--- Log opened Tue Feb 22 00:00:06 2022
00:42 -!- catalase [catalase@stable.bnc.by.bnc4you.xyz] has quit [Quit: Stable ZNC by #bnc4you]
00:51 -!- catalase [catalase@stable.bnc.by.bnc4you.xyz] has joined #hplusroadmap
00:56 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has joined #hplusroadmap
02:29 -!- Malvolio [~Malvolio@user/malvolio] has quit [Ping timeout: 272 seconds]
05:19 < kanzure> nsh: i think the name i was looking for was instapaper?
05:19 < kanzure> and stuff like https://mercury.postlight.com/
05:19 < nsh> ah, both new to me i think and ty
05:48 -!- flooded is now known as _flood
06:34 < nsh> .t https://physicsworld.com/a/complex-systems-approaches-to-information-processing/
06:34 < saxo> Complex systems approaches to information processing – Physics World
08:22 < nsh> good example of a url where reader mode button works: https://sam4k.com/a-dummys-guide-to-disclosing-linux-kernel-vulnerabilities/
09:08 < fenn> some alt-modernist propaganda
09:08 < fenn> "Remember when gas was only 25-cents a gallon? You could take a dollar down to the gas station and buy four gallons for a buck! At that time our dollar was backed by silver - real money. Guess what? That same amount of silver still buys four gallons of gas today! That just proves that real money like gold and silver holds its value and it is the US dollar that buys less and less. As a matter of
09:08 < fenn> fact, when you think about it, you realize that gas, food, and almost everything else has NOT gotten more expensive. It only seems that way because the value of the US dollar is worth less and less"
09:09 < fenn> or maybe silver has just gotten more expensive, though i don't see why that would be the case
09:09 < fenn> it isn't being used up in an irreversible process as far as i know
09:10 < muurkha> well, gas *has* gotten more and more expensive
09:10 < muurkha> if we're talking in terms of EROEI
09:11 < muurkha> but both oil and silver are relatively volatile markets
09:19 < muurkha> https://www.macrotrends.net/1470/historical-silver-prices-100-year-chart says silver has gone beyond US$16/ounce and US$30 over the last five years, and https://www.macrotrends.net/1369/crude-oil-price-history-chart shows oil going beyond the range US$21/bbl to US$91/bbl
09:19 < muurkha> over the same time period
09:20 < muurkha> and the biggest drop in the price of oil during that time, when covid hit the US, was also the timing for the biggest rise in the price of silver
09:23 < muurkha> so during early 02020 an ounce of silver bought amounts of crude oil ranging from roughly 50 liters to roughly 200 liters
09:25 < muurkha> ugh, sorry, that was inflation-adjusted, but presumably they use the same deflator for both series, and the inflation from 02020 is pretty mild anyway
09:28 < muurkha> in, say, 01960 oil was US$2.97/bbl (not inflation-adjusted!) and silver was US$0.92/ounce, so at that point an ounce of silver bought roughly 50 liters.  now it's down to 40
09:33 < muurkha> I think silver probably *has* gotten more expensive, both in terms of the effort and resources required to extract an ounce of it (because, as with oil but to a smaller degree, now we're exploiting even less concentrated deposits than 60 years ago) and in terms of the other goods and services you have to forgo in order to afford an additional ounce of silver (fresh vegetables in midwinter, trips on 
09:33 < muurkha> jet planes, pocket supercomputers, IKEA bookshelves)
09:33 < nmz787_> maaku: I guess Micron is another US brand, not sure about SSD branding (I haven't bought one in 5-ish years)
09:37 < nmz787_> Western Digital, Seagate... I think are both US brands too
09:39 < muurkha> Seagate: "Since 2010, the company has been incorporated in Dublin, Ireland, with operational headquarters in Fremont, California, United States."
09:40 < muurkha> WDC is in San José (United States)
09:40 < lsneff> What’s the etiquette for setting up an introduction between someone and someone they might want to work for?
09:41 < lsneff> Do I email the person I’m introducing someone to first?
09:49 < nsh> a letter of introduction is usually a one-way transaction between an endorsing party a service provider and a prospect
09:50 < nsh> s/service provider/employer/
10:00 < lsneff> Yes, I'm aware of that. I'm just curious if I should check with the employer before setting up the intro.
10:02 < nsh> probably not but might depend a bit on the people i guess
10:03 < nsh> you're just adding some weight to a decision 
10:21 < kanzure> lsneff: often you ask both parties if they are okay and then you make the intro
10:21 < kanzure> sometimes optional
10:21 < lsneff> Okay, I'll ask the employer then, thank you.
10:54 < xaete[m]> re: that proprietary/closed source reaction database; it was probably reaxys, that's the most common one i've heard of
11:02 < muurkha> oh thanks, xaete[m] 
11:03 < xaete[m]> and yea iirc the 4 thieves guy was p. dumb, u can make basically anything from immediate precursors that r commercially available, but u can also just buy the chemical in bulk from china in most cases
11:03 < xaete[m]> and in most cases thats what you're gonna want if you want a chemical, time is money and so on
11:18 -!- Codaraxis [~Codaraxis@user/codaraxis] has joined #hplusroadmap
11:21 < xaete[m]> fenn: there's build instructions and CAD drawings n shit for the reactor thing in the SI here https://www.nature.com/articles/s41557-020-00596-9#Sec24
11:25 < muurkha> is that the original Chemputer paper?
