--- Log opened Fri Dec 30 00:00:00 2022
01:39 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has quit [Ping timeout: 264 seconds]
02:19 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has joined #hplusroadmap
06:26 < kanzure> protein folding for google colab https://github.com/sokrypton/ColabFold
07:25 < kanzure> 🧬🧬🧬
07:33 < kanzure> .title https://nitter.nl/mimi10v3/status/1608566685258375168
07:33 < saxo> mimi-maxxx (@mimi10v3): 'Lmao that all the big GCs I’m in have had a discussion of “who’s the grimes alt in here?” in the past 24 hours… like, even if you could figure it out would you really want to..?'|nitter
07:39 < kanzure> https://adrianbowyer.blogspot.com/2022/12/phaseshift.html
07:40 < lsneff> I saw that article, trying to think if it would actually work
07:47 < lsneff> I don’t think it would work, emitting antenna have to be smaller than a wavelength of the light they’re emitting
07:48 < lsneff> Otherwise, you just get a bunch of different beams
09:00 < kanzure> .title https://nitter.nl/ElliotHershberg/status/1608553519103238144
09:00 < saxo> Elliot Hershberg (@ElliotHershberg): 'Another recent study from the lab of @MichaelCJewett at @NorthwesternU proposed an approach for using cell-free protein synthesis to produce peptide hormones at the point-of-care.This could eliminate the need for a cold chain (freezers in supply chain) for biologics.'|nitter
09:01 < kanzure> .title https://nitter.nl/ElliotHershberg/status/1608553531086364673
09:01 < saxo> Elliot Hershberg (@ElliotHershberg): 'Last example, this time not from an academic lab but directly in production!Lack of cold chain infrastructure has been a huge bottleneck for distributing mRNA vaccines for COVID.Now @BioNTech_Group is shipping a GMP compliant modular factory:'|nitter
09:03 < kanzure> https://centuryofbio.substack.com/p/atoms-are-local
09:27 < kanzure> .title https://nitter.nl/kanzure/status/1608849763826569217
09:27 < saxo> Bryan Bishop (@kanzure): 'e/acc seems to be mostly about writing banger tweets and not so much about, you know, achievement in effective acceleration. prove me wrong.'|nitter
09:27 < kanzure> i guess every social movement needs something for the dulls to do (make funny tweets)
09:31 < kanzure> .title https://nitter.nl/evanjconrad/status/1608556607117590528
09:31 < saxo> evan conrad is unavailable for a few days (@evanjconrad): 'its genuinely amazing you can just put a website on the internet without asking anyone There’s 100% an alternate path in which you would need to get an “internet license” to get an IP address'|nitter
10:07 < superkuh> That will happen when the major corporate browsers drop HTTP 1.1 support.
10:07 < superkuh> At least re: the web.
11:16 -!- cthlolo [~lorogue@77.33.23.154.dhcp.fibianet.dk] has joined #hplusroadmap
11:29 < L29Ah> > you would need to get an “internet license” to get an IP address
11:29 < L29Ah> i'd say ASNs are very much like "internet licenses"
11:49 < muurkha> good point
12:03 < nsh> yeah i used to think about low level tiering class structure a lot in early days of internet
12:03 < nsh> how things like cloudflare gained dominance is exactly via that 
12:03 < fenn> weird that this hasn't been trumpeted from every rooftop. no updates since 2010?
12:03 < fenn> .wik griffithsin
12:03 < saxo> "Griffithsin is a protein isolated from the red algae Griffithsia.  It has a 121-amino acid sequence which exhibits a Jacalin-like lectin fold." - https://en.wikipedia.org/wiki/Griffithsin
12:04 < nsh> a lot of this dynamics is less overtly discussed
12:04 < fenn> "It has been shown in vitro to be a highly potent HIV entry inhibitor. It is currently being investigated as a potential microbicide for use in the prevention of the transmission of HIV. [3]
12:04 < fenn> Griffithsin shows a broad spectrum ability to bind to the glycoproteins of other viruses, such as the coronavirus."
