--- Log opened Sun Jan 31 00:00:31 2021
00:02 -!- ottavio [~m0ttv@unaffiliated/m0ttv] has joined ##hplusroadmap
00:59 -!- darsie [~kvirc@84-113-55-200.cable.dynamic.surfer.at] has joined ##hplusroadmap
02:52 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has quit [Ping timeout: 272 seconds]
03:08 -!- preview [~quassel@2407:7000:8423:b79:29e:ecd2:bbd0:8cfd] has quit [Ping timeout: 260 seconds]
03:18 < juri_> lsneff: neat paper.
03:20 -!- preview [~quassel@2407:7000:8423:b79:29e:ecd2:bbd0:8cfd] has joined ##hplusroadmap
05:32 < kanzure> "Universal DNA methylation age across mammalian tissues" https://www.biorxiv.org/content/10.1101/2021.01.18.426733v1
06:36 -!- pompolic [~A@unaffiliated/pompolic] has quit [Ping timeout: 240 seconds]
06:38 -!- pompolic [~A@unaffiliated/pompolic] has joined ##hplusroadmap
08:06 -!- sanehatter [~sanehatte@141.98.255.150] has joined ##hplusroadmap
08:36 -!- prometheus_1 [~root@88.230.146.48] has joined ##hplusroadmap
08:47 -!- prometheus_1 [~root@88.230.146.48] has quit [Quit: leaving]
09:08 -!- Yarrbeard [~Yarrbeard@2603-7081-6701-2976-d474-9e60-5ce3-9342.res6.spectrum.com] has joined ##hplusroadmap
09:27 -!- HumanG33k [~HumanG33k@82-64-99-84.subs.proxad.net] has quit [Remote host closed the connection]
09:27 -!- HumanG33k [~HumanG33k@82-64-99-84.subs.proxad.net] has joined ##hplusroadmap
09:33 < docl> .tw https://twitter.com/lsparrish/status/1355928635178438661
09:33 < saxo> A blockchain where proof of work involves generating a molecular simulation based on the current ledger and making a physics prediction of its properties. Block verification involves synthesizing tiny quantities of the molecule and validating the prediction empirically. (@lsparrish)
09:34 < fenn> a blockchain where proof of work involves getting a job, going to work, earning money, and blowing it all on scamcoins
09:36 < docl> do you really think it's that lame of an idea?
09:41 < docl> everyone seems to think scientifically productive blockchain mining would be cool, but other than primecoin I've never seen it done. molecular dynamics sims are an obvious use case though.
09:42 < fenn> it's now year ten of me being sick of hearing about blockchains
09:42 < docl> lol fair enough
09:45 < fenn> actually i was sick of it before bitcoin, with marc fawzi and his energy based currency crap
09:45 < docl> I'm just thinking of how quickly people switched from amateur GPU mining to group buying a bunch of specialized ASICs and setting up shipping containers full of equipment near power plants. hard not to think "if only that was solving something productive"
09:46 < fenn> it's designed to be a waste
09:47 < docl> right, but you'd apparently get the same functionality if the computational task used were something productive.
09:47 < fenn> read nick szabo's history of money, especially the part about the wampanoag (?) cowry shell money makers
09:48 < fenn> there was this one tiny tribe that specialized in bead making
09:48 < fenn> it's a complete waste of effort, but it's only a tiny fraction of the overall economy
09:49 < fenn> if there were ever something useful, it would be subsidized by almost exactly the monetary value of the usefulness, and then you'd have this waste residual. the amount of activity would balloon to the point where the waste residual is the same as if you were doing a 100% wasteful activity
09:50 < fenn> likewise for computational efficiency improviments, power generation improvements, etc.
09:50 < fenn> currently due to stupid wallet defaults the bitcoin fees are way higher than they need to be
09:52 < fenn> anyway the point is that proof of work really means proof of waste
09:53 < fenn> if you mine ten identical altcoins in parallel, each block costs 1/10 as much, or conversely you can afford to do 10x more mining, so the total waste per coin ends up the same
09:54 < docl> wait, is waste including spillover benefits? because synthesizing millions of novel molecules you aren't allowed to patent would seem like waste from the inside.
09:54 < fenn> if the molecules are useful you can sell them or otherwise benefit economically from them, which means that eventually you can afford to do more of that
09:55 < docl> you don't have a monopoly on the molecules, they are part of the public blockchain. and we can always say using the system involves agreeing to a waiver of monopoly rights.