11:26 < xaete[m]> i believe so
11:26 < xaete[m]> or at least it was linked in the twitter thread
11:46 < xaete[m]> another way to look @ stuff like this is that it's trying to scale /down/ the sorta fancy automated equipment from chemical engineering to benchscale
11:52 < muurkha> that's an interesting perspective
11:53 < muurkha> my thought was that bulk plants can use specialized equipment optimized to, e.g., support a particular catalyst for a particular reaction at a particular temperature and pressure
11:54 < muurkha> while benchwork is all about using very generic equipment like heating mantles, boiling flasks, filters, separatory funnels, and rotovaps
11:54 < muurkha> and the Chemputer approach is an attempt to go even further in that direction
12:07 -!- Malvolio [~Malvolio@user/malvolio] has joined #hplusroadmap
12:57 < xaete[m]> w. the chemputer thing tho they use /alot/ of v. general equipment to support custom synths
12:57 < xaete[m]> and also ur correct (generally) that industry has specialized equipment, but w. pilot plant scale stuff that isnt true
12:58 < xaete[m]> like u can (if you have like 10k at least) buy a fully generic temp-controlled 100L reactor, which also works as a sep funnel, or buy a totally-generic like 50L filter funnel
12:59 < xaete[m]> chromatography is harder (mostly because you need absolutely massive chromatography columns) but for sth small scale like this you could use a cyclograph or somethin
12:59 < xaete[m]> although ofc identifying fractions wouldnt be perfect
13:02 -!- spaceangel [~spaceange@ip-78-102-216-202.net.upcbroadband.cz] has joined #hplusroadmap
13:04 < muurkha> hmm, that's interesting
13:04 < muurkha> "fully generic" I doubt though
13:05 < muurkha> are you saying it can handle *both* 1600° molten oxide electrolysis *and* room-temperature triflic acid *and* cryogenic reactions with liquid ozone?
13:07 < muurkha> what range of conditions *do* they tolerate?  can you give me a brand name?
13:07 < xaete[m]> well... "fully generic" as in "this works for any reaction that primarily consists of mixing, heating, cooling, and/or dropwise addition between something like -80 to 250 degrees and pressures between 1atm and 5kPa"
13:07 < xaete[m]> so not /everything/, just a high range
13:07 < xaete[m]> so lets see... one of the 3 you listed?
13:07 < muurkha> that's pretty nifty
13:08 < muurkha> probably teflon-coated instead of pyrex to handle that kind of pressure?
13:08 < xaete[m]> (for brand names i think chemglass sells em)
13:08 < xaete[m]> (you can find them on ebay/aliexpress/alibaba tho, just search "xL jacketed reactor" where x is 2,5,10,etc. normally goes up to 100, tho i think ive seen a 200 before)
13:08 < xaete[m]> ,title https://www.alibaba.com/product-detail/Jacketed-Reactor-Manufacturer-Sale-High-Effective_1600266269160.html
13:08 < muurkha> aha, thanks!
13:08 < xaete[m]> is one
13:09 < muurkha> that looks like it's glass, though
13:09 < xaete[m]> (of course, they also make them in Much Smaller, which is what you want for benchtop work)
13:09 < xaete[m]> and yea, its all glass
13:09 < muurkha> amazing that you can get a 100ℓ glass reactor that can handle 5kPa
13:09 < xaete[m]> all "generic chem" setups are gonna be some combo of glass and teflon just because of chemical compatibility
13:10 < xaete[m]> this one claims 2kPa, which i dont really doubt from a mechanical standpoint, only issue is how well they can deal with gas leaking in
13:10 < muurkha> (not saying I doubt it!  I just had no idea such things existed, but that's hardly surprising given my level of ignorance)
13:11 < xaete[m]> (i mean, ive never used one, but yea)
13:11 < xaete[m]> dw abt not knowin stuff, this is p. niche stuff i think
13:11 < muurkha> glass probably excludes reactions using fluorosulfonic acid and molten lye, but those are also pretty niche
13:11 < xaete[m]> plus we all have our times... i'm pretty rusty w. my bio stuff
13:11 < muurkha> I think it would be more interesting to figure out how we could scale this stuff *down*
13:12 < xaete[m]> and yes i woudlnt recommend running molten lye reactions in this thing, nor your 200 bar hydrogenation kek
13:12 < muurkha> haha
13:13 < muurkha> suppose manufacturing cost was no longer an obstacle, how would you do generic reactions at picoliter scales rather than kiloliter?
13:13 < xaete[m]> the generic answer to this is "microfluidics"
13:13 < muurkha> right, of course, but I mean how do you design the reactions?
13:14 < xaete[m]> the less-generic answer is ig more, "why are you doing chemical synthesis on the pL scale?"
13:14 < muurkha> temperature becomes a much bigger obstacle, because you can't vary temperature by gigakelvins per meter
13:14 < xaete[m]> seems v.v. hard to get useful/repeatable results there
13:14 < xaete[m]> re: resigning reactions?
13:14 < xaete[m]> *designing
13:14 < xaete[m]> on the pL scale you're probably going to be doign combichem of some sort
13:15 < muurkha> but you can vary pressure and electrode potentials enormously
13:15 < xaete[m]> so unless you're running 30k different reactions on the same chip (dont do this)
13:15 < xaete[m]> i'd say just run them all the exact same way, so do 30k steglich esterifications at once, or 30k palladium cross-coupling reactions, and so on
13:15 < muurkha> yeah, 30k different reactions on the same chip, or 300k, or 3M, could be super useful — but only if you *can* get repeatable results
13:16 < xaete[m]> realistically speaking you're going to want to run the same reactions anyways, uses for this would be something liek testing substituents on a small-molecule core for instance
13:16 < muurkha> well, that's mostly a matter of manufacturing cost, isn't it?
13:17 < xaete[m]> well other thing is that you might be limited by reagent selectivity
13:17 < xaete[m]> good luck asking sigma to sell you a bajillion different chemicals
13:17 < xaete[m]> i guess you could use combichem to make a massive chemical library in situ, but then you run to the same issue on the other end,etc.
