--- Log opened Sat Aug 30 00:00:25 2025
00:07 < jrayhawk> I don't understand how endocytosing mitochondria could possibly be worth the increased risk of endocytosing stuff like rickettsia (again)
00:11 < hprmbridge> .monokhrome> maybe it plays some vital role in organism development
00:11 < jrayhawk> unless it counteracts cancer risk by restoring apoptosis/glucose/amino acid transporter regulation or something, but papers just say it makes tumor cells grow even faster
00:20 <+gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=ecb52992 fenn: overview of factors affecting mitochondrial efficiency. this is the same as https://rentry.co/b8ypp6ye in the logs >> http://diyhpl.us/diyhpluswiki/users/fenn/mitochondrial-efficiency/
00:25 < hprmbridge> nmz787> @kanzure I'm 3 episodes into pantheon... Pretty darn good so far
00:32 <+gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=d13bf6c0 fennwiki: better formatting for ikiwiki markdown >> http://diyhpl.us/diyhpluswiki/users/fenn/mitochondrial-efficiency/
00:33 <+gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=1ca15bb1 fennwiki: the results of today's rabbit hole >> http://diyhpl.us/diyhpluswiki/mitochondria/
00:34 < fenn> i thought pantheon was an MMO
00:34 < fenn> "Set in a world where mind uploading technology is on the verge of mass adoption"
00:36 < fenn> jrayhawk: i mean the rebuttal says "they're not running" but if you put them in a cell i bet they'd run again. that doesn't mean it's a good idea, they've probably been evicted for a reason
00:37 < fenn> mitochondria move all over the body, it's normal
00:37 < jrayhawk> the implication may be that these are just cast-offs from exploded cells with no particular physiological significance
00:37 < fenn> sure
00:37 < fenn> apoptotic bodies would have a layer of cell membrane, right?
00:38 < MuaddibLLM> Yes - apoptotic bodies are membrane-bound vesicles that externalize phosphatidylserine and can encapsulate intact organelles (including mitochondria), which aids clearance and can also confound “cell-free mitochondria” measurements versus EVs or platelets.
00:38 < jrayhawk> as in a cellular phopholipid bilayer to go with the outer mitochondrial phospholipid bilayer?
00:38 < fenn> yes
00:39 < fenn> and typicall markers saying "eat me, i'm dead"
00:41 < hprmbridge> docl> I wonder if you can do it cheaply without brain autopsies by just having this big colony of genetically identical inbred mice that run mazes every day, by varying exposure to antivirals and/or immunosuppressants in their food depending on their performance. Select for viruses that enhance performance and against the ones that degrade it.
00:42 < fenn> what could possibly go wrong
00:43 < fenn> this is hilariously bad
00:43 < jrayhawk> if the vesicle just has typical cell surface expression, i suppose that resolves the "how does the immune system distinguish this from rickettsia-like parasites" problem
00:43 < fenn> docl: do you want fast-running zombies?
00:43 < hprmbridge> docl> I want something that increases working memory
00:44 < fenn> how do you select for that vs fast zombies?
00:44 < hprmbridge> docl> doesn't get them the immunosuppressant if the mouse keeps running into walls because it can't remember where the turn is supposed to be
00:44 < hprmbridge> docl> also if it kills the mouse before it completes the maze we select that out
00:46 < fenn> jrayhawk: Phosphatidylserine, loss of CD47, calreticulin, and oxidized lipids trigger engulfment and clearance
00:48 < fenn> skull impact resistant fast zombies
00:49 < fenn> call me a luddite but i don't think breeding brain altering viruses is wise
00:50 < hprmbridge> docl> thing is, these would be just normal viruses selected not to kill or enfeeble some already feeble inbred critters. we're not doing anything to make them more virulent out of context.
00:51 < jrayhawk> you need something like a virus to modify biology fast enough to keep up with silicon neural networks
00:51 < jrayhawk> germline ain't gonna cut it at this point
00:51 < fenn> use the silicon neural networks to devise a gene therapy plan
00:51 < fenn> and/or brain implants (more realistic)
00:52 < jrayhawk> which gene therapy plan doesn't involve viruses
00:52 < fenn> see, there's a difference
00:52 < fenn> yes, vectors are technically viruses. what docl is talking about is doing evolution on wild type viruses
00:53 < hprmbridge> docl> eh the way I picture this we're making them less fit in the wild. more docile.
00:54 < jrayhawk> we have to evolve the delivery vectors, too, because they're just not adequate for introducing systemic change
00:54 < fenn> well that's a different problem entirely
00:55 < fenn> you're breaking the problem down into manageable steps with an engineering mindset (good)
00:56 < fenn> instead of throwing raw terror at the wall and hoping something good comes out of the portal
00:56 < hprmbridge> docl> eh, no need to talk about inflicting high working memory capacity on people involuntarily, people want this. but might be worth going on immunosuppressants for a few weeks to get suddenly good at math.
00:57 < hprmbridge> docl> also I really think people will prefer their gene therapy as short lived rna packaged in non-viral vectors. less ick attached.
00:58 < hprmbridge> docl> but you have to have those genes in the first place
00:58 < jrayhawk> fuck that. short-lived RNA is expensive
00:58 < fenn> and lucrative
00:58 < hprmbridge> docl> every cell in my body churns out short lived rna all the time. super cheap for a nanomaterial
00:59 < fenn> what i'm saying is i expect life to find a way, and the entire virus breeding program to result in entirely unintended consequences, despite your careful selection criteria
00:59 < fenn> if it works, sure that'd be great
00:59 < jrayhawk> okay, let me rephrase that: delivering short-lived RNA is expensive
01:01 < hprmbridge> docl> fair enough
01:02  * fenn grumbles about minicircle
01:02 < fenn> have we really not found a safe landing pad site in the human genome yet?
01:02 < fenn> somewhere in one of the fourteen* copies of salivary amylase perhaps
01:03 < fenn> (varies by ethnicity, i can't be bothered to look up the actual value)
01:03 < jrayhawk> presumably the vaccine companies are working on DMCA-circumvention-clause-able proprietary genetic delivery platforms
01:03 < hprmbridge> nmz787> Why aren't y'all talking about engineered farts?