12:04 < fenn> nsh: out of sight, out of mind
12:05 < fenn> cloudflare routinely blocks people like me for lame reasons
12:05 < fenn> with a proudly displayed logo and everything
12:05 < fenn> i wouldn't notice some other collusion of ISPs behind the scenes
12:06  * nsh nods
12:06 < nsh> it's often the case that the louder and more obvious that politics is happening the less actually it is determining the course of events
12:06 < nsh> real decisions being made in smoky back rooms being the cliche
12:07 < nsh> all sleight of hand
12:07 < nsh> or necessary theatre 
12:09 < fenn> "These studies support further development of GRFT as a systemic antiviral therapeutic agent against enveloped viruses, although deimmunizing the molecule may be necessary if it is to be used in long-term treatment of chronic viral infections."
12:09 < fenn> GRFT = griffithsin
12:10 < fenn> i didn't even realize twitter had group chats
12:10 < fenn> seems like a bad idea for the platform overall
12:11 < kanzure> when e/acc talks about group chats i am pretty sure they mean signal or discord
12:11 < kanzure> maybe telegram.
12:13 < lsneff> it does have gcs, but yeah, they usually mean discord or signal
12:13 < lsneff> people love talking about how "it's time to build" and such
12:14 < kanzure> ok what have they built
12:14 < kanzure> show me
12:14 < fenn> gingerbread castles
12:14 < lsneff> that's what i'm sayinh
12:14 < lsneff> they love talking about it
12:15 < lsneff> they may or may not have actually built anything
12:15 < kanzure> :|
12:15 < fenn> it's signalling that you're against postmodernism
12:15 < kanzure> batsignaling for elon
12:16 < lsneff> ea is one of the few groups that actually turned into physical action irl
12:16 < lsneff> e/acc may or may not, remains to be seen
12:18 < kanzure> what precisely have the altruists achieved
12:18 < lsneff> I would say a lot of the ML field has drawn from offshoots of ea
12:19 -!- Llamamoe [~Llamamoe@46.204.76.114.nat.umts.dynamic.t-mobile.pl] has joined #hplusroadmap
12:20 -!- yashgaroth [~ffffffff@2601:5c4:c780:6aa0::4249] has joined #hplusroadmap
12:21 < fenn> kanzure if EA had saved 10 million peoples' lives would you even know about it?
12:21 < kanzure> would they?
12:22 < fenn> knowledge is not evenly distributed
12:22 < fenn> maybe some would and some wouldn't
12:22 < kanzure> maybe if you just smile a bit more you can save 5 million people/year
12:23 < kanzure> actually how do charities do lives-saved attribution anyway?
12:23 < fenn> that is literally the whole point of EA
12:23 < kanzure> the churches will tell you the attribution goes to god
12:24 < fenn> they have spreadsheets of QALYs per dollar estimates for each charity org
12:24 < kanzure> well i'm looking at a bunch of aggregate statistics and i get to take credit for trend reversals
12:24 < fenn> yes, if you're responsible for funneling money to the right place and it causally reverses the trend
12:25 < fenn> do you believe in causality?
12:26 < kanzure> i don't believe in altruists
12:27 < fenn> unfortunately EA has gone off the rails with "cause areas" prioritizing a particular cause instead of raw QALY optimization, because the singularity fucks up all the models
12:31 < fenn> ok this is just bad. there appears to be no obvious public-facing list of charities and their impact per dollar
12:31 < fenn> there used to be such a list, for example https://www.givewell.org/international/technical/programs/2009-version
12:31 < lsneff> that is bad
12:33 < muurkha> postmodern architects built plenty
12:34 < fenn> https://a16z.com/2020/04/18/its-time-to-build/
12:35 < muurkha> what's e/acc?
12:35 < muurkha> from the context I'm thinking something like "effective accelerationists" or "elon and accordants"
12:35 < fenn> the former
12:36 < lsneff> effective accelerationism
12:36 < fenn> way too much cultural cachet is being loaded on elon
12:36 < lsneff> i don't actually think e/acc has anything to do with elon, they just use him as an example of someone who builds, I guess
12:37 < lsneff> whether that's true or not is unclear
12:40 < fenn> are cell-free systems better than just growing a modified tobacco plant from seed? faster? less purification needed or something?
12:41 < fenn> i'm imagining a packet of seeds you let sprout overnight and they ooze your desired biomolecule into the water
12:41 < kanzure> that's a question for yashgaroth
12:41 < fenn> muurkha: "The problem is desire. We need to *want* these things. The problem is inertia. We need to want these things more than we want to prevent these things."