13:17 < muurkha> I mean the standard process for producing ethanol uses 0.04 picoliter reaction vessels
13:17 < muurkha> and has for thousands of years
13:18 < xaete[m]> basically, not super sure if scaling down to pL is the best way to go here
13:18 < xaete[m]> and well. i guess you /are/ correct about the ethanol production process
13:20 < xaete[m]> but lets see.... to make this "useful" you'd want a fully-end-to-end connected setup
13:20 < muurkha> yeah, to avoid the sigma problem
13:20 < muurkha> especially if sigma doesn't ship to mars
13:22 < xaete[m]> such a small scale implies combichem of some sort, so you want jacks to pipe in your "big" reactants (stuff like your organic building blocks, your fluorophenols or boronic acids or whatever), a few "bigger" bottles for your coupling reagants or catalysts or whatever (if we're not doing green chem, you can fill these up with stuff like SOCl2, acetic anhydride, various grignard reagents and so on and so forth. if you have safety concerns,
13:22 < xaete[m]> best to use something nicer)
13:23 < muurkha> safety concerns are less significant at the picoliter scale than at the kiloliter or megaliter scale
13:23 < muurkha> also, on mars
13:23 < xaete[m]> then i suspect you're going to want some form of soild-phase synthesis, so have each individual well be prepared to attach your first substrate on, run a coupling step to connect it, then flow through the other chemicals
13:24 < xaete[m]> if youre doing this /on mars/ its pretty much the same, maybe a bit easier even due to low gravity
13:24 < xaete[m]> if youre doing it in space just add enough inert air lines to make sure your liquids are all going one way
13:25 < muurkha> yeah
13:25 < xaete[m]> but maybe you can run the substrate to be coupled in the vertical direction, then the chemicals in the horizontal one... seems like the easiest way to make a matrix out of two sets of chemicals
13:25 < muurkha> you could maybe do centrifugal distillation in space but no living cells do distillation on earth afaik
13:25 < muurkha> gravity is too weak at that scale
13:25 < xaete[m]> eh, no need for distillation on this scale
13:26 < xaete[m]> we're going to be doing all solvent chem, and selecting the reactions carefully so all byproducts are soluble
13:26 < muurkha> you still need to separate things sometimes
13:26 < xaete[m]> since your "product" is still attached to substrate, you just rinse everything else off
13:26 < xaete[m]> if you need to separate things... maybe you could build a pL-scale chromatography column, but i would really prefer not to
13:27 < xaete[m]> distillation is definitely out, i have no idea how you would even do that, maybe liquid-liquid extraction?
13:27 < muurkha> then how do you separate the liquids?
13:28 < muurkha> surface tension I guess
13:28 < xaete[m]> Very Carefully basiaclly
13:28 < muurkha> heh
13:28 < xaete[m]> but yea that leads to the biggest "issue" i see with this which is... whatre you going to do with these tiny amounts of chemicals? which means you have to do a layered chip/highly integrated approach
13:29 < xaete[m]> so you drop em directly onto the detectors/biological screening arrray/whatever that you wanna do
13:29 < xaete[m]> on the uL or mL scale you'd have alot more freedom, but dropping down 9 orders of magnitude means that making use of the chemicals is Hard
13:30 < muurkha> well, if manufacturing cost is not a problem, you can just build however many reaction cells you need to get the amount of materials you want
13:30 < xaete[m]> hm.. how much do you want to make at the end?
13:31 < xaete[m]> parallelism always demands a cost after all
13:31 < muurkha> a zettagram seems like a good initial goal
13:32 < xaete[m]> if you want 1e21 g you're gonna want a base layer wayyy larger than pL. the chip i'm describing is hugely wasteful due to the fact that on the pL scale waste doesnt mater at all
13:32 < xaete[m]> so we'd flow through things like 10x equivalents just to make sure everything is filled, etc.
13:33 < xaete[m]> the smallest scale i think we could do "do this and multiply a bajillion times to scale up to arbitrary high amounts" is probably on the uL or mL scale, but even the uL scale is sketchy
13:33 < muurkha> the terrestrial biosphere is already on the order of hundreds of zettagrams, almost entirely in sub-picoliter reaction vessels
13:33 < muurkha> so we already have a feasibility proof
13:34 < xaete[m]> well at that point you're just talking about engineering artificial life... in which case i'd go completely differently than if i were designing a combichem setup box for small scale screening
13:34 < muurkha> yeah, being hugely wasteful in that context is bad
13:35 < muurkha> but I don't know if "engineering artificial life" is a very good way to think about it
13:35 < xaete[m]> well if you can find me some other pL-scale arbitrary catalyst synthesizer i'll call it something else :P
13:36 < muurkha> the next-generation industrial infrastructure will have some things in common with living cells, but will differ in other ways
13:36 < xaete[m]> (and at the same time... whenever we want to leverage those tiny reactors, we tend to just think of them as just another reactant in dm^3-m^3 scale reactors anyways)
13:36 < muurkha> but calling it "artificial life" is potentially as misleading as calling computers "giant electronic brains" or submarines "synthetic fish"
13:38 < xaete[m]> hm tru, but at the same time if you're looking for devices to do the same thing /as/ those, it might be better to change the sense of scale
13:38 < xaete[m]> after all, planes fly and are much larger than birds and bugs
13:38 < muurkha> possibly, yes
13:38 < kanzure> https://en.wikipedia.org/wiki/Retrosynthetic_analysis
13:38 < muurkha> the reason I think fine scale is important is that learning increases with feedback speed
13:39 < xaete[m]> i suspect w. modern equipment due to deadspace concerns you can't really get universal- or universal-like synthesis capability on an artifical scale below around the 1mL mark
13:39 < xaete[m]> wrt feedback concerns, i think the "how" of reactions is actually much less important than the "what" of reactions
13:39 < xaete[m]> run one grignard, steglich esterification, or suzuki and you've run em all
13:39 < muurkha> also, I think nanostructured metamaterials are likely to be important
13:39 < muurkha> heh
13:40 < xaete[m]> but the point is the chemical space you can get /by/ running those reations
13:40 < muurkha> well, up to a point?
