--- Log opened Thu Mar 09 00:00:05 2023
00:06 -!- codaraxis [~codaraxis@user/codaraxis] has quit [Quit: Leaving]
00:08 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has quit [Ping timeout: 255 seconds]
00:33 -!- AMG [ghebo@user/amg] has quit [Ping timeout: 260 seconds]
02:18 < nsh> mellow: https://web.archive.org/web/20170830040110/https://users.dsic.upv.es/~flip/papers/TR-upsycho2012.pdf
03:58 -!- alexbfi [~alexbfi@dzdkdfyy6mygg63t2xykt-3.rev.dnainternet.fi] has joined #hplusroadmap
03:58 -!- alexbfi_ [~alexbfi@dzdkdfyy6mygg63t2xykt-3.rev.dnainternet.fi] has joined #hplusroadmap
04:01 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has joined #hplusroadmap
04:42 < hprmbridge> mellow> wayback method, smart, thanks
05:26 -!- AMG [ghebo@2605:6400:c847:1449::9441] has joined #hplusroadmap
06:09 -!- alexbfi_ [~alexbfi@dzdkdfyy6mygg63t2xykt-3.rev.dnainternet.fi] has quit [Ping timeout: 252 seconds]
06:09 -!- alexbfi [~alexbfi@dzdkdfyy6mygg63t2xykt-3.rev.dnainternet.fi] has quit [Ping timeout: 252 seconds]
06:10 -!- yashgaroth [~ffffffff@2601:5c4:c780:6aa0:a42b:b972:f1fe:455] has joined #hplusroadmap
06:38 < kanzure> grimes has some weird singularity energy going on https://twitter.com/Grimezsz/status/1616264128376745985 https://twitter.com/Grimezsz/status/1633703659359068169
07:03 < kanzure> met someone who is working on a macro-scale von neumann mechanical self-replication system.. vitamin parts are raw materials and CPUs. not bad.
07:28 < muurkha> for my next trick, I will build an automobile from scratch, starting with only sheet metal, a file, and a Porsche Boxster
07:57 < kanzure> "Structural basis for bacterial energy extraction from atmospheric hydrogen" https://www.nature.com/articles/s41586-023-05781-7
08:53 < kanzure> ansa says they have synthesized a 1,005mer with terminal transferase https://twitter.com/sethbannon/status/1633848116154880001
08:53 < kanzure> https://www.genengnews.com/topics/genome-editing/synthetic-biology/dna-synthesis/worlds-longest-oligo-produced-using-de-novo-synthesis/
09:09 < muurkha> nice
09:28 < kanzure> sir reprap speculating about reverse translation - he calls it a "reverse ribosome" or "emosobir" https://twitter.com/adrianbowyer/status/1633881289982869514
10:02 -!- cthlolo [~lorogue@77.33.23.154.dhcp.fibianet.dk] has joined #hplusroadmap
10:11 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has quit [Ping timeout: 248 seconds]
10:55 < hprmbridge> kanzure> "make people better" has a podcast now https://twitter.com/antonioregalado/status/1633902411268931590
11:29 < hprmbridge> nmz787> https://www.raspberrypi.com/products/raspberry-pi-global-shutter-camera/
11:30 -!- TMM_ [hp@amanda.tmm.cx] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
11:30 -!- TMM_ [hp@amanda.tmm.cx] has joined #hplusroadmap
11:35 < hprmbridge> nmz787> Wow that ansa tdt yield is pretty terrible
11:35 < hprmbridge> nmz787> 28%
11:37 < kanzure> compared to phosphoramidite at 1000 bp it's miraculous
11:37 < kanzure> phosphoramidite chemistry is, what, 99.5% yield per step?
11:38 < kanzure> so you get 28% at 250 bp
12:18 -!- cthlolo [~lorogue@77.33.23.154.dhcp.fibianet.dk] has quit [Read error: Connection reset by peer]
12:22 < hprmbridge> nmz787> All terrible, or possibly my Intel engineering experience has me biased about what good manufacturing standards are
12:22 < kanzure> hey if we could print 100x more transistors but have to design things to tolerate a higher % bad transistors.. that might be a reasonable trade.