09:55 -!- catalase [catalase@unaffiliated/catalase] has quit [Quit: ZNC provider BNC4you]
09:56 < fenn> hmm
09:57 < kanzure> docl: the problem with "useful work" was described by andytoshi (here) in 2014
09:57 < kanzure> i think it was in https://download.wpsoftware.net/bitcoin/alts.pdf
09:58 < fenn> i'm not sure about a "returning benefit to the commons scenario" - it's plausible that governments could subsidize the beneficial activity as a whole, but typically they don't do that sort of thing
09:58 < kanzure> or https://download.wpsoftware.net/bitcoin/asic-faq.pdf
09:58 < kanzure> anyway there's various problems with tying it to some "useful" calculation
09:58 < kanzure> 09:45 < fenn> actually i was sick of it before bitcoin, with marc fawzi and his energy based currency crap
09:58 < fenn> if fawzi hadn't been around i probably would have taken it more seriously
09:59 < kanzure> ironically satoshi directly referenced marc fawzi https://diyhpl.us/~bryan/irc/bitcoin-satoshi/email-p2presearch-2009-02-13-023120.txt
09:59 < fenn> yes i know
09:59 < kanzure> yeah i was pretty fed up with him
09:59 < kanzure> i think i hadn't banned him yet?
10:00 < fenn> you did ban him
10:00 < fenn> that's how he ended up on p2presearch
10:00 < kanzure> oh that makes sense
10:00 < fenn> so congratulations, you helped create bitcoin :)
10:00 < docl> lol nice
10:01 < apotheon> exciting
10:01 < apotheon> butterfly wings an' all that
10:01 < fenn> actually i don't know if he was there already or not
10:02 < kanzure> that's all a little weird really
10:03 < docl> basically I'm trying to get two things done: generate a lot of molecules without monopoly value, and do proof of work in a way that is useful in some way. since proof of work needs to be proof of waste, but giving away free molecules looks like waste, the downsides seem to cancel
10:03 < fenn> .title http://nakamotoinstitute.org/shelling-out/
10:03 < saxo> Shelling Out: The Origins of Money | Satoshi Nakamoto Institute
10:06 < kanzure> andytoshi: your presence is necessitated^
10:06  * andytoshi appears
10:07 < andytoshi> yeah, the "useful PoW" thing was in asic-faq.pdf
10:07 < andytoshi> docl: so, if your goal is to generate a lot of molecules, spawning a cryptocurrency that finds them by mining might be useful
10:07 < docl> reading
10:07 < andytoshi> but if your goal is to get a functioning cryptocurrency
10:07 < andytoshi> you should not do this :)
10:08 < docl> why would the cryptocurrency not work?
10:08 < andytoshi> essentially the issue is that the PoW creates incentives for miners to always extend the longest chain, ensuring that you have (eventual) consensus
10:08 < kanzure> page 7 question 7
10:08 < andytoshi> but if people are doing the work for scientific (or any other) purpose, and the crytocurrency aspect is a side effect
10:09 < andytoshi> then they are (much more easily) subject to economic pressure to do bad things to the cryptocurrency. e.g. mine forks and stuff
10:10 < kanzure> maybe docl should look at contingent zero-knowledge payments or zero-knowledge proofs of (molecular research?)
10:11 < docl> not sure I follow, maybe I need to read the whole faq
10:11 < andytoshi> that'd get you the "pay people to do the research" effect .. but probably isn't practical to implement with 2021 tech
10:13 < docl> hmm. the issue is that taking away monopoly value for the molecules may not remove all of the economic incentives?
10:14 < fenn> right. there is still a societal benefit so presumably there would be a societal investment in it
10:15 < fenn> on the other hand, the US only invests a tiny fraction of what it ought to into science and tech development
10:15 < andytoshi> you may not have a monopoly of knowledge of the molecules ... but if you have significant capital investment in synthesis or filtering isomers or whatever
10:16 < andytoshi> then you may have an economic interest in this sort of work
10:16 < andytoshi> similar to why companies contribute to open source software
10:16 < docl> but there's societal benefit to having a valid blockchain as well, so one would expect public funding not to be used to override that
10:16 < andytoshi> yeah, but you need all the social benefit to be about having the blockchain
10:17 < andytoshi> in that case there is parity between the work an honest miner does and the work a dishonest miner does
10:17 < andytoshi> but if you introduce non-blockchain benefit .. this helps the dishonest miner, (potentially) without helping the honest ones who only care about the blockchain
10:18 < andytoshi> there are similar incentive problems with merged mining, where the "external benefit" is just mining a separate chain
10:18 < docl> not sure I buy that... you get the same benefit from solving molecules regardless of the validity of the block involved
10:19 < kanzure> if you solve malicious blocks then you get much more value (economic plus molecular benefit)
10:20 < docl> it seems like the beneficiaries of the molecules don't have incentive to prefer dishonest over honest blocks in any way.
