TFTC: A Bitcoin Podcast - Tales from the Crypt #18: The Brothers O'Beirne Pt I

Episode Date: March 28, 2018

Join Marty as he sits down with brothers James and Will O'Beirne to talk about Bitcoin and Ethereum. James is now a full-time Bitcoin Core contributor, working at Chaincode Labs. Will is a part of the MyCrypto.com team and finally gives us a peek into the mind of someone building on and around Ethereum. The conversation meanders through many topics including prediction markets, Proof-of-Work v. Proof-of-Stake, the philosophy behind these technologies, and the current state of the space.

what is up freaks welcome back to tales from the crypt very excited for this week's episode we have the brothers o'burn in the studio unfortunately uh the first five minutes of our conversation got cut off so you're gonna miss the intro so i'm gonna do that now uh we have Will O'Byrne and James O'Byrne. Will is working on MyCrypto.com, working on apps and sort of the tangential apps of Ethereum. So we're going to get an Ethereum perspective on this episode. And then James, his older brother, is working at Chaincode Labs, contributing to Bitcoin Core full-time. So we have a pretty polarizing pair of brothers here. We get into a very interesting conversation. We talk for about three hours. So I had to cut it up into two episodes. So here is the first

hour and a half. Hope you guys enjoy. We've moved on from, maybe not entirely from that conversation, but it certainly carries less weight now compared to all the good. And I think prediction markets are going to go the same way where, you know, assassination markets is the first thing that we talk about now, but later we're going to see all sorts of interesting byproducts of it. yeah i mean so i come from a background with futures markets and like that's that's how you set prices for goods and sort of predict how much corn there's going to be going you use price price pointers and if you could if you can make this more equitable and more more distributed

that's fucking incredible i mean yeah because markets aren't always efficient for what it's worth i just want to put a footnote in there that that um we should stick up for drugs you know um I think facilitating a nonviolent crime is great. Yeah, I'm all for drugs. Sorry, whenever I think drug money, I immediately jump to cartels, which is not. Oh, well, these markets actively help to probably remove some power from cartels, right? Oh, you're preaching to the choir. I just mean when people, you know, state like, oh, drugs are bad for this or that, I feel like more often than not what they're communicating, I mean, obviously there are some people who think for whatever reason drugs are morally reprehensible, but sort of empowering, like, shady figures to further remain under detection.

Right, right. Yeah, the drug war is fucking stupid. I can't believe it's still going on. Like, how long has it been now? like nixon or not nixon reagan started the drug war that sounds right i'm not uh not versed in the history of that and it's i mean let's be frank we're fucking living in a time where like legal drugs are killing more people than illegal drugs i would argue like we have an opioid crisis epidemic some would argue in this country um people get on the heroin uh the gateway drug

heroines like oxycontin opiate painkillers like that and it's gotten fucking out of hand and if we were to hold bar like full just fucking legalized drugs i'm not i'm not even talking decriminalization i'm just saying make them legal one most people won't want to do them like it'll be that that like oh they're they're legal now like they're not cool that social aspect and then for the people that do do them you set up safe environments for them In Philadelphia, I'm actually proud to be from Philly, they set up a safe opiate clinic where people can inject heroin in a safe environment with clean needles. And that goes a long way to sort of making sure people don't die. People are going to do drugs no matter what.

Humans have done drugs forever. Ronald Reagan's not going to stop human nature. Right, right. Yeah, I don't want someone to feel afraid to reach out for help. Because, I mean, that's another thing is, like, if you check into a clinic, you might be worried about losing your job or something like that. You know, I think that having it be legal has all these positive side effects. Again, you know, so similar to what we were just talking about, the first thing you jump to is, oh, but then everybody's going to be doing drugs. But I think once that happens, that conversation will also kind of die down as we see all the positive benefits, something like that.

I heard this quote that's simultaneously funny and sad the other day. And it's something like Aldous Huxley thought that television would be the opiate of the masses. But it turns out that opiates are the opiate of the masses. Too soon. Who would have thought? Too soon. It's, yeah. But one note on the auger end of the discussion, I guess.