01:04 < hprmbridge> docl> I suspect that if they get into the wild the evolved virus just mutates and loses its selected-for properties because it doesn't need that stuff
01:04 < jrayhawk> it'll be like having Microsoft Windows running on every cell in your body
01:06 < hprmbridge> nmz787> "limited edition pumpkin spice....  farts"
01:06 < jrayhawk> gotta update the EFI firmware on your balls to get an updated platform key to secureboot your gamete cells
01:07 < hprmbridge> docl> liposomes are just carefully sonicated lipids though, not some proprietary magic
01:07 < hprmbridge> nmz787> Designer baby farts
01:08 < hprmbridge> nmz787> Sounds like something the average babyholic would love
01:09 < jrayhawk> delivery platforms would potentially involve proprietary genetic boilerplate to avoid offtarget modification and improved means of targeting specific tissues
01:11 < jrayhawk> COMIRNITY and friends had some, uh, slight tissue targeting problems, for instance
01:11 < fenn> @nmz787 AI beat us to it: https://old.reddit.com/r/ChatGPT/comments/1mv0lth/i_trained_an_ai_on_3_months_of_continuous_fart/
01:12 < hprmbridge> nmz787> I got gpt5 to talk about in utero electroporation for idiocracy 2.0 designer baby farts
01:13 < fenn> no joke armpit odor genetics is real
01:14 < hprmbridge> docl> viruses can target specific tissues, at least. just selecting for not making the mouse too sick to run the maze despite being immunocompromised is going to tend to keep it focused on tissues where it can make favorable edits to the task. I think at least you want a way to disable its replication capabilities before using in humans though.
01:14 < jrayhawk> i wrote an OBS-Websocket object model to dynamically create and tear down libavcodec video sources to create a live stream that were just spammed fart effects
01:15 < jrayhawk> i feel that this would synergize well with the above
01:15 < jrayhawk> wait, what channel am i in again
01:16 < hprmbridge> nmz787> /that/ channel
01:16 < hprmbridge> docl> the publicly logged one that will probably be used to train models on
01:16 < fenn> stop giving them ideas!
01:18 < fenn> wouldn't designer farts be more about the microbiome
01:19 < fenn> there's a lot of room for innovation in engineered microbiomes
01:19 < fenn> currently 99% of probiotics are the same ~10 yogurt strains
01:19 < hprmbridge> docl> not that I would necessarily say no to some good microbiome mods
01:20 < fenn> i am currently brewing up some general biotics equilibrium half-n-half yogurt or something, it may end horribly, but hard to get worse than my current microbiome
01:21 < hprmbridge> docl> microbiome could be a relatively low risk way to make use of some of the weirder ideas we have like putting a computer in the cell
01:22 < fenn> i have this problem where i can't use fresh milk older than 2 days. yogurt hard mode
01:22 < fenn> what's this about tumor microbiomes? why are there even bacteria (?) there in the first place?
01:23 < MuaddibLLM> fenn: Short version - tumors are leaky, hypoxic, and locally immunosuppressed, so translocated gut/oral bacteria can seed them and even persist intracellularly, sometimes modulating therapy response; overview in Nejman et al., Science 2020: https://pubmed.ncbi.nlm.nih.gov/32467386/
01:24 < hprmbridge> docl> hmm, presumably they get infected at some rate? hadn't thought of it
01:24 < hprmbridge> docl> yeah they're going to be locally immunosuppressed because otherwise the immune system would kill them
01:24 < fenn> heh i just opened that paper
01:28 < hprmbridge> docl> oh, brain tumors in smart people might be a way to get intelligence enhancing genes. IDK, maybe they would in some cases get a food supply benefit by making you think a lot?
01:28 < fenn> MuaddibLLM: i'm lazy, please summarize any interesting observations in that paper
01:28 < MuaddibLLM> r.jina.ai HTTP 451, retrying in a bit...
01:29 < fenn> oh right it's paywalled
01:29 < MuaddibLLM> r.jina.ai HTTP 451, retrying in a bit...
01:29 < jrayhawk> surely it knows how to search libgen/anna's/sci-hub
01:29 < hprmbridge> docl> MuaddibLLM: locate a non paywalled version please
01:29 < fenn> i dunno
01:29 < fenn> no stop poking it
01:30 < fenn> MuaddibLLM: can you read this and summarize any interesting observations https://files.catbox.moe/7v9c04.pdf
01:30 < MuaddibLLM> grabbing the PDF and skimming for the juicy bits...
01:30 < MuaddibLLM> Key juicy bits from that tumor microbiome paper: bacteria are mostly *intracellular* in cancer/immune cells (not just hanging around), breast tumors have the richest microbiomes, each tumor type has distinct bacterial signatures, and - coolest part - lung tumors in smokers are enriched for bacteria that can degrade cigarette smoke chemicals like nicotine and toluene, suggesting these metabolites create bacterial niches.
01:31 < MuaddibLLM> Science paywall being annoying, trying the free PMC version...
01:31 < MuaddibLLM> fenn: Here's the TL;DR of that massive Nejman paper - they found tumor-specific bacterial signatures across 1526 tumors from 7 cancer types, with bacteria mostly living inside cancer and immune cells (sometimes as cell wall-deficient L-forms), breast cancer had the richest microbiome, smokers' lung tumors harbored bacteria that can degrade cigarette chemicals, and different tumor types had distinct bacterial communities 
01:31 < MuaddibLLM> that correlated with therapy response.
01:31 < jrayhawk> huh, neat
01:32 < fenn> ah i didn't see the "free full text" button
01:32 < fenn> i hate that interface
01:36 < hprmbridge> docl> I wonder if you can select for bacteria that make the tumor less immunosuppressive? That's kind of uphill though, hmm. Ideally you want the tumor and bacteria to co-evolve their way into being easy to kill somehow. Hyper sensitive to ROS's maybe
01:37 < fenn> tumors are already hyper sensitive to ROS
01:37 < jrayhawk> and tumors are highly mutagenic to where you never, ever want to provide them with a gradient of survival
01:37 < hprmbridge> docl> yeah but if you hit them and don't kill them entirely they will evolve resistance
01:37  * fenn mumbles something about cancer stem cells
01:41 < hprmbridge> docl> that they have decentralized data storage is part of the problem. If you could get it such that they rely on a master cell to replicate you might be able to nail them that way
01:42 < hprmbridge> docl> maybe if the local immunosuppression depends on a bacterium that you can kill off predictably with antibiotics?