13:41 < xaete[m]> (wellll also as long as the molecule is flat enough and sterochemistry isnt too big of an issue)
13:41 < xaete[m]> (but hey nature gets by with only 20 amino acids so)
13:41 < muurkha> also, I think conflict is likely to continue, and larger-scale infrastructure is a more appealing target for disruption
13:41 < muurkha> the humans have just been through 75 years of relative peace
13:41 < xaete[m]> hm. in this case you're going to be directly participating in the scale-efficenciy tradeoff more or less
13:42 < muurkha> that is a potential problem, especially with reactions that rely on temperature changes
13:42 < xaete[m]> true, although i haven't seen anyone attack a refinery yet (fingers crossed)
13:43 < xaete[m]> but like, there are economies of scale you can get at the MT scale you can't really rely on on the lab scale; at the same time, scaling up lab reactions is hard
13:44 < muurkha> .t https://en.wikipedia.org/wiki/Abqaiq%E2%80%93Khurais_attack
13:44 < saxo> "On 14 September 2019, drones were used to attack oil processing facilities at Abqaiq (Arabic: بقيق, romanized: Buqaiq) and Khurais (خريص) in eastern Saudi Arabia." - https://en.wikipedia.org/wiki/Abqaiq%E2%80%93Khurais_attack
13:44 < muurkha> "Both facilities were shut down for repairs, cutting Saudi Arabia's oil production by about half – representing about five percent of global oil production"
13:44 < xaete[m]> ah shit, shouldve known
13:45 < muurkha> .t https://en.wikipedia.org/wiki/Operation_Tidal_Wave#Second_Steaua_Rom%C3%A2n%C4%83_attack
13:45 < saxo> "Operation Tidal Wave was an air attack by bombers of the United States Army Air Forces (USAAF) based in Libya on nine oil refineries around Ploiești, Romania on 1 August 1943, during World War II." - https://en.wikipedia.org/wiki/Operation_Tidal_Wave
13:46 < muurkha> .t https://en.wikipedia.org/wiki/Operation_Boomerang#Attack
13:46 < saxo> "Operation Boomerang was a partially successful air raid by the United States Army Air Forces' (USAAF) XX Bomber Command against oil refining facilities in Japanese-occupied Dutch East Indies during World War II." - https://en.wikipedia.org/wiki/Operation_Boomerang
13:46 < xaete[m]> ye makes sense, didnt remember the recent ones tho
13:46 < muurkha> .t https://en.wikipedia.org/wiki/Battle_of_Baiji_%282014%E2%80%932015%29
13:46 < saxo> "The Battle of Baiji was a battle that took place in Baiji, Iraq, lasting from late October 2014 to late October 2015. In mid-November 2014, Iraqi forces retook the city of Baiji, and re-entered the Baiji Oil Refinery." - https://en.wikipedia.org/wiki/Battle_of_Baiji_%282014%E2%80%932015%29
13:47 < muurkha> .t https://en.wikipedia.org/wiki/Lago_Oil_and_Transport_Company#Wartime
13:47 < saxo> "Lago Oil & Transport Co. Ltd. had its beginning in 1924 as a shipping company carrying crude oil from Lake Maracaibo to its transshipment facility on the island of Aruba. /  With the discovery of a vast amount of crude oil under Lake Maracaibo, but unable to send it to [...]" - https://en.wikipedia.org/wiki/Lago_Oil_and_Transport_Company
13:47 < muurkha> "The importance of the Lago refinery was well known to the German High Command and an Attack on Aruba was performed. On February 16, 1942 the site was attacked by the German submarine U-156. The submarine's deck gun exploded due to mistakes by the German deck gunner, and the refinery was not damaged."
13:47 < xaete[m]> lmfao
13:48 < xaete[m]> at the same time, this shows an issue w. "scaling down", which is that running a thousand 10L reactors is much harder than running a 10kL reactor
13:48 < xaete[m]> which is the reason those refineries exist inthe first place
13:49 < muurkha> well, remember that most of these refineries were built basically by hand
13:49 < xaete[m]> mhm
13:50 < muurkha> dudes running arc welders around huge steel pipes and tightening bolts
13:50 < xaete[m]> ye
13:50 < xaete[m]> ig issue w. paralleism stems down to the fact that its easier to make one big pipe than a hundred small ones
13:50 < muurkha> the whole second industrial revolution was largely a question of scaling processes *up* to "mass production"
13:50 < xaete[m]> so economies of scale generally favors the large ones
13:51 < xaete[m]> w. the fine chemical market it might be different tho
13:51 < xaete[m]> and yea that makes sense
13:51 < muurkha> bigger shovels, bigger ore crushers, assembly lines, bigger blast furnaces, etc.
13:51 < xaete[m]> the old joke about "whats the difference between a chemist and a chemical engineer?" "9 orders of magnitude" applies here i think
13:51 < muurkha> and yeah bigger distilleries, in which the humans distilled their gasoline
13:52 < xaete[m]> ye, and you can capture lots more effiency that way, w. stuff like heated/jacketed columns, more layers, and so on and so forth
13:52 < muurkha> well, as I said, I think that's mostly a question of temperature differences?