12:41 < L29Ah> fenn: you don't ooze proteins
12:42 < yashgaroth> plants aren't good at secreting proteins, and we're even less good at engineering it into them
12:43 < yashgaroth> which one's better heavily depends on your specific application
12:46 < kanzure> wait are there really no plants that ooze
12:46 < fenn> forgotten history: "Leaf Protein International (LPI) was founded in 1979 ... tobacco F-1-p may be the best nutritional and functional food protein.  1982, LPI demonstrated that the extraction process for obtaining crystalline F-1-p and other raw materials from tobacco was commercially feasible. The yield of F-1-p was estimated at 530 lbs/acre (600 kg/hectare). LPI's goal was to transform tobacco
12:47 < fenn> into a commodity used on a large scale for human food"
12:50 < fenn> this is comparable with soybeans
12:50 < yashgaroth> usually they ooze small molecules, probably some (eg carnivorous plants) have decent secretory pathways but that doesn't scale well. Especially vs yeast or something
12:52 < yashgaroth> should be easy to knock out nicotine production from tobacco, but leaf-based protein doesn't store well compared to a dried soy bean
12:53 < fenn> why can't you dry the protein or if that's not immediately possible dry the leaves
12:54 < fenn> i'm surprised that they used protein crystallization and claimed to be commercially feasible
12:55 < fenn> these quotes are from https://www.acsh.org/news/1992/01/01/food-from-tobacco-a-well-kept-secret
12:55 < yashgaroth> if you have to extract it that'll make the cost uncompetitive with soybeans, similarly you can dry the leaves (we already do for smoke leaf production) but soybeans already do it themselves in the field
12:55 < yashgaroth> yeah a crystal food would be sick but like lab-grown meat, not economically viable
12:57  * L29Ah ingests a crystal of sucrose
12:57 < fenn> it's a rare human that eats unprocessed soybeans
12:57 < L29Ah> surprisingly cheap pure-ish chemical
12:57 < L29Ah> organic even!
12:58 < yashgaroth> haha fair
12:58 < fenn> made of mostly carbon
12:58 < yashgaroth> isn't tofu just milled soybeans
12:58 < fenn> no it's extracted and re-coagulated and then pressed and aged
12:59 < fenn> well, usually not aged i guess
12:59 < fenn> i've made tofu from soybeans, it's not hard, but it is something
12:59 < yashgaroth> tbf the vast majority of soybeans are for animal feed, and they don't care much
13:03 < yashgaroth> the question is, will nicotine knockout tobacco be immediately destroyed by every insect
13:04  * L29Ah applies DDT to yashgaroth
13:05 < fenn> you don't need to do a nicotine knockout
13:05 < L29Ah> also i'm pretty sure you don't care much about nicotine if you're going to extract your specific protein anyway
13:06 < yashgaroth> I'm working on the assumption that we'd be feeding dried leaves to livestock, as the extraction costs would be infeasible
13:07 < fenn> i think the idea is more of a dietary supplement or processed food additive like whey protein
13:08 < yashgaroth> will see if I can find their extraction method for the pilot plant, but nicotine and protein are both hydrophilic and nicotine is extremely toxic (I say as I inhale some)
13:08 < fenn> lol
13:08 < yashgaroth> at least whey protein is floating around freely in water, whereas with tobacco protein it's locked up inside cells so there's probably mechanical maceration involved
13:09 < fenn> yeah i found this because i remembered something about how easy it was to get tobacco to secrete protein into its growth medium
13:11 < fenn> another paper i just found "(SP)32 tag (32 tandem repeats of "Ser-Pro" motif) ...  generated a high secreted protein yield (125 mg/L) with a low cell biomass accumulation (~7.5 gDW/L)"
13:12 < fenn> now 125 mg/L is not great for food production but may work fine for vaccines
13:12 < yashgaroth> implying cell tissue culture if there's growth medium. That's been a sticking point for mammalian, since adherent cells are much happier and more productive but don't scale vs suspension cells in a bioreactor
13:13 < fenn> well i dunno what to call the water that you grow a plant in
13:13 < yashgaroth> you end up with massive stacks of t-flasks, I used to work with a device we called the rollercoaster that held thousands of them for trypsinization. The size of a large van
13:13 < fenn> is "tobacco hairy root culture" really all that different than a plant with roots?