12:24 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has joined #hplusroadmap
12:27 < lsneff> I dunno, really hard to make traditional computers deal with any significant rate of bad transistors. You can have a huge number of independent processing units and just seal off the bad ones, but if the rate is too high, then nothing will work
12:27 < kanzure> pretty sure von neumann wrote about this
12:28 < kanzure> "Synthesis of reliable organisms from unreliable components" http://fab.cba.mit.edu/classes/862.16/notes/computation/vonNeumann-1956.pdf
12:30 < TMA> I have read recently that the overhead for correcting errors is on the order of O(logN) where N is the number of unreliable components
12:35 < lsneff> Oh, if it's log, then high error rates are still usable
12:35 < lsneff> Hmm
12:41 < muurkha> if you have *uncorrelated* errors in your transistors, it's easy to design gates that tolerate them
12:44 < muurkha> CMOS uses transistors as switches.  If you replace such a transistor with four identical transistors with a common gate, and wire those transistors as a parallel combination of two series combinations of two transistors, you have a "fault-tolerant transistor" which acts like a single transistor with four times higher gate charge
12:45 < muurkha> so if any one of those four sub transistors fails closed (the usual way for MOSFETs to fail) or open, it will still operate as before, though with half the current capacity if one fails open
12:45 < muurkha> if you want to additionally tolerate gate oxide failures (which short the gate to the drain or source), you need to put resistors on the gates
12:47 < hprmbridge> nmz787> If you have higher gate capacitance, then you need higher voltage to run at a given speed
12:47 < muurkha> yes.  also if you add gate resistors
12:47 < hprmbridge> nmz787> And if things fail enough, drive strength compromise could mean your signal never gets to where you want it
12:47 < muurkha> yes
12:48 < muurkha> I think the bigger issue, though, is that an awful lot of possible failures aren't uncorrelated
12:48 < muurkha> a misaligned mask, an uneven deposition step, ESD damage, etc.
12:59 < L29Ah> makes more sense to make mosfets bigger
13:08 < muurkha> yeah, what I described is basically a bigger mosfet with some discontinuities in the channel for no good reason
13:21 < hprmbridge> nmz787> Kanzure has asked about making DNA bigger too
13:21 < hprmbridge> nmz787> But that isn't backwards compatible at this point
13:21 < hprmbridge> nmz787> And would need some translation invented anyway
13:21 < kanzure> yes like transfer printing from larger molecules to smaller ones (obviously, very different structures)
13:22 < hprmbridge> nmz787> Getting manufactuable yield in semiconductors is achievable, however
13:22 < hprmbridge> nmz787> (and I think the same for DNA, we're just obviously not there yet)
13:33 < kanzure> i was also wondering about transfer printing of DNA from some other objects, like vinyl records or something
13:33 < kanzure> once you transfer the data then it is a lot cheaper to copy
13:37 < fenn> how do you transfer print at a different size?
13:37 < fenn> or do you mean like, using DNA tiles to paint goop in nanoscale patterns?
13:37 < muurkha> sometimes you can elastically or plastically strain a material, but I don't know if that's what is meant here
13:37 < kanzure> well, say you represent data on a vinyl surface as an imprint of a certain size, and uh.. something something... and then that region of DNA does or does not get marked.. something something.
13:39 < kanzure> there's methylation, irradiation, cutting, base editing (substitutions), or you could have photocleavable nucleotides, a few other things that could be done..
13:39 < fenn> is this DNA stretched out in a line?
13:39 < kanzure> yes
13:41 < kanzure> it should be a bulk or batch copy operation instead of iteratively encoding bits into DNA (which is what my patented version does: https://patents.google.com/patent/US11339423/en )
13:41 < fenn> oho you've joined the dark side
13:42 < kanzure> well i figured i should at least see what it's all about from the other side
13:43 < fenn> ok so this is just the idea to use DNA copy errors as the data encoding method
13:43 < kanzure> yes
13:44 < fenn> anyway. you want to imprint big dents onto some goop with DNA. why not attach bigger blobs to the DNA like DNA binding proteins
13:44 < fenn> it would be difficult to resolve atom-scale dents
13:44 < kanzure> so you stretch out a DNA molecule and you run it by an inkjet printhead, and it jets out DNA binding proteins?