13:52 < xaete[m]> those small "generic reactors" that work for chem generally all work making fine chemicals, which (generally) need less-harsh conditions and probably dont need to be fractionally distilled
13:53 < xaete[m]> the way i see it its not just temp differences, but also just a boring sense of scale
13:53 < muurkha> larger scales mean lower temperature gradients for the same ΔT and thus, crudely, lower power usage
13:53 < xaete[m]> accurate frac distillation requires lots of "stages", which you generally make by... building a bigger condenser column
13:53 < muurkha> yeah, true
13:53 < xaete[m]> and yea i think ik what u mean
13:53 < xaete[m]> now
13:54 < muurkha> so, even today, most equipment is made by hand-built equipment
13:54 < muurkha> your 100ℓ jacketed reaction vessel may be mass-produced
13:54 < xaete[m]> well... "hand"-built, i think CNC is pretty advanced nowadays, but yes
13:54 < muurkha> but the machine that made it is not
13:54 < xaete[m]> yea for sure
13:55 < muurkha> yeah, but you can't carve the whole glassblowing machine out of a single block of aluminum
13:55 < muurkha> but yeah, CNC is beginning to bring recursive automation into the human manufacturing systems
13:55 < xaete[m]> ye, gotta assemble it from parts n bits
13:56 < xaete[m]> "automating" that process would probably mean changing the final product in fairly significant ways
13:56 < muurkha> as is pick-and-place, and the humans have some pretty impressive welding robots
13:56 < muurkha> anyway, by going to small scales, you give up large ΔT
13:56 < xaete[m]> mhm
13:57 < xaete[m]> ig energy costs arent sth ive considered much, considering in lab-scale chem u can just pretend electricity (and water) is free :P
13:57 < muurkha> and you give up gravity-driven processes, and ultimately even surface tension
13:57 < xaete[m]> mhm
13:57 < muurkha> but what else?
13:58 < xaete[m]> well, at some point you /can/ give up on stirring
13:58 < muurkha> right, diffusion is fast
13:58 < muurkha> also you *must* give up on stirring because high Reynolds numbers are just inaccessible
13:58 < xaete[m]> some things genuinely will get easier, for example scaling up from a 500mL round bottom to a 100L reactor, you have to take into consideration heat flow
13:58 < muurkha> everything is laminar
13:58 < xaete[m]> suddenly heating up your reaction starts looking alot easier again
13:58 < xaete[m]> (if you scale down)
13:58 < muurkha> right, at fine scales you can very reasonably control the temperature to millikelvins
13:59 < xaete[m]> scaling up means that going from say, 0 C to 100 C can take iirc hours, not minutes
13:59 < xaete[m]> going down would reverse that
13:59 < muurkha> right, which makes new reactions accessible actually
13:59 < muurkha> and you can still thermocycle, you just can't do it over space, just time
13:59 < xaete[m]> hm, if you really wanna heat-shock your shit, yea
14:00 < xaete[m]> never really saw a reaction that called for that tho, altho i can guess one (ik sth like kinetic sulfonation relies on short time @ high temps or whatever)
14:00 < muurkha> well, of course neither human chemists nor cells have explored whateve reactions would require that
14:01 < xaete[m]> PCR comes close
14:01 < muurkha> yes
14:01 < xaete[m]> as does the thing where you transform bacteria by blastin em or whatever
14:01 < xaete[m]> but the tendency (in lab-scale chem anyways) is for reactions to tend "boring"
14:02 < xaete[m]> removing things like refluxes or gas-additions or whatever for just
14:02 < muurkha> but if you could do that kind of thing by cycling pressure or pH instead, you could probably do them at submillisecond timescales instead of minutes
14:02 < xaete[m]> "add, mix, cool, wait"
14:02 < xaete[m]> oh and scaling down i think you can (on some level) replace stirring with ultrasound or shaking, and as you go even slower you dont need to stir at all
14:02 < xaete[m]> your bodies cells dont have mini stir bars in em after all
14:02 < muurkha> right
14:03 < muurkha> the humans have done a fair bit of exploration of pH-driven equilibrium shifts already, but not much in the way of pressure except for pure-substance phase changes AFAIK
14:03 < xaete[m]> well, lots of industral scale processes are pressure-driven equilbirum reactions in a sense
14:04 < muurkha> yes!
14:04 < xaete[m]> the haber process is one, uses high P to accomidate the cruddy thermodyanmics of the temp
14:04 < xaete[m]> the wgsr is one too maybe? but unsure
14:04 < muurkha> right, and that's really where chemical engineering came from
14:05 < muurkha> the wgsr works fine at atmospheric pressure, as many humans have discovered by dying
14:05 < muurkha> but at pℓ scales you can easily go to a GPa
14:05 < muurkha> and you can do it in nanoseconds
14:05 < xaete[m]> yep, where you need to make tons, theres no real good way to have a bunch of small reactors, you can only use a big one
14:05 < xaete[m]> at pL scales you can amp up the pressure a fuckton mhm
14:05 < xaete[m]> most obvious reaction that would make sense for is hydrogenation id bet
14:06 < muurkha> yeah!
14:06 < xaete[m]> reduction of ketone to alcohol or whatever (or nitriles or imides or whatever) is gonna b way "easier" or "green" if you can just blast it with high pressure
14:06 < muurkha> ooh, I never thought of that!
14:06 < xaete[m]> rather than having to use say, nabh4 or lialh4
14:07 < xaete[m]> .t https://www.science.org/content/blog-post/hydrogenating-ball-mill
14:07 < saxo> Science | AAAS
14:07 < muurkha> also of course different solvent systems become accessible, supercritical CO₂ being the most popular
14:07 < muurkha> fuck2aaas
14:07 < xaete[m]> (well, that didnt work), but that's an example of sth similar happening
14:07 < xaete[m]> although instead of microscale, it uses the pressure generated between ball bearings
14:08 < xaete[m]> and yea supercritical co2, must be a dream to work with if i could, just open a tap and all the solvent leaves....