01:44 < hprmbridge> .monokhrome> Can you get healthy cells to express some sort of dna checksum, to signal that they're ok, and have the immune system kill the cells that fail?
01:45 < hprmbridge> docl> even if they just can't replicate without passing the checksum would be cool
01:46 < fenn> https://www.youtube.com/watch?v=w20TL4avw9M Starship Flight 10 landing footage, stabilized
01:47 < hprmbridge> .monokhrome> How do you get someone like elon musk interested in radical life extension, or at least curing cancer?
01:48 < fenn> i'd have to know more about why he's not already interested. his friends and acquaintances are involved in this stuff
01:51 < hprmbridge> docl> You could in principle make DNA-free cells that just have optically controlled RNA synthesis in the nucleus, and function exactly like normal cells. Then send out a copy of all the genes you want expressed, with uuid like tags specific to each cell, from a bunch of LED devices. The whole genome could be sent every few seconds on an optical frequency.
01:52 < hprmbridge> docl> Or maybe just very minimal DNA rather than none
01:52 < hprmbridge> .monokhrome> That sounds rather ambitious. We want something that's going to be simple enough to be put into clinical use against cancer
01:53 < fenn> yeah i don't see the point
01:53 < jrayhawk> fenn: that's some cartoony levels of jank on that rocket
01:54 < fenn> it's supposed to do that silly dogleg maneuver because it's pretending there's a barge on the ocean to land on, and you don't want to have an engine failure and go boom on the barge so it slides over sideways at the last possible second
01:54 < fenn> still not sure what the orange crap is
01:55 < jrayhawk> if it did actually have a barge to land on, it would've reported a nominal landing and then every single component would suddently detach from every other component and collapse into a neat pile
01:56 < fenn> the fire is probably because of a small explosion in orbit in a vent line. that's actually surprising
01:56 < jrayhawk> Wil-E Coyote would be sitting in what used to be the payload area blinking haplessly
01:59 < jrayhawk> then the neat pile would catch on fire
01:59 < hprmbridge> docl> the nucleus is the control center of the cell. it's not designed to be controllable, so why not replace it with something that is?
02:01 < fenn> why not design something not made out of spaghetti and barbed wire
02:01 < hprmbridge> .monokhrome> the cancer cells could develop mutations in whatever genes are supposed to make them undergo apoptosis. We really need an external mechanism to decide whether the cell lives or dies, based on examining its dna
02:02 < fenn> @monokhrome that's a killer T-cell
02:03 < fenn> it uses surface antigens mostly
02:03 < hprmbridge> .monokhrome> yes, that's what I was thinking of, the immune system can recgonise rogue polypeptides inside cells, it just has to recognise mutated dna
02:04 < jrayhawk> archaea and prokaryotes see nuclei as proof of the coddled decadence of eukaryotes
02:05 < hprmbridge> docl> ok so let's see. the cancer is basically limited by food and ROS production. its replication speedups tend to have the side effect of more ROS's which tend to kill it. so it tends to retain ability to make antioxidants to mitigate that. say we make a bacterium that releases antioxidants when exposed to light. keep it dark, gradually over time ramp that up with an LED. the cancer loses ability to
02:05 < hprmbridge> docl> synthesize antioxidants as it mutates, becomes dependent on the bacterium. then suddenly switch the light off and hit it with increased ROS.
02:06 < fenn> hmm. i haven't formulated my conspiracy theory (stand-alone-complex theory?) completely yet, but something like, public health leaders are incentivized by regulatory capture to ignore the actual causes of cancer, and instead promote a false paradigm that doesn't threaten business interests
02:06 < fenn> cancer isn't caused by DNA mutations, usually
02:08 < jrayhawk> well, there are certainly a lot of billions "invested" in sequencing cancer cells to finally, this time we swear, it'll happen guys, find out what the oncogenes that are definitely causing this problem are
02:08 < jrayhawk> we can't stop spending those billions now, guys, we've made it so far!
02:09 < fenn> forgetting about cancer for a sec, it would be cool to pyrosequence individual cells and have a nanobot nearby watching the pretty lights blink, and kill the cell if it doesn't match the checksums
02:09 < fenn> or at least if the fraction is too low to be worth editing
02:09 < hprmbridge> .monokhrome> Cancer's a difficult disease to cure completely. Medical science has made progress against it. There's no need for conspiracy theories as to why it hasn't been eradicated yet though, it's just a very challenging disaease.
02:10 < fenn> monokhrome they're barking up the wrong tree and demonizing anyone who tries to point it out
02:10 < jrayhawk> pathological incentives impact every field of medicine
02:10 < hprmbridge> docl> what's the right tree?
02:10 < fenn> pathological incentives impact every field
02:10 < fenn> the last few hours of mitochondrial metabolism stuff
02:12 < fenn> it would be nice to at least acknowledge that mitochondrial hormesis is a thing, and we can finally start building nuclear reactors again and save the planet
02:13 < fenn> wow it really is two birds with one stone
02:14 < jrayhawk> haha
02:14 < fenn> MuaddibLLM: please educate the channel about why ALARA is stupid
02:14 < MuaddibLLM> ALARA often turns a tiny, poorly evidenced LNT-risk at low doses into massive cost and operational drag, so a risk-informed approach with hard dose limits, transparent cost-benefit, and not chasing microSieverts would cut fear, speed work, and likely reduce total harm compared to fossil-heavy alternatives.
02:14 < fenn> MuaddibLLM: please educate the channel on why linear no threshold radiation dose is a bad model
02:15 < MuaddibLLM> grabbing a couple solid references so I don’t just hand-wave it
02:15 < MuaddibLLM> LNT ignores cellular repair and adaptive responses - reanalysis of atomic bomb survivor data actually shows hormesis (reduced cancer) at low doses when you correct for baseline bias, plus high-background areas like Ramsar show no increased cancer rates despite 10x normal radiation (https://pubmed.ncbi.nlm.nih.gov/24298226/)
02:16 < fenn> MuaddibLLM: explain what effects this knowledge would have on nuclear power plant construction if it were mainstream knowledge
02:16 < MuaddibLLM> If low-dose risk were treated as threshold/hormetic instead of LNT, you’d drop a lot of ALARA-driven overbuild and micromonitoring, likely shrink EPZs toward the site boundary for many designs, allow higher worker dose budgets to speed outages, trim shielding and rad-waste handling, streamline licensing/decommissioning, and with simpler civils and fewer delay points you’d finish plants faster and cheaper without 
02:16 < MuaddibLLM> sacrificing safety.