13:45 < kanzure> how would it be read back?
13:45 < fenn> you could encode data into the DNA and selectively mess up the binding regions, then in bulk mix that modulated DNA with binding proteins, and then stretch the DNA-protein complex out on a surface
13:46 < fenn> read out with an AFM or SEM
13:46 < kanzure> ok. some really big DNA binding protein.
13:46 < kanzure> unfortunately this can't be copied by PCR
13:54 < kanzure> maybe you could cut and insert a fragment at each location to indicate a bit
13:55 < fenn> i must have missed the part where the goal was specified
13:56 < fenn> DNA is already a fantastic data storage medium, and it comes with lots of tools for free
13:57 < fenn> are you trying to write data to DNA, read data from DNA, or use DNA as a template for fabricating inorganic electronic circuits?
13:57  * L29Ah disapproves DNA r/w error rate
13:57 < kanzure> 13:33 < kanzure> i was also wondering about transfer printing of DNA from some other objects, like vinyl records or something
13:57 < kanzure> 13:33 < kanzure> once you transfer the data then it is a lot cheaper to copy
13:57  * L29Ah prefers MOSFET gates
13:58 < kanzure> in this context i was thinking about just cheap redundant data storage by "imprinting" from another data source
13:58 < fenn> but copying DNA is already ridiculously cheap, for all intents and purposes "free" and it has a very low error rate
13:58 < kanzure> however, it does look like nmz787 was referring to DNA origami for device manufacturing
13:58 < kanzure> fenn: indeed. but writing the data into DNA is an issue right now... once it's in DNA then very cheap to copy.
13:59 < fenn> how about selectively abusing the DNA with surface-bound enzymes or chemicals
14:00 < kanzure> trouble at silicon valley bank (seems to be a consequence of silvergate "voluntarily" going into liquidation) https://twitter.com/howard/status/1633908101534367746
14:01 < fenn> super resolution masks to focus 280 nm UV patterns on 10 nm ish stretches of DNA to make thymidine dimers in those stretches
14:01 < fenn> evanescent wave stuff
14:02 < kanzure> making masks is hard.. you could use a pinhole shutter..
14:05 < fenn> thymine dimer*
14:05 < kanzure> more animal uplifting https://twitter.com/mattparlmer/status/1633929583555747843
14:05 < fenn> the mask would just be a grid of pinholes, and you do the modulation with a digital micromirror array
14:06 < fenn> position the mask, expose all the pinholes, step over 10nm, reapeat
14:07 < fenn> you'd want the DNA to have some sort of pseudorandom code between the writable regions, to keep track of the position you're at
14:08 < fenn> aligning all the separate strands of DNA seems like a pain
14:09 < fenn> nanopores is probably the smart way to do this
14:09 < fenn> feed dna through the nanopore and damage it as it goes by
14:10 < kanzure> something something cross-species young blood rejuvenation https://twitter.com/mujaya_/status/1633551625204953088
14:10 < kanzure> no reference tho
14:16 < kanzure> "make people better" podcast 1 https://twitter.com/MakePeople_Film/status/1633915801219768321 https://www.buzzsprout.com/2142608/12370908
14:44 < fenn> the tweet about young blood rejuvenation refers to harold katcher, he's a regular feature on josh mitteldorf's blog, e.g.: https://joshmitteldorf.scienceblog.com/2022/06/27/lifespan-of-harold-katchers-rats/
14:45 < fenn> he's kinda secretive because he's trying to commercialize it
14:47 < L29Ah> so, diy blood plasma dilution
14:48 < L29Ah> is it enough to insert some sort of a bypass with a couple of needles and let your heart pump blood through a membrane, removing part of your plasma while you sit behind your computer flaming in IRC?
14:49 < L29Ah> how fancy is this membrane? can it be bought for a few bucks off ebay?