10:20 < andytoshi> they don't, but non-beneficiaries (which there will be many, e.g. small players who don't care to set up the relationships they'd need to profit from the molecules) prefer honest blocks
10:20 < andytoshi> more than beneficiaries do
10:21 < kanzure> docl: by the way, we're not discounting your goal, as we share it, just commenting on blockchain design issues
10:22 < kanzure> for context, andytoshi is much too humble to point out that his code has directly facilitated hundreds of billions of dollars of bitcoin transactions on mainnet
10:24 < andytoshi> well, there's an interesting argument that if you just want to make molecules, and maybe skim off some of the blockchain hype, and don't care about creating a blockchain that has any lasting future
10:24 < andytoshi> you could do this. i would submit that this is unethical, but there are utilitarian arguments for both sides
10:26 < andytoshi> by "you could do this" i mean you could deploy some unmaintained clone-coin of bitcoin where the mining logic was replaced with a molecule search
10:28 < kanzure> oh for further context, dishonest miners can double spend
10:28 < kanzure> which is ungood
10:29 < kanzure> they do this by rewriting recent blockchain history
10:37 -!- catalase [catalase@pls.just.stfu-kthx.bnc4you.xyz] has joined ##hplusroadmap
10:37 -!- catalase [catalase@pls.just.stfu-kthx.bnc4you.xyz] has quit [Changing host]
10:37 -!- catalase [catalase@unaffiliated/catalase] has joined ##hplusroadmap
10:43 -!- yashgaroth [~ffffffff@c-73-147-55-120.hsd1.va.comcast.net] has joined ##hplusroadmap
11:15 < docl> hmm. beneficiaries of the work function profit from honest blocks because it causes high market cap and thus more total effort will go towards mining. colluding with dishonest miners by paying them to throw computing power at dishonest blocks seems like it would be killing the golden goose.
11:16 < andytoshi> i think, in practice arguments of the form "but that would kill the golden goose" don't work well, eventually somebody shows up with a short enough time horizon
11:20 < docl> perhaps, but I'm having trouble imagining what that would look like in the real world when beneficiaries are trying to use molecules for scientific research. they'd have to go to a lot of trouble to incentivize fake blocks at a high rate, and it would have to occur over enough time for substantial fraud to be an issue.
11:24 < andytoshi> no, it'd look like people running mining rigs who have two income streams -- one from the scientists who want the molecules, one from the rewards for mining honest blocks
11:24 < andytoshi> as margins go to zero, both income streams will be necessary for miners to survive
11:25 < andytoshi> but then an attacker who wants the miner to do something bad only needs to pay a bit more than the block-reward income stream, since that is the only one disrupted by bad behavior
11:26 < andytoshi> and it's not logistically that hard for the attacker (or the miner), the attacker just provides a bad block to grind on in place of a good one
11:26 < docl> that assumes the scientists cooperate with the bad miners, which they might not do
11:26 < andytoshi> no, the scientists don't need to be aware of this at all
11:26 < andytoshi> they just see blocks and lift molecules out of them
11:27 < docl> they can't tell if they are bad blocks?
11:28 < andytoshi> no, bad blocks look the same as good ones
11:28 < andytoshi> but even if they could ... it's not clear why they'd care
11:28 < andytoshi> molecules are molecules
11:28 < docl> because the network going down means no more molecules?
11:28 < docl> brb, errand
11:28 < andytoshi> kk ttyiab
11:29 < andytoshi> but the network doesn't need to go down for hashpower attacks, even ongoing hashpower attacks, to be effective
11:41 -!- filipepe_ [uid362247@gateway/web/irccloud.com/x-iclziyvxnoawlbdt] has joined ##hplusroadmap
12:09 -!- TMA [tma@twin.jikos.cz] has quit [Ping timeout: 264 seconds]
12:09 -!- TMA [tma@twin.jikos.cz] has joined ##hplusroadmap
12:32 -!- ottavio [~m0ttv@unaffiliated/m0ttv] has quit [Quit: Leaving]
12:37 < docl> anything that causes the mining reward to be worth relatively less will in general make there be fewer molecules solved. this doesn't need years-long time horizons to be relevant, as a market crash would cause fewer molecules to be solved within a matter of days or less. it just doesn't make sense to hire someone to sabotage your source of new molecules.