02:17 < fenn> the only solution is to get rid of ALARA as a policy
02:19 < fenn> ALARA = "as low as reasonably achievable" meaning cost of nuclear plant construction rises until it is equal to other power sources. then when costs of other sources fall, the same safety threshold remains in effect, meaning nuclear cost will always be >= alternatives
02:20 < fenn> because even microscopic amounts of radiation are considered a risk by boneheads
02:20 < fenn> this is what i think when i hear about mutations causing cancer
02:22 < fenn> .t https://pmc.ncbi.nlm.nih.gov/articles/PMC2477708/
02:22 < saxo> Effects of Cobalt-60 Exposure on Health of Taiwan Residents Suggest New Approach Needed in Radiation Protection - PMC
02:23 < jrayhawk> i think of cardiologists causing heart disease via the American Heart Association, immunization experts causing spike protein pathology via the normalization of antibody trials and test-negative case control trials, and dieticians causing type 2 diabetes via the food pyramid.
02:24 < jrayhawk> epidemiologists and oncologists causing cancer via the somatic gene mutation theory is just another disastrous policy enshrined in perpetual marriage with institutional prestige
02:26 < jrayhawk> a faustian bargain for social power at the expense of being able to discern truth
02:28 < fenn> humans see the world not as it is, but how they wish it were
02:30 < fenn> reality is just too complicated to comprehend
02:37 < jrayhawk> docl: https://scholar.google.com/scholar?q=press-pulse+cancer+glutamine is what i am rooting for
02:43 < jrayhawk> i mean, for killing existing cancers. dealing with initiation is just the usual thing of conceptualizing food quality in terms of immunological safety, oxidative safety, micronutrient concentration, and micronutrient bioavailability, so as to keep the regulators of insulin resistance and apoptosis, mitochondria, contented and functional.
02:44 < fenn> it took me a long time (years) to understand and internalize what you're saying
02:47 < fenn> i found a discarded apocalypse stockpile of rice and beans today, and thought of the other rayhawk
02:48 < fenn> a certain type of person screams "just eat REAL FOOD" without ever explaining what "real food" means
02:48 < jrayhawk> i am sure society has found whole new forms of biological misalignment to obsolete what i am saying in those years
02:48 < jrayhawk> people are v. creative
02:49 < fenn> a pig farmer i know goes on and on about PFAS in the water
02:49 < jrayhawk> do we even have teflon manufacturers on the west coast?
02:50 < fenn> the main source is airports conducting fire drills with real PFAS containing foam. they dump thousands of gallons on the ground on a regular basis
02:50 < fenn> that won't stop people complaining about teflon though
02:50 < hprmbridge> .monokhrome> if it stays in the water forever, won't it eventually spread everywhere?
02:50 < jrayhawk> oh, wow
02:51 < fenn> the pig farmer is in romania. it's a global problem
02:51 < fenn> it's possible to destroy PFAS with ozone and UV
02:51 < fenn> (at the same time)
02:52 < hprmbridge> .monokhrome> is there pfas in bed pillows?
02:52 < hprmbridge> .monokhrome> I just read I'm not supposed to put pillows in recycling because they're full of toxic chemicals. But I sleep on them.
02:52 < fenn> you might be thinking of PBDE fire retardants
02:53 < jrayhawk> EU: "2026: PFAS will be banned in PFHxA foams for training, testing, and public fire rescue. 2029: PFAS will be banned in PFHxA foams in civil aviation."
02:53 < jrayhawk> well, at least the hyperregulatory state is good for *something*
02:56 < hprmbridge> .monokhrome> I'm sure I've seen teflon labels in my bedding. It's either  the pillow or the quilt or something
02:56 < fenn> well teflon makes stuff slippery
02:56 < fenn> where do you live such that pillow recycling is a thing?
02:57 < jrayhawk> the teflon itself is not the big problem, just the means of attaching the teflon to other things
02:57 < jrayhawk> which is a manufacturing step
02:57 < hprmbridge> .monokhrome> I don't think it is recyclable, I just wondered if it might be
02:58 < fenn> when you burn teflon or overheat it, it offgasses toxic stuff
02:58 < hprmbridge> .monokhrome> the alternative to have teflon is dust mites, which are bad for my asthma
03:00 < fenn> there can be trace residue of manufacturing inputs to the teflon coating process that didn't get washed off, and those can be toxic to birds, but i dunno about humans
03:01 < fenn> anyway they make a big deal about teflon so i assume it's not a big deal
03:01 < fenn> nobody says shit about all the houses with PVC siding outgassing phthalates 24/7 into the HVAC intake
03:02 < fenn> or the receipts literally covered with BPA
03:03 < fenn> and they wonder why the fertility rate is plummeting
03:04 < fenn> the conspiracy theory i'd like to talk about has been self-censored by the new mob
03:04 < fenn> uh, i mean. i am self censoring because i don't want to deal with it, and there's nothing i can do about it anyway
03:09 < TMA> Several girls in our hackerspace mentioned how the discourse about "safe sex" in the 1990's instilled into them instinctive fear of pregnancy and STDs when they were teens and that that might be a contributing factor as well
03:11 < TMA> in that the women (formerly girls scared of pregnancy) don't update the attitude to being pregnant well until their 40's when it is already somewhat late
03:11 < TMA> s/in the 1900's/since the 1990's/
03:15 < fenn> pregnancy *is* dangerous
03:16 < fenn> also means your social life is effectively over
03:17 < fenn> did they have to carry around the fake baby doll
03:17 < jrayhawk> .title https://www.sciencedirect.com/science/article/abs/pii/S0171298509001521
03:17 < saxo> ScienceDirect
03:18 < jrayhawk> hmm. i wonder if another request fixes it.