14:49 < L29Ah> i guess if you dump your plasma slow enough, you can replace it with oral isotonic solution instead of injecting special sterile stuff
14:51 < L29Ah> albumin? fuck albumin
14:52 < fenn> it's called plasmapheresis and is a standard medical procedure
14:53 < fenn> the thing is, we don't know what the bad fraction is, or at least not when i looked into this a couple years ago
14:53 < L29Ah> pulling 4L of plasma isn't standard
14:53 < fenn> yeah
14:54 < fenn> sterile IV bags aren't *that* expensive
14:54 < fenn> but you can't buy them as a mere citizen
14:54 < L29Ah> the plasmapheresis equipment doesn't boil down to sterile IV bags
14:54 < fenn> i meant for simply diluting with saline
14:54 < L29Ah> diluting what?
14:55 < L29Ah> you need to discard stuff
14:55 < fenn> your blood, so you don't die of low blood pressure
14:55 < muurkha> fenn: yeah, I noticed that yesterday; why is it that you can't buy them?
14:55 < fenn> i guess you'd want ringer's, not saline. i'm no EMT
14:56 < fenn> muurkha: because it's a medical thing, and we're all supposed to just ... die
14:56 < fenn> honestly i don't know if there even is a rationale
14:57 < L29Ah> anyway, i don't see why i need to mess with sterility when i can use my own intestines and immune system to ensure sterility
14:57 < fenn> oh, so you just drink it. right
14:57 < fenn> good idea
14:57 < L29Ah> humans can routinely pump 20L of water from mouth to sweat in a hot day
14:58 < L29Ah> i think that means it ends up as plasma at some point
15:00 < L29Ah> so the thing we need to keep sterile is a tube, two needles and some membrane setup that dumps some of the plasma; i wonder if it needs any external pumping, if it does, probably a simple peristaltic pump will do
15:06 < muurkha> 20ℓ seems unlikely
15:06 < muurkha> maybe 8ℓ
15:06 < muurkha> probably more like 4ℓ
15:08 < L29Ah> > Maximum sweat rates of an adult can be up to 2–4 liters per hour or 10–14 liters per day (10–15 g/min·m2), but is less in children prior to puberty.[3][4][5]
15:08 < L29Ah> okok
15:09 < L29Ah> still seems enough
15:10 < L29Ah> and there should be an observer with a quick-access saline reserve during trials
15:11 < L29Ah> or just blood pressure monitoring every 10 mins or so
15:11 < L29Ah> to terminate the procedure if the blood pressure gets alarmingly low
15:15 < fenn> seems fine to take all day long to do the procedure
15:16 < fenn> a human has 6 liters of blood, some of that is red blood cells, and you probably don't want to get rid of more than half of the plasma at a time
15:17 < fenn> 5 liters*
15:17 < fenn> "Plasma makes up 55% of your blood's total volume"
15:21 < streety> Would you not also need an anti-coagulant to prevent exposure to the tubing and membrane prompting clot formation?
15:23 < L29Ah> streety: seems like vein to vein blood transfusions don't require anticoagulants
15:23 < L29Ah> some aspirin won't hurt i guess
15:24 < muurkha> 14 liters per day is still a lot higher than I would have thought
15:26 < streety> Is that generally true or with specific materials?
15:33 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has left #hplusroadmap []
15:34 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has joined #hplusroadmap
15:40 < TMA> there is probably some medical reason you are not allowed to "donate" plasma more frequently than once per fortnight
15:41 < L29Ah> depends on the country
15:41 < L29Ah> in US: The maximum plasma donation frequency is once in 2-day period, and no more than twice in a 7-day period.
15:43 < L29Ah> and it's not like we're planning to dump plasma more often than once a year or so
15:44 < L29Ah> personally i don't wanna lose all the antibodies too quickly
15:44 < TMA> It can be a matter of economy. (if frequent donations cause more doctor visits afterwards, the incentive might be different in US or here because of the different ways of financing the medical care)
15:52 < L29Ah> for now seems like the biggest hazard is getting sepsis from poor sterility
15:53 < kanzure> for uplift: rat gestation is 3 weeks and breeding age is 10 weeks https://twitter.com/srnorty/status/1633956897123336198
15:53 < L29Ah> although i don't recall exceptional antiseptic procedures in moscow blood donation center
15:54 < kanzure> "Human. I think there is yet hope to make them sentient and civilized." - pmetzger
15:58 < L29Ah> they wiped my arm with alcohol, inserted a single use needle and covered the thing with a wide band-aid
15:59 < L29Ah> i wonder how do they sterilize the plasma-separating membrane
16:00 < L29Ah> i guess some 70% ethanol will do
16:00 < muurkha> so you can get 5 generations per year
16:00 < muurkha> no, almost 4
16:04 < kanzure> that's still too slow. you could do mutants of a clone a bit faster.
