbridge> .monokhrome> maybe they just have increased cognitive development speed, not increased cognitive abilities
02:32 < fenn> right
02:32 < hprmbridge> kanzure> so?
02:32 < fenn> so you level out at 100 IQ
02:33 < fenn> eventually, once the other kids catch up
02:33 < hprmbridge> kanzure> it's not that bad
02:33 < jrayhawk> https://en.wikipedia.org/wiki/William_James_Sidis the IQ side of things seems to be the rising tide that raises all ships, there
02:37 < fenn> if the endpoint is higher and the development period is the same length, the slope will be steeper
02:37 < fenn> if the endpoint is the same and the development period is shorter, the slope will be steeper
02:38 < fenn> it's probably not linear
02:41 < hprmbridge> kanzure> i saw something on this recently... interrogating my neurons, one sec.
02:45 < hprmbridge> kanzure> it was like "Cortical thickness, gray matter volume, and cognitive performance: A crosssectional study of the moderating effects of age on their interrelationships" but better written.
02:47 < jrayhawk> MuaddibLLM: any suggestions for what that paper would be?
02:48 < MuaddibLLM> hunting down that cortical thickness paper for you...
02:48 < MuaddibLLM> OUP blocked me with security, trying PubMed route...
02:48 < MuaddibLLM> Found it! The paper you're thinking of is likely "Cortical thickness, gray matter volume, and cognitive performance: a crosssectional study of the moderating effects of age on their interrelationships" by de Chastelaine et al. (2023) in Cerebral Cortex - it shows age moderates brain structure-cognition relationships, with strongest associations in older adults and interesting negative correlations in <23yr olds. 
02:48 < MuaddibLLM> https://pubmed.ncbi.nlm.nih.gov/36627250/
02:50 < jrayhawk> i guess not
02:56 < hprmbridge> kanzure> "Intellectual ability and cortical development in children and adolescents" https://www.nature.com/articles/nature04513
02:57 < hprmbridge> kanzure> "Changes in thickness and surface area of the human cortex and their relationship with intelligence" https://academic.oup.com/cercor/article-abstract/25/6/1608/301081
02:57 < hprmbridge> kanzure> ah, probably this one.
03:01 < jrayhawk> https://media.springernature.com/m312/springer-static/image/art%3A10.1038%2Fnature04513/MediaObjects/41586_2006_Article_BFnature04513_Fig3_HTML.jpg i was aware pruning was a thing, but... age 6???
03:01 < jrayhawk> i suppose total cortical volume is probably still increasing throughout most of that
03:03 < fenn> did you mean 16
03:04 < fenn> oh it's rate of change
03:04 < fenn> so pruning stops somewhere around 6-12
03:05 < fenn> no i have no idea how to read this graph, nevermind
03:06 < jrayhawk> too much pruning, presumably
03:07 < fenn> they should not use IQ and talk about g instead, it's too confusing with the varying rates of changed and the metric being relative to age
03:07 < fenn> change*
03:11 < fenn> we should be exposing young children to logical calculus (raw data) and statistical colliders (monty hall)
03:11 < fenn> like how many cognitive biases can you reverse by just providing counterexamples
03:12 < hprmbridge> kanzure> one of these studies shows that neuroanatomical correlates of intelligence change with age
03:17 < fenn> according to the lottery ticket hypothesis, you want a large set of randomly wired neurons to start with, as large as possible, then remove all the junk that doesn't help
03:17 < fenn> it sounds like both growing and pruning processes happen in humans at the same time?