03:18 < jrayhawk> .title https://www.sciencedirect.com/science/article/abs/pii/S0171298509001521
03:18 < saxo> ScienceDirect
03:18 < jrayhawk> nope
03:18 < jrayhawk> bummer
03:18 < fenn> "Choline attenuates immune inflammation and suppresses oxidative stress in patients with asthma"
03:19 < jrayhawk> what sort of lunatics do RCTs without a placebo group
03:19 < TMA> you still cannot avoid both low fertility and pregnancy at the same time
03:20 < fenn> well technically...
03:20 < TMA> the social impact just means that whe have successfully optimized social interaction against including children
03:21 < TMA> fenn: of course, death solves both problems, but I thought suggesting that in *this* particular channel is a bad taste
03:21 < fenn> i think there should be a test of personal responsibility and judgement which grants access to anyone of any age to do the dangerous activities
03:22 < TMA> that would be discriminatory
03:22 < fenn> if you can't pass the test, you don't get to drink and drive, shoot guns, build atomic weapons, etc
03:22 < TMA> hence politically infeasible
03:22 < fenn> of course it's politically infeasible. everything's politically infeasible
03:23 < fenn> also it would not be a simple pass/fail test, there would be multiple levels and perhaps you'd have to take additional tests if you score too high
03:23 < TMA> except for more regulation "for the greater good" like e.g. chat control
03:24 < fenn> also politically infeasible. i wonder how it got through
03:24 < TMA> fenn: it is by repeating "think of the children!!!" in scared voice over and over again
03:25  * fenn is reminded of the patriot act
03:26 < fenn> a manufactured crisis? well CONVENIENTLY we have this pre-written 738 page document to add to your laws, no changes needed!
03:29 < fenn> "public consultation had revealed that a majority of respondents, both citizens and stakeholders, opposed imposing an obligation to use chat control"
03:30 < fenn> "Over 80% of respondents opposed its application to end-to-end encrypted communications."
03:30 < fenn> not a random sample
03:34 < fenn> meanwhile, roblox has 400 million users and zero monitoring
03:34 < fenn> shows how serious they are
04:14 -!- Malvolio [~Malvolio@idlerpg/player/Malvolio] has joined #hplusroadmap
04:56 < L29Ah> 10:22:10]<fenn> i have this problem where i can't use fresh milk older than 2 days. yogurt hard mode
04:56 < L29Ah> don't you have cheap sterilized milk everywhere?
05:00 < L29Ah> 10:47:39]<hprmbridge> .monokhrome> How do you get someone like elon musk interested in radical life extension, or at least curing cancer?
05:00 < L29Ah> give him cancer
05:01 < L29Ah> although that didn't help with that apple guy...
05:05 -!- TMA [tma@twin.jikos.cz] has quit [Ping timeout: 260 seconds]
05:13 -!- TMA [tma@twin.jikos.cz] has joined #hplusroadmap
05:41 < hprmbridge> kanzure> interest is not enough; look at all the recent longevity companies funded by billionaires.
06:24 -!- Pixi [~Pixi@user/pixi] has joined #hplusroadmap
06:46 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has quit [Ping timeout: 272 seconds]
07:15 < hprmbridge> kanzure> for non-resuscitation vitrification of human brain, shouldn't we first be doing the expansion microscopy trick before vitrification....?
07:27 < hprmbridge> kanzure> "Three-dimensional intact-tissue sequencing of single-cell transcriptional states" https://pmc.ncbi.nlm.nih.gov/articles/PMC6339868/ (2018)
07:30 < hprmbridge> kanzure> "axonal BARseq" https://pmc.ncbi.nlm.nih.gov/articles/PMC9949159/
07:32 < hprmbridge> kanzure> apparently there is now a BARseq2 and a MAPseq2 improvements on BARseq and MAPseq for connectome sequencing and mapping.
07:41 < L29Ah> https://tinystash.undef.im/il/43jDXBMCFkcwULefCVEEfsnVUwRo3uQ8pHE9TNKX3dKBRqTCQMQZBjRWZF5uSGMYPv94YzbJUuwQHyQKuPgvDZWM.png haha data science go brrr
08:29 < hprmbridge> kanzure> https://www.reuters.com/world/americas/worlds-first-gene-edited-horses-are-shaking-up-genteel-sport-polo-2025-08-30/
08:29 < hprmbridge> kanzure> "These five 10-month-olds are the world’s first genetically edited horses: cloned copies of a prize-winning horse named Polo Pureza, or Polo Purity... Kheiron Biotech is the Argentine company that created the clone horses."
08:30 < hprmbridge> kanzure> "While Argentina, regarded as the global capital of polo, has long welcomed reproductive technologies – including cloning – to breed elite horses, the sport’s national body and breeding association are putting up hurdles to prevent GE horses from joining the game. The Argentine Polo Association has banned GE horses from competition."
08:31 < hprmbridge> kanzure> "The world’s first cloned horse was born in 2003. Adolfo Cambiaso, widely considered the world’s top player, helped popularize polo clones. When a clone of Cambiaso’s prized Cuartetera sold at an auction in 2010 for $800,000, the amount caught the attention of Vichera, then a biotech doctoral student. Vichera went on to co-found Kheiron the next year with backing from businessman Daniel
08:31 < hprmbridge> kanzure> Sammartino. Its first cloned horse was born in 2013."
08:31 < hprmbridge> kanzure> "Cloned horses weren’t an easy sell at first. To get going, Sammartino said he provided free cloning services to top breeders, who allowed him to keep some of the newborn clones in exchange. By this year, the company says it will produce 400 clones, more than half of all the cloned horses born in Argentina in 2025, according to the breeders’ association’s estimates. The cloned foals sell for an
08:31 < hprmbridge> kanzure> average of $40,000, Sammartino said. In 2017, the Kheiron lab used CRISPR genetic editing to produce nine GE horse embryos for research purposes."
08:32 < hprmbridge> kanzure> "The association decided to tread cautiously" inappropriate pun
08:33 < hprmbridge> kanzure> "Sammartino admitted that plans to commercialize their gene-editing service are on hold until the polo authorities are on board" welp. we are doomed.