12:38 < andytoshi> the scientists and the people doing the sabotage don't need to be the same people
12:41 < docl> how do miners get paid enough for grinding on a bad block to justify the cost, if they don't get the subsidy from the scientists?
12:44 < andytoshi> they get the subsidy from the scientists no matter what
12:46 < docl> why would they?
12:46 < andytoshi> so they can get the molecules
12:47 < docl> they get the molecules no matter what, so why would they pay for them if doing so means fewer molecules in the future?
12:48 < docl> if someone is publishing bad blocks with free molecules, then you can just take the molecules and not pay for them, is how I'm modeling this in my mind
12:48 < andytoshi> sure, you can free-ride on this system
12:49 < andytoshi> but presumably the scientists are paying miners to produce these because it's cheaper than doing it themselves (due to the cryptocurrency subsidy that the miner isgetting)
12:49 < andytoshi> and because they're well-positioned enough to accept loss due to free riders
12:55 < docl> um. it's not free riding to refuse to pay someone to sabotage your business...
12:56 < andytoshi> it's not clear to me at all that the price-per-molecule would be meaningfully affected by miners' behavior toward the blockchain
12:56 < andytoshi> and even if it was, it's not clear to me at all that it's in-principle possible to distinguish malicious behavior from honest behavior at the level of an individual miner
12:56 < kanzure> you can't refuse to pay someone who mines a valid block
12:57 < docl> more people mining = more molecules per unit time
12:57 < andytoshi> and it's not clear at all that the scientists at the molecule company would be doing the detective work here
12:57 < kanzure> the problem is that a double spend appears as a valid block (erasing recent history where something was previously spent)
12:57 < kanzure> "Evidence of two deeply divergent co-existing mitochondrial genomes in the Tuatara reveals an extremely complex genomic organization" https://www.nature.com/articles/s42003-020-01639-0
12:58 -!- apotheon [~apotheon@copyfree/founder/apotheon] has quit [Ping timeout: 240 seconds]
13:02 < docl> this seems easy to solve: make the subsidy payment require several validations before it can be used in an outgoing transaction.
13:03 -!- apotheon [~apotheon@copyfree/founder/apotheon] has joined ##hplusroadmap
13:14 < andytoshi> yes, you cannot spend subsidy payments til they are buried by 100 blocks. that doesn't solve the issue here
13:14 < docl> that would be the solution for the case where you make the subsidy part of the block solution itself (akin to tx fees in bitcoin). however, paying block solvers after their block has been verified 10x and all transactions checked for validity is not that hard either.
13:16 < docl> what am I missing here?
13:17 < andytoshi> i think, that there is no way you can "validate" a subsidy payment, or even an entire block, to verify that it was not part of a hashpower attack
13:18 < andytoshi> if you're suggesting that the molecule company would only pay for deeply-buried blocks, that's an inconvenience to the miner (who now needs to front the money, and is subject to substantial variation in income because blocks are very rare and there's no notion of "partial work" for deeply-buried blocsk)
13:19 < andytoshi> so the scientist would get better value for their money by paying a miner without that inconvenience
13:19 < docl> not if they run a bigger risk of rewarding an attacker
13:20 < andytoshi> they're not the police, they have no incentive to avoid rewarding attackers
13:20 < andytoshi> except for "killing the golden goose" which is several layers abstracted away from them
13:33 < docl> rewarding an attacker means you have less money to work with, for one thing. for another, devaluing the currency by rewarding attacks would tend to cause the number of molecules to go down, defeating the whole point of paying miners to begin with. that doesn't seem particularly buried in abstraction. choosing to pay instead of free riding is itself more abstract than that.
13:34 < docl> but we might be getting a bit into the dismal social science side here
13:49 -!- Dunas [~Dunas@170.81.211.52] has joined ##hplusroadmap
13:49 -!- Dunas [~Dunas@170.81.211.52] has quit [Read error: Connection reset by peer]
13:52 < andytoshi> yeah :)
13:52 < andytoshi> i guess, one final point is that bitcoin miners, who you'd think would have an interest in protecting the bitcoin ecosystem for the reasons you describe, have nonetheless attacked bitcoin many times
13:53 < andytoshi> so if those miners' interest was split bitcoin bitcoin and something else, it stands to reason that the situation would be even worse
14:03 < L29Ah> > have nonetheless attacked bitcoin many times
14:04 < L29Ah> can you please provide some examples?