14:08 < muurkha> haha
14:10 < xaete[m]> i think sth like easily-scalabe generic reactors work well for fine chems, but i'm not sure for commodity chemicals
14:11 < muurkha> well, I think it'll be a long time before you can compete with Monsanto on superphosphate
14:11 < muurkha> but there are lots of markets where you don't have to
14:11 < xaete[m]> like if you had a "pilot plant" you could arbitrarily scale which was just a few thousand small-scale reactors that somehow could all be automated (with cleaning included... although i'm not sure what to use for that, sth like hot ozone might work) it might help with scale-up
14:12 < xaete[m]> like for pfizers new small molecule going from the kg scale to the Mg scale
14:12 < muurkha> yeah, cleaning is a potential big problem
14:12 < muurkha> though if manufacturing cost is not a problem, you can just manufacture new cells when you want to do a new reaction
14:13 < xaete[m]> ah yes, the good ol "reaction byproducts were cleaned by melting down the flask"
14:13 < xaete[m]> i think somethin like hot nitric could work tho
14:13 < muurkha> reaction byproducts were disposed of with the ICP
14:13 < xaete[m]> ICP?
14:14 < xaete[m]> if not theres always piranha, but idk how well PTFE stands up to that
14:14 < muurkha> inductively coupled plasma
14:14 < xaete[m]> ah lol
14:14 < muurkha> teflon and, astoundingly, delrin are apparently fine in regular piranha
14:14 < muurkha> not sure about lye piranha
14:15 < xaete[m]> you basically just want something to oxidize everything to hell and back, carbon to co2, ntirogen to n2, oxygen/hydrogen to h2o, and chlorine to hcl etc
14:15 < xaete[m]> jesus christ derlin survives that? i dont trust it tbh, ive herad stories of ppl sniffin formaldehyde by accidentally dropping keck clips in sulfuric
14:15 < muurkha> carbon to CF₄, nitrogen to NF₃, ...
14:15 < xaete[m]> (oh and you want hot pihrana too, so somethin like 100 degree angry caros acid)
14:16 < muurkha> well, supposedly people use delrin tweezers in piranha in some labs
14:16 < xaete[m]> clf3 is also another uh, potential cleaning solution, although id prefer to not need it for obvious reasons :P
14:16 < muurkha> yeah, that was what I was suggesting ;)
14:17 < muurkha> I wouldn't be worried about a picoliter of ClF₃ at all though.  I'd even swallow it
14:17 < xaete[m]> lol
14:17 < muurkha> even in a human body
14:18 < xaete[m]> we're probably only going to be dealing with organics for this tho, so you might be even to use something nicer, like just pouring acetone through everything
14:18 < xaete[m]> only issue would be if youre repurposing the machines for something else (depending on scale ofc, im thinkin rn of the 500mL flask scale)
14:18 -!- Malvolio [~Malvolio@user/malvolio] has quit [Quit: beware the living]
14:18 < muurkha> dunno, I suspect inorganics may be pretty important for settling the asteroid belt
14:19 < xaete[m]> and.... turns out PTFE can survive hot perchloric acid (100C)
14:19 < xaete[m]> and iirc all organics get oxidized with that
14:19 < muurkha> ptfe is a motherfucker
14:19 < muurkha> basalt fiber is already stronger than any organic polymer, though of course not stronger than nanotubes or carbon fiber
14:19 < muurkha> and that's using crude pultrusion techniques
14:20 < xaete[m]> for stuff like the asteroid belt, you're going to have to look beyond what we have currently in the lab/industry for running reactions i suspect
14:20 < muurkha> imagine if you could explore the space of fluorinated phyllosilicates
14:20 < xaete[m]> stuff like chemputer prob too limited for space usage
14:20 -!- Malvolio [~Malvolio@user/malvolio] has joined #hplusroadmap
14:21 < xaete[m]> i wonder if in space you can just like. do chem in a bubble suspended in space
14:21 < xaete[m]> probably not, cuz its too chaotic, but it'd be cool to see
14:22 < muurkha> oh, that's another thing: very low pressures are also potentially a lot more accessible at the picoliter scale
14:22 < muurkha> by sorption
14:22 < xaete[m]> hm tru
14:23 < xaete[m]> will not lie tho my "background" as a chemist makes it really hard to imagine pL scales, in chem "microscale" is iirc defined as anything under 1ml or somethin
14:23 < muurkha> which may be even more important for shifting reaction equilibria, since if I understand things correctly the ΔG of anything goes to infinity as the pressure goes to zero?
14:23 < muurkha> I mean, assuming nonzero vapor pressure
14:24 < xaete[m]> but for sth like materials chem i think something like containerization and quick-deployment of larger-scale devices is going to be more important for industial chem
14:24 < muurkha> yeah, as long as manufacturing cost matters at all, I agree
14:24 < xaete[m]> so instead of scaling down the reactors, just put them all in boxes; if you're going to be making a bajillion of something youre inevitably going to need alot of Reactor
14:24 < xaete[m]> uhhhh tbh not sure abt reqction equilibria
14:24 < muurkha> I might be misunderstanding that
14:25 < xaete[m]> thats the other thing abt chem, chemists really /dont/ like dealing with equilbira, it makes the chem messy
14:25 < xaete[m]> well ig reason i'm not sure is that by moving P to 0 you effectively have just removed all of your reactants in a sense
14:25 < muurkha> but breweries don't worry about the cost of manufacturing yeast cells
14:25 < muurkha> yes, but some of them get removed faster than others
14:26 < xaete[m]> at which point you're going back to molecular distillation
14:26 < muurkha> yes!