08:40 < hprmbridge> docl> I think elon not trying to cure aging or cancer would probably comes down to a futility argument, lack of belief that he can do anything about it. It's weird to see that contrasted with the space stuff though. Also maybe picturing a finite lifespan is somehow part of his motivational structure, dying on mars means getting there in the first place.
08:43 < hprmbridge> docl> That makes me imagine he's super fatalistic about everything on the inside and all the optimism is forced/localized, like ok the world is an unsolvable mess in most ways but at least I can ensure humans get boots on mars before I die
08:43 < hprmbridge> kanzure> making a rocket and making cells live forever are not even close to the same kind of complexity
08:49 < hprmbridge> kanzure> https://www.kheiron-biotech.com/
08:54 < hprmbridge> docl> I guess if you are solving something that's only as complex as a self landing rocket robot the size of a skyscraper it might be more ego flattering in a way than reverse engineering an insanely complicated alien technology that the universe handed you with no instruction manual
08:56 < hprmbridge> kanzure> the kinds of things we need to do to expand lifespan are the kinds of things that are considered off limits (most longevity gains presently come from germline engineering)
08:58 < hprmbridge> kanzure> other cloning companies: ViaGen Pets & Equine (dog, cat, horse), Trans Ova Genetics (cow, pig, sheep, goat, deer), Sooam Biotech Research Foundation (dog), Sinogene Biotechnology (dog, cat, horse), Kheiron Biotech (horse), Clonargen Biotech (horse), Avantea (horse, cow), Ovohorse / Ovoclone (horse), Replica Farm (horse), Cryozootech (horse), Fortgen (cow, sheep, goat), Crestview Genetics (horse),
08:58 < hprmbridge> kanzure> Boyalife/Sinica (dog, horse, cow).
08:58 < hprmbridge> docl> nonviral vector rna gene therapy is the least off limits variation on gene therapy
08:58 < hprmbridge> kanzure> delivery is the problem.
08:58 < hprmbridge> kanzure> (and, long term expression)
08:59 < hprmbridge> kanzure> there are a lot of difficult hard to reach cells.
09:14 < hprmbridge> kanzure> for longevity, i have noticed recently that early stage investors are moving more towards the ancephalic human clone whole-body transplantation strategy.
09:14 < hprmbridge> kanzure> there are two or three companies pursuing this direction.
09:24 < hprmbridge> kanzure> "Reproductive Biotechnology Centre (RBC) in Dubai, UAE is the primary commercial provider cloning camels (dromedaries; they’ve also produced Bactrian clones via interspecies transfer) for racing and beauty-pageant clients; it takes outside orders and reports cloning “a few dozen” camels a year. Closely related is Dubai’s Camel Reproduction Centre (CRC), the research group behind the world’s first
09:24 < hprmbridge> kanzure> cloned camel (“Injaz,” 2009); CRC underpins the UAE program, while RBC is the public-facing service arm."
09:29 < hprmbridge> kanzure> "In 2013, the The International Federation for Equestrian Sports (FEI) lifted the ban on cloned horses competing internationally following a review which found they were unlikely to have any advantage over horses bred traditionally."
09:31 < L29Ah> https://www.enhanced.com/newsroom/enhanced-games-files-800-million-antitrust-lawsuit-against-world-aquatics-the-world-anti-doping-agency-and-usa-swimming
09:32 < hprmbridge> kanzure> revivicor (pig) and jackson labs (JAX) (mice, maybe rats) is missing from my above list of cloning companies.
09:33 < hprmbridge> kanzure> " to ban from all of its games, including the Olympics, athletes and non-athletes (including coaches, trainers, doctors, physiotherapists and others) who participate in, actively support or even endorse any sporting event or competition that embraces scientific enhancements, whether or not they use any performance enhancing substances or commit any anti-doping violation under the WADA Code"
09:34 < hprmbridge> kanzure> last grasps of a dying empire right there... wtf. even the non-athletes?
10:09 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has left #hplusroadmap []
10:46 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has joined #hplusroadmap
11:13 -!- stipa_ [~stipa@user/stipa] has joined #hplusroadmap
11:16 -!- stipa [~stipa@user/stipa] has quit [Ping timeout: 248 seconds]
11:16 -!- stipa_ is now known as stipa
13:58 < hprmbridge> Eli> I believe this might fall under the definition of serial passage. That’s how they do gain of function research.
14:03 < hprmbridge> Eli> That’s sort of how things work in the body. If mutations alter the MHC, the immune system will see self as non-self and destroy the cell. Cancer is very sneaky though.
14:08 < hprmbridge> Eli> Some of pharmas biggest advances against cancer are immunotherapy, where we basically use the immune system to annihilate cancer cells. So, the checksum metaphor is sort of what they are using
14:08 < hprmbridge> .monokhrome> I know, that's why I wanted some sort of cryptographic checksum of the dna, the same way computer antivirus detects malware
14:09 < hprmbridge> Eli> Yeah, the body does this naturally. And then we enhance this with checkpoint inhibitors, car t, monoclonal antibodies.
14:11 < hprmbridge> Eli> Cancer really sucks though. Currently, lifestylemaxxing and prevention are basicaa as lot the way to “cure” most cancer by preventing it. Fenn is right that the financial incentives in cancer are weird.
14:13 < hprmbridge> Eli> Like, the biggest “breakthrough” in the war on cancer was getting people to quit smoking.
14:15 < hprmbridge> Eli>  https://cdn.discordapp.com/attachments/1064664282450628710/1411459251443863582/cancer-death-rates-by-type-who-db.png?ex=68b4bb5d&is=68b369dd&hm=7009a124daf658856c4a9a184d0395f9749bc9f8a5752c93e0a632ea3b3e5275&
14:16 < hprmbridge> Eli> colon/rectum decline, I think is due to screenings.