16:08 < muurkha> it's pretty slow yeah
16:11 < kanzure> maybe if you had a billion mutant rats.
16:12 < muurkha> the idea is that you can clone them earlier than 10 weeks after birth?
16:16 < L29Ah> Muaddib: how many lines of chat context are you given?
16:16 < Muaddib> L29Ah: Enough to make a witty remark.
16:18 < muurkha> I think it's ten?
17:19 < kanzure> "Those participants with a bachelor’s degree were more likely to agree that gene editing and PGT-P (via IVF) for selecting “smart” embryos were morally acceptable, compared to their less-educated peers, and were more likely to use them for their kids."
17:19 < kanzure> "Peer pressure made a difference. When parents were told that 90% of those similarly situated would opt for PGT-P or gene editing, they were more likely to go along. When others were told that only 10% would use the procedure, most similarly rejected the approach. People are heavily influenced by what they presumed their peer group would do."
17:19 < kanzure> from https://www.acsh.org/news/2023/03/09/designing-super-smart-kids-should-public-have-say-16926
17:19 < kanzure> "Shortly before IVF was made available, a 1969 Harris poll found that most Americans opposed the procedure, and the American Medical Association imposed a moratorium on IVF research. In 1978, one month after Louise Brown, the first IVF baby, was born in the UK, the same poll found that 60% supported IVF and would use it themselves.    "
17:21 < muurkha> I remember in the 01990s in a US middle school and high school that schoolkids would use the term "test tube baby" to deprecate one another
17:21 < muurkha> I think the implication was that being a test tube baby made you "retarded"
17:22 < kanzure> apparently there was a rule that an ethics board had to approve it and then they disbanded the ethics board but kept the rule https://www.science.org/doi/10.1126/science.3393906
17:44 < fenn> "hormonal birth control preserves fertility longer because you aren't ovulating"? is this true?
17:53 < fenn> this is a huge variation (note that it's on a log scale) https://en.wikipedia.org/wiki/Ovarian_reserve?useskin=vector#/media/File:WallaceKelseyModel.jpg
18:36 < fenn> how did IVF ever become a thing? at every step of the process we had nearly identical ethical qualms as with human cloning and germline modification. "only the rich will get IVF" and "eugenics zomg" and "it's not proven safe and never can be" (unless you just do it)
18:36 < fenn> have institutions just become much more risk averse since the 1990s?
18:37 < fenn> specifically was there some reason to believe that IVF wouldn't result in some horrible mutated baby that would suffer for life
18:39 < hprmbridge> kanzure> antonio says JK announced the pregnancies 2 or 3 hours after his article (with speculation about current pregnancies) was posted https://twitter.com/antonioregalado/status/1617870011602980865
18:40 < hprmbridge> kanzure> no idea why we achieved IVF escape velocity.
18:41  * L29Ah hopes to achieve nosepicking escape velocity
18:45 < muurkha> fenn: IVF was mostly invented before progress ended 50 years ago
18:45 < fenn> hm yes the 1969 singularity
18:46 < fenn> i really ought to have a wiki page on this huh
18:56 < muurkha> I was thinking 01973, which is when the first successful IVF oocyte implantation happened.
19:00 < fenn> eesh. "Interviewer: Can you be sure that the three children were neither killed nor sterilized? He: Let's talk about something else.
19:02 < fenn> "possible victims of genetic mayhem" would be a great band name
19:06 < hprmbridge> kanzure> how you gonna sterilize someone these days, isn't that really hard with all the genetic material flying around
19:08 < fenn> uh it's pretty easy
19:08 < hprmbridge> kanzure> we have startups trying to make synthetic human embryos from somatic cells.