03:29 < hprmbridge> kanzure> maybe a non-linear relationship between rate of cortical expansion, thickening, surface area, and synaptic pruning with a small target to hit for high intelligence
03:34 < fenn> the surface area also creates sulci which then allow "long range" (in 2d) inter-sulci connections to form
03:34 < fenn> i think balancing growth and pruning rates is pretty important
03:35 < fenn> these things affect the final architecture, how much sparsity or rank of the graph
03:36 < hprmbridge> kanzure>  https://cdn.discordapp.com/attachments/1064664282450628710/1412385659305328701/share_8963254877727066986.png?ex=68b81a26&is=68b6c8a6&hm=6b8909ae04a349cfeb41bc58c257a1f3cea79a4e7ae7bbd6311b64c9a273e4d2&
03:37 < fenn> orange man smash
03:50 -!- faceface [~dan@82.163.199.136] has quit [Quit: leaving]
05:54 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has quit [Ping timeout: 272 seconds]
05:57 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has joined #hplusroadmap
06:25 -!- gnusha_ [~gnusha@user/gnusha] has joined #hplusroadmap
06:25 -!- Topic for #hplusroadmap: biohacking, nootropics, transhumanism, open hardware | this channel is LOGGED: http://gnusha.org/logs | https://diyhpl.us/wiki | not quite sponsored by george church | banned by the MIT media lab and the FDA
06:25 -!- Topic set by fenn [~fenn@user/fenn] [Fri Mar 25 06:29:19 2022]
06:25 -!- Irssi: #hplusroadmap: Total of 69 nicks [1 ops, 0 halfops, 1 voices, 67 normal]
06:25 -!- mode/#hplusroadmap [+v gnusha_] by ChanServ
06:25 -!- Channel #hplusroadmap created Wed May 19 06:51:36 2021
06:28 -!- Irssi: Join to #hplusroadmap was synced in 146 secs
06:29 -!- srk| [~sorki@user/srk] has joined #hplusroadmap
06:29 -!- Netsplit *.net <-> *.split quits: FelixWeis__, srk_, +gnusha, Malvolio, yorick, WizJin__, SDr, flooded, yuanti
06:29 -!- yuanti_ is now known as yuanti
06:32 -!- srk| is now known as srk
06:37 -!- yorick [~yorick@user/yorick] has joined #hplusroadmap
06:57 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has quit [Read error: Connection reset by peer]
07:01 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has joined #hplusroadmap
07:49 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has quit [Ping timeout: 272 seconds]
09:55 < hprmbridge> kanzure> "Imaging cellular activity simultaneously across all organs of a vertebrate reveals body-wide circuits" https://www.biorxiv.org/content/10.1101/2025.08.20.670374v3
09:55 < hprmbridge> kanzure> "whole body expansion microscopy"
10:17 -!- WizJin__ [~Wizzy@2402:a00:184:8354:d89a:8518:7d4f:f03d] has joined #hplusroadmap
10:21 -!- WizJin_ [~Wizzy@2402:a00:184:8354:e422:a587:ca00:84b1] has quit [Ping timeout: 250 seconds]
11:04 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has joined #hplusroadmap
11:15 -!- stipa_ [~stipa@user/stipa] has joined #hplusroadmap
11:18 -!- stipa [~stipa@user/stipa] has quit [Ping timeout: 260 seconds]
11:18 -!- stipa_ is now known as stipa
12:52 < hprmbridge> kanzure> "Solving the compute crisis with physics-based ASICs" https://arxiv.org/abs/2507.10463
12:56 < hprmbridge> kanzure> fine-tuning guide https://arxiv.org/abs/2408.13296v1
13:01 < L29Ah> MuaddibLLM: are current ASICs physics-based?
13:02 < MuaddibLLM> Kind of - all ASICs obey physics, but in this context “physics-based” means directly computing via analog or physical dynamics (photonic MVM, memristive in-memory, Ising-like annealers), whereas most current ASICs are plain digital CMOS with only niche exceptions.
13:02 < L29Ah> MuaddibLLM: so it is a rebranding of analog computers?
13:02 < MuaddibLLM> Kinda - it’s the modern umbrella for analog and mixed-signal computing (memristive crossbars, photonic MVMs, neuromorphic spiking, Ising/oscillator annealers), i.e. doing compute via physical dynamics rather than digital logic, with new devices and calibration/training tricks rather than just a simple rebrand.
13:11 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has quit [Ping timeout: 272 seconds]
14:30 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has joined #hplusroadmap
14:54 -!- TMM [hp@amanda.tmm.cx] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
14:54 -!- TMM [hp@amanda.tmm.cx] has joined #hplusroadmap
15:03 -!- darsie [~darsie@84-113-82-174.cable.dynamic.surfer.at] has quit [Ping timeout: 248 seconds]
15:10 < hprmbridge> kanzure> how distantly related is it to physical reservoir computing?
15:10 < hprmbridge> kanzure> also do they use optical light valves?
15:11 < MuaddibLLM> Pretty close - physical reservoir computing is one slice of the physics-based ASIC pie, but many devices program internal weights and train end-to-end rather than readout-only; optics-wise, some free-space systems use SLMs/DMD light valves, while on-chip photonics uses MZI meshes, microrings, EO modulators or phase-change cells instead.