14:26 < muurkha> so I think it doesn't make sense to calculate an equilibrium *at* vacuum?  but at low pressures maybe
14:26 < xaete[m]> ye... provided youre doing a gas-phase reaction
14:26 < muurkha> it's interesting what you say about not liking dealing with equilibria
14:27 < xaete[m]> the prototypical example i think is pcl3+cl2<->pcl5
14:27 < xaete[m]> a good example of that is ester formation
14:27 < muurkha> I don't know anything about chlorides of phosphorus
14:27 < muurkha> (or much of anything else, thus my nick)
14:27 < xaete[m]> nah  u dont need to, its just a textbook example
14:27 < xaete[m]> fischer esterification for example is an equilibrium
14:27 < xaete[m]> and it works.... on the industral scale
14:28 < xaete[m]> because you can nicely drive the equilibrium to one side or another with distillation or whatever
14:28 < xaete[m]> but on the /lab/ scale well... you can't always do that effectively
14:28 < muurkha> because you don't have time?
14:29 < xaete[m]> not always just that, but also just that somtimes the setups are more complicated for example
14:29 < xaete[m]> or requires distillation, or whatever
14:29 < xaete[m]> so if youre making fine chemicals, its often more productive to get rid of the water with some chemical way, instead of dealign with moving the equilibrium around (which can also hurt your yield)
14:29 < xaete[m]> so instead of refluxing with excess of alcohol in sulfuric acid (which absorbs some water)
14:30 < muurkha> oh, sure
14:30 < xaete[m]> you run a steglich esterification, which irreversably removes the water in the urea byproduct
14:30 < muurkha> right
14:30 < xaete[m]> one way of lookin @ this is to think of this as adding another reaction alongside yours in order to drive the delta G up
14:30 < muurkha> right
14:30 < muurkha> instead of a distillation column
14:31 < xaete[m]> and you can run the steglich at 0 degrees (its preferred actually), with one equivalet of alcohol
14:31 < xaete[m]> the only issue is that now youre removing water via chemical means, so you have to pay for it
14:31 < muurkha> people still use equivalents in 02022?
14:31 < muurkha> right, you have this DCU waste product
14:32 < muurkha> ?
14:32 < xaete[m]> yes, and also you have to literally pay for DCC
14:32 < muurkha> sure
14:33 < muurkha> biochemical pathways are full of this kind of hting
14:33 < muurkha> *thing
14:33 < xaete[m]> eh biochemical pathways use another dirty trick
14:33 < xaete[m]> which is that they generally use really good catalysts
14:33 < muurkha> yeah
14:34 < muurkha> but that doesn't change the equilibrium by itself, does it?
14:35 < xaete[m]> well yea
14:35 < muurkha> I guess in biochemical systems the standard term for "equilibrium" is "death"
14:36 < xaete[m]> not quite, its more that we've evolved to always be in a state of equilibrium
14:36 < xaete[m]> and they have their ways of shifting equilibira around too
14:36 < muurkha> hmm, really?  that's surprising to me
14:36 < xaete[m]> mostly by doing the same thing i described earlier, but the extra bit of delta-g theyre adding comes from ATP (iirc, anyways)
14:36 < xaete[m]> or NADH/NAD+
14:37 < muurkha> where do the concentration gradients in organisms come from if not from non-equilibrium reactions?
14:37 < xaete[m]> ah
14:39 < xaete[m]> one way 2 see it is that its a shifting equilibrium ig
14:39 < xaete[m]> which ig ik what u mean when u say equilibria =death
14:40 < muurkha> thought-provoking
14:41 < xaete[m]> lol
14:41 < xaete[m]> but yea
14:42 < xaete[m]> ig the bigger poitn i hvae abt this is that i wouldn't be too concerned/aiming for/etc uber tiny scale, just because at some poitn you need to make the chemicals for /something/
14:42 < xaete[m]> and well humans are always going to be around 60kg
14:43 < muurkha> only until we upload them
14:43 < muurkha> ;)
14:43 < ^ditto> muurkha: Error: ")" is not a valid command.
14:43 < xaete[m]> well if youre uploading them i suspect you'll have bigger issues, but yes
14:44 < muurkha> undoubtedly
14:44 < xaete[m]> at the same time, w. chemical synthesis (constrained by "on earth, in the near-short-term, for humans/modern technology") i suspect the minimum amount needed to have "useful" data output is going to be around the ug mark
14:44 < xaete[m]> to run things like enzyme screens and whatnot
14:45 < muurkha> but, if in a kilogram of material you can carry out 500 1mℓ concurrent experiments with 100-ms time constants for diffusion due to adequate stirring, or 10¹² concurrent experiments with 1-ns time constants, I think you can probably learn more from the second
14:45 < xaete[m]> (its going to be lower for stuff like DNA, but i'm talking about small-mol, and plus, dna can replicate /itself/. totally unfair)
14:45 < muurkha> totally!
14:46 < xaete[m]> and yea, ur thing abt the kg of material makes sense
14:46 < xaete[m]> at the same time, if we're lookin at stuff like fine chemicals, unless you're making stuff thats ug-active, to actually put it in humans you're gonna need at least a few gs
14:46 < muurkha> similarly I think if you can transform a kilogram of material into a kilogram of one pure compound, or a chosen set of 10⁹ different pure compounds, the second one may be useful in situations where the first one isn't
14:47 < muurkha> oh, I'm not interested in drugs
14:47 < xaete[m]> and even on the ug-scale you're still gonna need mgs, and very few things are ug-active (maybe the most famous is lsd, but if youre not into that, vitamin b7)
14:47 < xaete[m]> hm if not drugs/pharma what do u envision this for?