13:15 < fenn> think hydroponics, not tissue culture
13:15 < fenn> remember the context here is talking about "scaling out" biology to the edges of industrial infrastructure
13:15 < yashgaroth> yeah definitely not in soil, imagine the topsoil loss from extracting root systems from the ground
13:16 < yashgaroth> hydroponics though yeah it might be workable
13:19 < yashgaroth> again it depends on the specific application, yeast eat sugar and divide every couple hours in a scalable bioreactor and they can handle most of the same post-translational modifications as plants
13:19 < fenn> fish tanks have bubble column filters which skim off protein in the water. maybe something like that could be used so the hydroponics don't get funky bacterial growth
13:20 < fenn> so anyway, cell free systems are a one-shot sterile thing that can sit on a room temperature shelf until needed
13:20 < fenn> is that an accurate portrayal?
13:21 < yashgaroth> nah they're protein solutions stored at -80
13:22 < yashgaroth> I wouldn't want to try lyophilizing a ribosome
13:22 < fenn> alright then i don't understand this tweet
13:22 < fenn> .tw https://nitter.nl/ElliotHershberg/status/1608553519103238144
13:22 < saxo> Another recent study from the lab of @MichaelCJewett at @NorthwesternU proposed an approach for using cell-free protein synthesis to produce peptide hormones at the point-of-care. // This could eliminate the need for a cold chain (freezers in supply chain) for biologics. https://pbs.twimg.com/media/FlK65YJagAApacc.jpg (@ElliotHershberg, in reply to tw:1608553509234040833) psa from .tw cmd: fuck twitter
13:24 < fenn> .title https://pubs.acs.org/doi/10.1021/acssynbio.2c00392
13:24 < saxo> Just a moment...
13:25 < yashgaroth> getting a therapeutic dosage from cell-free is also laffo, "heat induced purification" like ok I guess you can denature ribosomes easily enough but no way that plus a single-step polish will get to injectable levels of quality
13:25 < fenn> "Here, we identify maltodextrin as an additive to cell-free reactions that can act as both a lyoprotectant to increase thermostability and a low-cost energy substrate. ... We anticipate that our low-cost, thermostable cell-free glycoprotein synthesis system will enable new models of medicine biosynthesis and distribution that bypass cold-chain requirements."
13:26 < yashgaroth> maybe for a vaccine since quantity is lower and purity matters less (though try convincing the FDA of that), but even so
13:30 < lsneff> i had essentially zero problem focusing really well this fall but getting back into my undergrad thesis is exposing my inability to actually do anything
13:31 < fenn> it's time to build lsneff
13:32 < lsneff> many such times
13:32 < yashgaroth> I'll give it more of a read, but initially: lyophilization is hard to scale, and they're using E. coli ribosomes which are probably more resilient but extremely incompatible with human injection without extensive purification. Is the plan to have a bunch of shipping containers sitting around in remote areas in case a new pandemic breaks out? Seems like it'd be easier to lyophilize the vaccine
13:33 < fenn> yes the plan is to have a bunch of shipping containers sitting around :(
13:34 < yashgaroth> I'm not always immediately suspicious when someone from Stanford posts fanciful sci-fi with cool cgi renders, but
13:39 < yashgaroth> also the amount of QC to make an automated container factory pass inspection is mind-boggling, I assume there aren't GMP-qualified engineers in a separate shipping container twiddling their thumbs between outbreaks. GE makes containerized manufacturing facilities which look neat in renders
13:43 < fenn> it's third world so anything goes :\
13:43 < fenn> like formula-1 for biology
13:53 < yashgaroth> probably be easier to ship them some freezers
13:54 < yashgaroth> (and batteries and solar panels)
14:10 < muurkha> yashgaroth: what kind of inspection?  food safety inspection?
14:10 < saxo> muurkha: <clsn> Re max number of sequences: WFM, what can I say?