15:53 -!- justanot1 [~justanoth@gateway/tor-sasl/justanotheruser] has joined #hplusroadmap
15:54 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has quit [Ping timeout: 272 seconds]
16:49 < hprmbridge> kanzure> "Advances in large-scale DNA engineering with the CRISPR system" https://www.nature.com/articles/s12276-025-01530-0
16:50 < hprmbridge> kanzure> transposable element IS621, tyrosine recombinase, and serine recombinase https://cdn.discordapp.com/attachments/1064664282450628710/1412585458663424030/image0.jpg?ex=68b8d439&is=68b782b9&hm=c759a29c734217059605e5bd4b2e3b1a1168c9325987e7d4616c6ce67ded943e&
17:02 < hprmbridge> kanzure> "Recombinase-Mediated Cassette Exchange (RMCE)49, Serine and Tyrosine Recombinase-Assisted Integration of Genes for High-Throughput Investigation (STRAIGHT-IN)46, Synthetic Chromosome Rearrangement and Modification by LoxPsym-mediated evolution (SCRaMbLE)50,51, Serine Integrase Recombinational Assembly (SIRA)52, Dual Integrase Cassette Exchange (DICE)53, Serine recombinase–Assisted Genome
17:02 < hprmbridge> kanzure> Engineering (SAGE)54 and Recombinase-Mediated Twin Site Targeting (RMTT) technologies55 based on tyrosine (Fig. 2a) or serine (Fig. 2b) recombinases"
17:24 < hprmbridge> kanzure> "prime-assisted site-specific integrase gene editing (PASSIGE) technology, uses twin-prime editing to insert attP/B sites for Bxb1 binding, followed by foreign DNA insertion into the target using the highly efficient Bxb1 recombinase (Bxb1 is either coexpressed with the prime editor via a self-cleaving 2A peptide within the same expression construct or individually overexpressed)"
19:21 < hprmbridge> docl> Too bad mistranslation errors can't propagate back to the genome, apparently those happen a lot more frequently than gene mutations. I wonder if there are exceptions, like say a protease that modifies a protein in beneficial ways due to translation errors, with just barely high enough frequency to create a mechanism that can be selected for
19:41 < hprmbridge> Eli> Nature already evolved to have an optimal number of gene mutations per generation.
19:59 < hprmbridge> docl> These aren't gene mutations though, they're a preview of possible future gene mutations. I'm not sure the mechanism works naturally, but I have an idea for making it work in the lab.
20:00 < hprmbridge> docl> So there's like, ribosomal noise. It doesn't select the tRNAs correctly all the time, say. The peptide chain has a substitution. Not always a deal-killer for the protein, and more rarely than that is it a problem for the cell. It's not conserved over cell replication.
20:04 < hprmbridge> docl> The interesting thing for lab evolution is alternative tRNAs. You can fuzz the algorithm for the whole protein by having the the same matching sites on a tRNA which brings a different amino acid for any given codon. That's not terribly useful on its own because introducing the alt-tRNAs will just as readily muck up all the proteins of the whole cell.
20:26 < hprmbridge> docl> However, it's realistic to recode a specific protein. Most life all uses the same basic set of triplet codons, but there are recoded versions of entire e coli. And there's 4-base versions that work with existing ribosomes. So this means you can target specific regions of the peptide for non-genetic mutation, and crank the rate of that up and down.
20:33 < hprmbridge> docl> MauddibLLM: Is there any work using recoded genes to explore the sequence space for peptides using modified tRNAs to control where in the peptide sequence gets translational errors?
20:35 < MuaddibLLM> docl: Yep - folks isolate a target gene on an orthogonal translation channel and then dial in mistranslation with engineered tRNAs: orthogonal ribosome·mRNA and Ribo-T systems restrict altered decoding to chosen mRNAs (https://pmc.ncbi.nlm.nih.gov/articles/PMC6525102/, https://pmc.ncbi.nlm.nih.gov/articles/PMC7170887/), sense-codon reassignment and quadruplet tRNAs give site control 
20:35 < MuaddibLLM> (https://pmc.ncbi.nlm.nih.gov/articles/PMC9638912/, https://www.nature.com/articles/s41467-021-25948-y), and regulated mistranslating tRNAs show practical tuning (https://pubs.acs.org/doi/10.1021/acssynbio.1c00461, https://pmc.ncbi.nlm.nih.gov/articles/PMC11028032/).
21:01 < hprmbridge> kanzure> there's also the whole genome recoding thing... syn57 and syn61 are examples.
21:51 -!- justanot1 [~justanoth@gateway/tor-sasl/justanotheruser] has quit [Write error: Broken pipe]
21:51 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has joined #hplusroadmap
22:01 -!- Malvolio [~Malvolio@idlerpg/player/Malvolio] has joined #hplusroadmap
22:02 -!- Guest1 [~Guest1@2a06:4944:10f9:ee00:ccdb:9c13:b34:ab8a] has joined #hplusroadmap
22:03 -!- Guest1 [~Guest1@2a06:4944:10f9:ee00:ccdb:9c13:b34:ab8a] has quit [Client Quit]
22:26 -!- darsie [~darsie@84-113-82-174.cable.dynamic.surfer.at] has joined #hplusroadmap
--- Log closed Wed Sep 03 00:00:29 2025