14:47 < muurkha> a dyson sphere
14:48 < xaete[m]> ohh ok
14:49 < xaete[m]> well in this case i assume you're thinking of nanobots to build the thing then?
14:49 < muurkha> that might be a bad idea
14:49 < xaete[m]> mhm
14:49 < xaete[m]> well for sth that big, honestly the best thing to do would be a large-number of mass-produced mid-scale automated facilities
14:50 < muurkha> yeah, that's more what I'm thinking
14:50 < xaete[m]> basically moving back the "human" element 2-3 steps back b/c the satellites are automated, the factories are automated, and the ships that deploy them are
14:50 < xaete[m]> but the things needed to make those ships arent completely automated
14:50 < xaete[m]> imo you should fit each dyson-sphere sat factory within oh idk
14:50 < xaete[m]> one 40-ft container?
14:51 < muurkha> I don't think 2-3 is enough
14:51 < muurkha> but we won't know until we get there
14:51 < xaete[m]> hm how big is a dyson sphere? you can prob fermi it
14:53 < muurkha> depends on radius and thickness, but on the order of 10²³ kg
14:53 < xaete[m]> 1e23 huh
14:53 < nmz787_> .title https://www.nsf.gov/pubs/2022/nsf22557/nsf22557.htm
14:53 < saxo> Semiconductor Synthetic Biology Circuits and Communications for Information Storage (SemiSynBio-III) (nsf22557) | NSF - National Science Foundation
14:54 < muurkha> but you surpass Terra's energy output nine orders of magnitude earlier
14:54 < nmz787_> I'm signed up for the mentioned webinar on March 17th
14:54 < xaete[m]> uber back of the envolope but if we assume 1 factory can in one year create 1e6kg of material, you're lookin at somethin like 5-6 steps or so
14:54 < xaete[m]> seems abt right
14:55 < muurkha> (current output 12 orders, potential geothermal output 5 orders)
14:57 < muurkha> adverse mutation is a problem with nanobots; larger-scale control systems controlling small-scale manipulators avoids that
14:57 < xaete[m]> eh with a big enough factory just add a CRC and SHA2 and you'll be fine
14:57 < xaete[m]> the satellites arent gonna be that small anyways, at least a few tonnes
14:57 < muurkha> right, exactly
14:57 < xaete[m]> just stick a few computers in it
14:58 < xaete[m]> plus if youre using a master-slave approach (and you probably will want to) its prob not a big deal if one of the tiny ones dies
14:59 < muurkha> human bodies have about 10⁹ cells with DNA, which undergo on the order of 10¹¹ replications before inevitably succumbing to cancer
14:59 < muurkha> they have an additional 10¹⁰ cells without DNA
15:00 < nmz787_> which cells lack DNA?
15:00 < xaete[m]> at the same time, human bodies dont have built-in forward error correction and sha-2 hashes
15:00 < muurkha> erythrocytes
15:00 < xaete[m]> plus, we'ere talking only one dyson swarm here
15:00 < muurkha> we might be able to do better for an artificial system thanks to SHA2, yeah
15:00 < nmz787_> .wik erythrocyte
15:00 < saxo> "Red blood cells (RBCs), also referred to as red cells, red blood corpuscles (in humans or other animals not having nucleus in red blood cells), haematids, erythroid cells  or erythrocytes (from Greek erythros for 'red' and kytos for 'hollow vessel', with -cyte translated as [...]" - https://en.wikipedia.org/wiki/Erythrocyte
15:00 < muurkha> but we might also do worse
15:00 < nmz787_> oh, RBC
15:00 < muurkha> yup
15:01 < nmz787_> quit trying to sound all academic in here, yo
15:01 < muurkha> sorry
15:01 < nmz787_> :P
15:01 < xaete[m]> well its basically a straight function of error rate * code (or meaningful code equivalent) size * probability
15:01 < xaete[m]> and you can make the final one arbitraily low
15:02 < muurkha> dunno, that doesn't take viruses into account
15:02 < muurkha> or polonium tea
15:02 < xaete[m]> lol
15:02 < xaete[m]> well if someones intentionally fuckin with em u got bigger problems
15:02 < muurkha> people will inevitably be intentionally fucking with them from the outset
15:03 < xaete[m]> yea then u gotta talk to the cybersec guys not me :P
15:03 < xaete[m]> im just a (bad) CS person who LARPs as knowng stuff abt biochem
15:03 < muurkha> in general when you push the probability of one kind of fault down low enough, the other kinds start to dominate
15:04 < nmz787_> .wik larp
15:05 < saxo> "A Live Action Role-Playing game (LARP) is a form of role-playing game where the participants physically portray their characters." - https://en.wikipedia.org/wiki/Larp
15:05 < muurkha> LARP is a useful lens for understanding human organizations in general
15:07 < xaete[m]> lol
15:08 < muurkha> Foresight's interest in infosec is not an accident
15:18 < muurkha> thanks for a very stimulating discussion!
15:45 -!- spaceangel [~spaceange@ip-78-102-216-202.net.upcbroadband.cz] has quit [Remote host closed the connection]
15:50 < xaete[m]> ye same here
16:29 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has quit [Ping timeout: 256 seconds]
16:45 < jrayhawk> https://www.lesswrong.com/posts/xwBuoE9p8GE7RAuhd/brain-efficiency-much-more-than-you-wanted-to-know
16:50 < Jay_Dugger> Good evening.
17:04 < lsneff> Indeed it is!
17:07 < kanzure> be greeted
--- Log closed Wed Feb 23 00:00:07 2022