14:16 < hprmbridge> Eli> Stomach cancer was the top cancer, but, apparently we deleted it by inventing refrigerators https://cdn.discordapp.com/attachments/1064664282450628710/1411459689669197965/stomach.png?ex=68b4bbc5&is=68b36a45&hm=c531c5ba41d826ba00e35f70194202226c47bc0dcfa6b383767e0b167705d320&
14:30 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has joined #hplusroadmap
14:44 < fenn> stomach cancer could also be due to changes in diet
14:45 -!- darsie [~darsie@84-113-82-174.cable.dynamic.surfer.at] has quit [Ping timeout: 260 seconds]
14:46 < fenn> and smoking, apparently
14:47 < fenn> "The decline in the West is believed to be due to the decline of salted and pickled food consumption, as a result of the development of refrigeration as a method of preserving food"
14:47  * fenn high fives
14:48 <+gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=b0fa8223 Bryan Bishop: 3d bioprinting, cloning, FISSEQ, MAPseq, BARseq, in situ sequencing, lineage tracing, etc >> http://diyhpl.us/diyhpluswiki/optogenetic-axon-guidance/
14:51 <+gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=3afafac0 Bryan Bishop: and mapseq >> http://diyhpl.us/diyhpluswiki/mapseq/
14:54 < kanzure> https://spacetx-starfish.readthedocs.io/en/latest/ "starfish is a Python library for processing images of image-based spatial transcriptomics. It lets you build scalable pipelines that localize and quantify RNA transcripts in image data generated by any FISH method, from simple RNA single-molecule FISH to combinatorial barcoded assays"
14:59 < kanzure> MuaddibLLM: write me a script to go into the PMC articles linked from the wiki page and embed the images into the wiki just below or after the pubmedcentral (PMC) link.
15:00 < MuaddibLLM> diving into the diyhpl wiki to see how PMC links are structured...
15:00 < kanzure> O_o
15:00 < MuaddibLLM> cooking up a script to auto-embed PMC figures into wiki pages...
15:01 < MuaddibLLM> fixing those pesky regex escapes and polishing the script...
15:01 < fenn> please don't do vibe coding right in the channel
15:02 < MuaddibLLM> Created a script that auto-embeds PMC figures into wiki pages - it finds PMC links, downloads the figures via NCBI API, and inserts them right after each link with captions. Here's the fixed version:; ; ```python; #!/usr/bin/env python3; """; PMC Wiki Image Embedder - grabs figures from PMC articles and embeds them in wiki pages; Usage: pip install requests && python pmc_wiki_embedder.py; """; import re; import os; import 
15:02 < MuaddibLLM> subprocess; import requests; import xml.etree.ElementTree as ET; import time; import tempfile; ; def extract_pmc_id(pmc_url):;     """Extract PMC ID from various PMC URL formats.""";     patterns = [;         r'articles/PMC(\d+)',;         r'pmc/articles/PMC(\d+)', ;         r'PMC(\d+)',;         r'/(\d+)/?$';     ];     ;     for pattern in patterns:;         match = re.search(pattern, pmc_url);         if match:;         
15:02 < fenn> it should be a bookmarklet anyway
15:02 < kanzure> quantiative imaging of synapses https://pmc.ncbi.nlm.nih.gov/articles/PMC8860300/ see figure 2 i guess?
15:03 < fenn> or userscript. i'm not sure what the diff is
15:03 < kanzure> pasky: what model is this?
15:03 < fenn> non-resuscitation vitrification of human brain doesn't really make sense to me. is the goal to do barcoding microscopy at some point? otherwise why not aldehyde stabilized?
15:04 < fenn> sometimes it's gpt-5 sometimes claude-4-sonnet
15:04 < fenn> https://github.com/pasky/irssi-llmagent
15:05 < fenn> config.json.example:    "sarcastic": "anthropic:claude-sonnet-4-20250514"
15:05 < fenn> config.json.example:    "proactive_serious": "openai:gpt-5"
15:06 < fenn> regular serious is also gpt-5 in the example config
15:09 < fenn> crosslinks to DNA/RNA molecules physically prevent the enzyme from copying past that point
15:09 < fenn> someone really should be working on enzyme-cleavable crosslinks
15:10 < fenn> or some fancy biology-compatible reversible chemistry (i have no idea how this might work)
15:38 < hprmbridge> kanzure> you probably want azobenzene mediated crosslinking and decoupling, enzymatic would be too slow during revival
15:38 < hprmbridge> kanzure> pantheon was really, really good
15:53 < pasky> yes
15:53 < pasky> (well didn't finish season 2 yet but it is, i wish they'd do with less technobabble but even so)
15:54 < pasky> MuaddibLLM: !h
15:54 < MuaddibLLM> default is serious agentic mode with web tools (['openai:gpt-5', 'anthropic:claude-sonnet-4-20250514']), !d is explicit sarcastic diss mode (anthropic:claude-sonnet-4-20250514), !a forces thinking, !p is Perplexity (prefer English!)
15:57 < pasky> kanzure: the script is at https://pasky.or.cz/cp/pmc.py (indentation issues are mine, i pulled it out of a debug print)
15:57 < pasky> i guess i could give it a tool to publish artifacts
16:12 -!- TMM [hp@amanda.tmm.cx] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
16:12 -!- TMM [hp@amanda.tmm.cx] has joined #hplusroadmap
16:26 < pasky> (fixed, next time you ask muaddib to code something it'll share a link with the result)
16:29 < hprmbridge> docl> Just doing a bunch of "gain of function" (tainted term for it) research in immune deficient rodents seems like underrated life extension methodology.
16:32 < hprmbridge> docl> it smells like the kind of thing that sounds bad/unthinkable on the surface but really isn't if you don't do psycho things with it on purpose.
16:32 < hprmbridge> kanzure> or selection in general
16:33 < hprmbridge> docl> selection on the rodents? that's more in the overrated category. select the viruses, get fast, deliverable results.
16:46 < hprmbridge> docl> well, maybe not that overrated, but in comparison
17:26 < fenn> pasky: wow that was quick
17:26 < fenn> MuaddibLLM: paste some code for me, let's try it out
17:26 < MuaddibLLM> cooking up a userscript + bookmarklet combo for PMC figure embedding...
17:26 < fenn> good bot
17:27 < MuaddibLLM> Error: No valid text or tool use found in response
17:27 < fenn> oh noes
17:27 < fenn> MuaddibLLM: better luck next time
17:27 < MuaddibLLM> Heh, want me to take another swing and drop a link to a working PMC figure-embed userscript/bookmarklet - which browser are you on?