14:10 -!- filipepe_ [uid362247@gateway/web/irccloud.com/x-iclziyvxnoawlbdt] has quit [Quit: Connection closed for inactivity]
14:29 < andytoshi> bitcoin cash
14:29 < andytoshi> the attempted interruption of segwit deployment
14:29 < L29Ah> i don't see how bitcoin cash is an attack
14:29 < andytoshi> ...
14:30 < andytoshi> in 2014 and 15 it was a very common thing for miners to have "obtain 50% of the hashpower" as a business model
14:30 < L29Ah> it's more like an alternative vision of bitcoin future that actually makes sense
14:30 < andytoshi> non succeeded ofc, but nonetheless that is an attack
14:30 < andytoshi> L29Ah: sure whatever
14:30 < L29Ah> so no attacks, i got it
14:31 < fenn> 50% hashpower is only useful as an attack
14:42 -!- Urchin[emacs] [~user@unaffiliated/urchin] has quit [Ping timeout: 264 seconds]
14:43 < fenn> .tw 1348458430604976128
14:43 < saxo> Could someone please create a network of real doctors willing to work with patients on longevity and nootropic therapies that isn't a nut job and isn't pushing testosterone supplements? How hard can this be? (@mmealling)
14:45 < apotheon> Pushing testosterone supplements is probably quite lucrative.
14:46 < apotheon> The way you get a "real doctor" that fits your requirements is to pay someone enough to make it a lucrative career all by itself.
14:47 < L29Ah> what's a "real doctor"?
14:47 < apotheon> I guess the way to really make that work without being a billionaire is to create a member-owned healthcare organization with its own doctors focused on the healthcare needs of transhumanists and cognitive performance health nuts.
14:48 < L29Ah> i guess i may pass as one in most of Africa, and i'd surely work on longevity and nootropic therapies for monies!
14:48 < apotheon> Call it Health+ (H+ for short).
14:49 < apotheon> I've suddenly got a great idea for a logo, too.
15:00 < fenn> the H+ thing was always lame
15:03 < apotheon> Do you mean using "H+" as an abbreviation?
15:05 < fenn> yes. the organization "H+" was lame too
15:05 < fenn> though i guess it was styled "h+"
15:08 < apotheon> For what I mentioned, though, it might just be good marketing.
16:34 -!- spaceangel [~spaceange@ip-94-112-205-34.net.upcbroadband.cz] has quit [Remote host closed the connection]
17:09 < kanzure> https://www.longevity.technology/whole-body-mitochondrial-transfusion-start-up-lands-funding/
17:11 -!- mauz555 [~mauz555@2a01:e0a:994:7ed0:2193:dba9:7955:2c6b] has joined ##hplusroadmap
17:26 < kanzure> .title https://github.com/bitcoin/bitcoin/pull/17920
17:26 < saxo> guix: Build support for macOS by dongcarl · Pull Request #17920 · bitcoin/bitcoin · GitHub
17:35 -!- darsie [~kvirc@84-113-55-200.cable.dynamic.surfer.at] has quit [Ping timeout: 256 seconds]
18:07 -!- yashgaroth [~ffffffff@c-73-147-55-120.hsd1.va.comcast.net] has quit [Quit: Leaving]
18:12 -!- mauz555 [~mauz555@2a01:e0a:994:7ed0:2193:dba9:7955:2c6b] has quit [Remote host closed the connection]
18:31 -!- Urchin [~urchin@unaffiliated/urchin] has joined ##hplusroadmap
18:53 -!- Yarrbeard [~Yarrbeard@2603-7081-6701-2976-d474-9e60-5ce3-9342.res6.spectrum.com] has quit [Read error: Connection reset by peer]
19:25 -!- Urchin [~urchin@unaffiliated/urchin] has quit [Ping timeout: 272 seconds]
19:27 -!- Urchin [~urchin@31.217.3.88] has joined ##hplusroadmap
19:27 -!- Urchin [~urchin@31.217.3.88] has quit [Changing host]
19:27 -!- Urchin [~urchin@unaffiliated/urchin] has joined ##hplusroadmap
21:03 -!- TMA [tma@twin.jikos.cz] has quit [Ping timeout: 272 seconds]
21:05 -!- TMA [tma@twin.jikos.cz] has joined ##hplusroadmap
22:44 -!- yonkunas [uid403824@gateway/web/irccloud.com/x-syznrhtlnmvlnjlk] has quit [Ping timeout: 260 seconds]
22:47 -!- yonkunas [uid403824@gateway/web/irccloud.com/x-jyhgmybeppgvumpo] has joined ##hplusroadmap
--- Log closed Mon Feb 01 00:00:32 2021