14:16 < yashgaroth> muurkha, proving that the entire factory can work reliably and automatically after several years of sitting in the kalahari, once some guy drives up and injects the vaccine mRNA code into it. And then dispensing a drug product than can just go into people's arms without quality control. Whereas the standard nowadays is years of setting up a facility that generates a million pages of documentation before opening, and generating 
14:16 < yashgaroth> thousands more after every batch, most of which is made and cross-checked by engineers and quality technicians
14:19 < yashgaroth> versus the current method of doing all that in a semi-centralized location and passing it off to some guy driving a jalopy loaded with dry ice for delivery
14:20 < yashgaroth> hopefully some MPH has done a study of how that turned out for covid vaccine distribution
14:24 < muurkha> yashgaroth: yeah, probably the places that decide not to allow that kind of thing will lose out
14:24 < muurkha> or you'll have to do it secretly
14:25 < yashgaroth> mhm
14:26 < muurkha> there's still the problem of getting it to work at all and finding out whether the result is okay
14:28 < yashgaroth> if regular vaccine manufacturing operated flawlessly I'd be more optimistic, but trust me it does not
14:30 < muurkha> it'll surely be difficult
14:30 < muurkha> I don't think GE is going to be able to manage it
14:31 < muurkha> maybe they should start with automated manufacturing of something simpler than vaccines
14:33 < yashgaroth> oh GE doesn't make automated factories, it's just a bunch of regular equipment shoved into containers that technicians operate
14:33 < yashgaroth> mostly meant to save on lead time for construction and fitting-out
15:02 -!- cthlolo [~lorogue@77.33.23.154.dhcp.fibianet.dk] has quit [Read error: Connection reset by peer]
15:53 -!- Llamamoe [~Llamamoe@46.204.76.114.nat.umts.dynamic.t-mobile.pl] has quit [Quit: Leaving.]
16:30 -!- A_Dragon is now known as gAy_Dragon
16:31 -!- gAy_Dragon [A_D@libera/staff/dragon] has quit [Killed (Stx (You knew this was coming :-) Happy new year!))]
16:31 -!- A_Dragon [A_D@libera/staff/dragon] has joined #hplusroadmap
16:32 -!- A_Dragon is now known as gAy_Dragon
17:45 -!- yashgaroth [~ffffffff@2601:5c4:c780:6aa0::4249] has quit [Quit: Leaving]
17:53 < nmz787> https://youtu.be/mN1uAVn0RwA
17:53 < Muaddib> [mN1uAVn0RwA] Clearing forest clear-cut brush for replanting seedlings --- Alibaba.com mini-ex --- (0:44)
18:06 < kanzure> tree harvesters https://www.youtube.com/watch?v=dyQMgoJjxWc
18:06 < Muaddib> [dyQMgoJjxWc] World's Modern Long Reach Excavator Machine Working - Heavy Equipment Cutting Big Tree Machine (11:26)
18:10 < kanzure> .wik harvester (forestry)
18:10 < saxo> "A harvester is a type of heavy forestry vehicle employed in cut-to-length logging operations for felling, delimbing and bucking trees." - https://en.wikipedia.org/wiki/Harvester_(forestry)
18:27 -!- luna_ [~luna@user/luna/x-4729771] has quit [Remote host closed the connection]
18:27 -!- luna_ [~luna@2a01:4c8:a3:acdf:69d9:bf2b:169f:d564] has joined #hplusroadmap
18:27 -!- luna_ [~luna@2a01:4c8:a3:acdf:69d9:bf2b:169f:d564] has quit [Changing host]
18:27 -!- luna_ [~luna@user/luna/x-4729771] has joined #hplusroadmap
18:33 -!- luna_ [~luna@user/luna/x-4729771] has quit [Ping timeout: 265 seconds]
18:56 -!- luna_ [~luna@2a01:4c8:a3:acdf:f18e:c8cc:b9fc:ae9] has joined #hplusroadmap
18:56 -!- luna_ [~luna@2a01:4c8:a3:acdf:f18e:c8cc:b9fc:ae9] has quit [Changing host]
18:56 -!- luna_ [~luna@user/luna/x-4729771] has joined #hplusroadmap
20:48 < lsneff> maaku: I’d love to chat in depth at some point about why our opinions on physical vs digital revival from cryonics differ so much
22:45 -!- jrayhawk [~jrayhawk@user/jrayhawk] has quit [Ping timeout: 260 seconds]
22:57 -!- jrayhawk [~jrayhawk@user/jrayhawk] has joined #hplusroadmap
--- Log closed Sat Dec 31 00:00:01 2022