17:27 < pasky> whoops
17:27 < fenn> firefox
17:28 < pasky>       "name": "execute_python",
17:28 < pasky>       "input": {}
17:28 < pasky> what the heck
17:29 < pasky> hm i should really loop this back to the llm
17:29 < fenn> the agent* :)
17:30 < pasky>   "stop_reason": "max_tokens",
17:30 < pasky> oh
17:31 < fenn> maybe have a different token limit for making artifacts
17:32 < pasky> well you don't know that's going to happen in advance
17:32 < fenn> another bot has this problem where all its artifacts get included with every line of chat history, until they scroll off the window, which leads to something like 80x duplication of reading its own output, and it's expensive. i think a better solution would be to make clear that the bot can look at the output it wrote before responding
17:33 < fenn> or use another smaller LLM to shape the context
17:36 < fenn> too bad commercial inference engines don't let you move pre-digested KV cache chunks around
17:37 < fenn> for many use cases it doesn't matter what order you read documents in, and it turns out you can just move chunks of memory around, causal model be damned, and it works fine
17:40 < pasky> here only the final chat messages are included (but i wanted to enhance this actually, but only on demand via tool use)
17:40 < pasky> MuaddibLLM: thid time's a charm?  (write me a script to go into the PMC articles linked from the wiki page and embed the images into the wiki just below or after the pubmedcentral (PMC) link.)
17:42 < MuaddibLLM> crafting firefox userscript magic to auto-embed PMC figures...
17:42 < MuaddibLLM> packaging up the userscript for easy firefox install...
17:43 < MuaddibLLM> There you go - install Tampermonkey/Violentmonkey in Firefox, then paste that userscript to auto-embed PMC figure thumbnails below any PMC links: https://pasky.or.cz/muaddib/5016cef30d87432d959155d92afc238f.txt
17:44 < pasky> hm downside of the bookmarklet is it didn't test it, while it actually did the python version and fixed some stuff
17:46 < L29Ah> bacteria are annoying: i did 8 recursive batches of yogurt from two different starter lines, and they are all noticeably different from each other
18:31 < fenn> in the script, s/www/pmc/ because the url is e.g. https://pmc.ncbi.nlm.nih.gov/articles/PMC7019143/ and delete /pmc from the regex and querySelector and pmcId = pmcMatch[2]
18:32 < fenn> in the above url the first figure is https://cdn.ncbi.nlm.nih.gov/pmc/blobs/3800/7019143/77eb4c56652a/ijbsv16p0849g001.jpg
18:33 < fenn> MuaddibLLM: after fixing the above issues i still get "No figures found" please look at the pmc link i provided and think about where the script might be going wrong and fix it. also fix the issues i pointed out
18:33 < hprmbridge> kanzure> wouldn't you have an easier time with a python script
18:33 < MuaddibLLM> debugging PMC DOM structure and fixing those URL issues...
18:34 < fenn> what would i do with a python script?
18:34 < fenn> i want link previews in the browser, is that not the goal?
18:34 < hprmbridge> kanzure> embedded images on the wiki page.
18:34 < MuaddibLLM> fixed the URL issues and DOM selectors - packaging the corrected script...
18:35 < MuaddibLLM> Fixed the URL regex, PMC ID extraction (pmcMatch[1] not [2]), and DOM selectors to target cdn.ncbi.nlm.nih.gov/pmc/blobs images - verified against https://pmc.ncbi.nlm.nih.gov/articles/PMC7019143/: https://pasky.or.cz/muaddib/162187844f784c1fbfbbe780e5aebdbc.txt
18:35 < hprmbridge> kanzure> "Cortex folding by combined progenitor expansion and adhesion-controlled neuronal migration" https://www.nature.com/articles/s41467-025-62858-9
18:35 < hprmbridge> kanzure> "Here, we show in mice that combining the Flrt1/3 double knockout with an additional genetic deletion that causes progenitor expansion, greatly enhances cortex folding"
18:36 < hprmbridge> kanzure> "Expansion of intermediate progenitors by deletion of Cep83 leads to a relative increase in Flrt-mutant neurons resulting in enhanced formation of sulci. Expansion of apical progenitors by deletion of Fgf10 leads to a relative reduction in Flrt-mutant neurons resulting in enhanced formation of gyri."
18:37 < fenn> looks like it rewrote the entire script from scratch
18:46 < fenn> i had to change fetch back to GM_xmlhttpRequest to allow CORS
18:46 < fenn> it works but it's super ugly
18:47 < fenn> oh it's showing each figure for each figure, that's why there's so many
18:55 < fenn> comment out the mutationObserver stuff
18:56 < hprmbridge> kanzure> "Altered cerebral cortical gyrification in ferrets with neonatal exposure to the bacterial endotoxin, lipopolysaccharide" https://pmc.ncbi.nlm.nih.gov/articles/PMC12331228/ cc @docl
19:14 < fenn> looks decent, it would be cool if hovering the link popped up the image box like wikipedia link previews but this is as far as i am going to go with it: https://files.catbox.moe/up2p45.png https://files.catbox.moe/91ece6.js
19:15 < fenn> oh right i have a website
20:25 < hprmbridge> .monokhrome> intermediate fasting may kill you: https://www.sciencedirect.com/science/article/abs/pii/S1871402125000955
20:45 < hprmbridge> docl> https://magazine.hms.harvard.edu/articles/good-viruses-do
20:57 < hprmbridge> .monokhrome> are there any human viruses that are indisputably symbiotic in benefits?
21:08 -!- MuaddibLLM [muaddib@pasky.or.cz] has quit [Ping timeout: 256 seconds]
21:11 -!- MuaddibLLM [muaddib@pasky.or.cz] has joined #hplusroadmap
21:15 < hprmbridge> docl> seems weird to me that we've known about evolution for 200 years but haven't been selectively breeding symbiotic viruses all this time
21:42 -!- TMM [hp@amanda.tmm.cx] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
21:43 -!- TMM [hp@amanda.tmm.cx] has joined #hplusroadmap
21:43 -!- Malvolio [~Malvolio@idlerpg/player/Malvolio] has quit [Quit: 602F Shaka, when the walls fell 2025-08-31 04:43:40:821]
22:50 < fenn> amazon overnight delivery of perishables is 25-50% cheaper than going to the cheapest grocery store here...
22:50 < fenn> what a time to be alive!
--- Log closed Sun Aug 31 00:00:26 2025