19:09 < hprmbridge> kanzure> https://conception.bio/
19:12 < fenn> most cases of involuntary sterilization were to prevent "accidental" pregnancies
19:13 < fenn> "Let's talk about something else."
19:15 < fenn> is it possible that He is the father of lulu and nana?
19:18 < fenn> it's somewhat surprising that 'There is no working "sterilization pill" that causes permanent inability to reproduce.'
19:21 < muurkha> It is, isn't it?
19:24 < muurkha> Hmm, the fact that the first IVF births were in 01978 means that I'm older than all test tube babies, and always will be.
19:25 < hprmbridge> kanzure> that's nice grandpa, now take your sterilization pills
19:25 < hprmbridge> kanzure> jk :/
19:26 < fenn> "1. Anything that is in the world when you’re born is normal and ordinary and is just a natural part of the way the world works.
19:26 < fenn> 2. Anything that's invented between when you’re fifteen and thirty-five is new and exciting and revolutionary and you can probably get a career in it.
19:26 < fenn> 3. Anything invented after you're thirty-five is against the natural order of things."
19:27 < hprmbridge> kanzure> that's why you need to start socializing all the 2068 tech today
19:27 < fenn> ah whoops i just said that. i grepped the wrong logs
19:29 < hprmbridge> kanzure> it's amazing how beyond just inventing stuff, it turns out that nearly ~everything needs a whole social movement or whatever (hopefully this can be reversed in time)
19:29 < fenn> i do think optimistic science fiction has a good role to play in getting people ready for future tech
19:29 < hprmbridge> kanzure> what if each new technology didn't require pulling teeth from the collective's face
19:30 < hprmbridge> kanzure> or what if people just did things and it is fine.
19:30 < fenn> this is why we need to go to mars
19:31 < hprmbridge> kanzure> why is it society's business if someone is a test tube baby or not.
19:32 < hprmbridge> kanzure> it's really not.
19:32 < muurkha> it's really not.
19:32 < fenn> is it society's business if they have a GMO baby?
19:33 < hprmbridge> kanzure> there might be an argument for saying their sexual partners should be informed, or even the child themselves.
19:33 < fenn> uh well yeah
19:33 < fenn> but what about other people
19:34 < hprmbridge> kanzure> no.
19:34 < fenn> but your crazy experiment is going to infect the rest of humanity with cooties
19:35 < muurkha> probably more important to ensure that the sexual partners of carriers of Tay-Sachs or sickle-cell disease should be informed
19:35 < fenn> tay-sachs heterozygotes have some crazy high IQ bonus like 10 points
19:35 < hprmbridge> kanzure> are those partners informed upfront or do they learn after fetal genetic screening
19:36 < fenn> both, because tay-sachs is only highly prevalent in certain communities, which have preemptive screening
19:36 < fenn> it's part of the dating culture apparently
19:39 < fenn> https://en.wikipedia.org/wiki/Prevention_of_Tay%E2%80%93Sachs_disease
19:40 < fenn> this reminds me of the blue eyed islanders problem
19:40 < fenn> if you call the hotline and get an "incompatible" answer, you know you're a carrier
19:46 -!- yashgaroth [~ffffffff@2601:5c4:c780:6aa0:a42b:b972:f1fe:455] has quit [Quit: Leaving]
21:19 -!- alexbfi_ [~alexbfi@dzdkdfyy6mygg63t2xykt-3.rev.dnainternet.fi] has joined #hplusroadmap
21:19 -!- alexbfi [~alexbfi@dzdkdfyy6mygg63t2xykt-3.rev.dnainternet.fi] has joined #hplusroadmap
22:17 < fenn> if having kids is so expensive, why isn't a large family a positional status symbol?
22:54 < muurkha> enjoying leisure time and/or a prestigious job is more desirable
22:56 < muurkha> having a lot of kids puts heavy demands on your time regardless of how much money you have to spend on piano teachers and nannies
--- Log closed Fri Mar 10 00:00:06 2023