a lot of people i think don't realize that bitcoin as a system is capable of facilitating the same kind of you know prediction market pattern that something like augur is doing maybe augur is much farther along than than anything on the bitcoin side but one really interesting piece of work to check out is a paper called discrete log contracts which basically outlines a means for for doing this kind of thing um uh using bitcoin primitives um and that's by uh taj drya who i think is one of the the co-inventors of the lightning network concept yeah that's one of the prediction markets and then we have paul paula stork sports whatever however you pronounce his last name is hive mind um that he wants to do in a drive chain but uh the the security of drive chains have not been

proven and you're basically trusting miners not to steal your money um yeah so let's dive into this so this is the big driver of tribalism and crypto is again going back to the move fast and break things versus the slow and steady uh what i why i hate on ethereum so much i try not to i'm i'm weak i'm a weak bitch i can't oh don't say that marty you got some biceps on you oh thanks yeah i don't know if it's ever been called out but um marty's a beefcake just to the listeners at home it's official certified bitcoin beefcake yeah i've got i've got a beefcake i come from a beefcake lineage broad shoulders never really get fat just get bulky um but going back to crypto cards you know um so the debate i fall back on like the heuristic so i am dumb all you

freaks out there know that i'm fucking dumb he's done and i fall back on heuristics like what you should build the foundation of like we're envisioning a world of new money and decentralized world built on these blockchains and my heuristic is that at the protocol level like people focus on the protocol too much and want to do everything on the protocol level but i think it should be slow dumb and very good at what it purports it will do and then you just anchor trust into it on other layers so that's why i have qualms with ethereum where they sort of hastily threw together their protocol level i would argue and it is is led to a huge attack surface at the protocol level and i mean they're trying to do everything at the protocol level from

what i understand um and james like you were saying like you can do these prediction markets on bitcoin it's just going to take time like to build out like what what are the trade-offs of of moving fast like ethereum like and moving slow like bitcoin so there is a chance that bitcoin moves too slow that ethereum figures it out and and beats bitcoin to market to an extent um so will what like what draws you towards that sort of at the protocol level with ethereum because one thing the uh clementia was that the uh the eclipse attacks on ethereum uh did you read that white paper no i have but the potential eclipse like this is one like one example of where like when vitalik and crew were building out the protocol level is it clementia is that what it

was i'm not sure chlamydia no it sounds like chlamydia we were joking about this last night but they use basically um sort of a well-known service to help nodes connect with each other quickly and it it the service made it so that like eclipse attacks uh were very like highly probable on ethereum's network and that's just like one thing they did that from like the get-go and that wasn't something that was taken into consideration and i think that like hastily thrown together at the protocol level just like doesn't sit well with me now can you clarify on that is that is that the the bootstrap node basically the first thing you connect to yes and is is that something that is um kind of software independent like you have many different

um you know node clients like geth and uh parity are kind of the two big ones because you know i often hear things about oh solidity is garbage or you know something something about um something else about the ethereum ecosystem is tough to work with and i don't hear a lot of complaints about um right you know i don't hear a lot of criticisms of actually the the core stuff those those are layers on top of ethereum i mean you have the evm which many languages could be built against. You have things like Viper and other languages that are trying to be a little bit more, like, less error-prone.

And you have, you know, different node clients. Like, those, to me, those kinds of problems aren't, like, multi-year problems. That's some of the software that was written today that could be changed tomorrow, you know, has this one potential issue. Yeah. And to be clear with everybody, this attack's been patched. I'm pretty sure. It's not possible anymore. But again, this is going. So here's from Bitcoin Magazine, article titled, Researchers Explore Eclipse Attacks on the Ethereum Blockchain.

If you guys want to look it up, here's just a snippet. But as it turns out, Ethereum was actually easier to attack mainly because while Bitcoin relies on an unstructured network where nodes form random connections with each other, Ethereum relies on a structured network based off the protocol called Kademlia, which is designed to allow nodes to connect to other nodes more efficiently. So nodes in Ethereum's peer-to-peer network are identified by their public key. Remarkably, Ethereum versions prior to Geth v1.81 allowed a user to run an unlimited number of nodes, each with a different public key from the same machine within the same IP address. yeah so break that down for us dumb people please um let's see i can't comment too much on the way that ethereum does things but what what i can tell you is that um this stuff is really unintuitive so actually bootstrapping the nodes that you have an outbound connection to

which is um basically the nodes that you you trust more than inbound connection right because if someone's trying to attack you it's really easy to open a connection to your node but if you actually have a choice of who you initially connect to then you can do a good job of choosing peers so ethan heilman who did a lot of early work on the eclipse stuff he may have even coined that that name i'm not sure but um yeah who wrote this paper yeah yeah he came and gave a really great he spent some time at the residency and gave a few really great talks on the thinking that went into countermeasures within bitcoin to avoid um these kind of you know eclipse attacks or i guess more generally sybil attacks um and so one naive way to do it which which may or may not be may or

not kind of resemble how for example geth was doing it is that you could select peers that you know you could prefer to peers with that that have a low latency relative to you are and that you know that might be efficient or you can you could you could peer with somebody who's you know delivering blocks very quickly um but obviously that's kind of gameable because if you know someone's in a certain data center you can spin up a node in that data center to um connect to them um in overwhelming number i guess that's disproportionate to to your hardware um and and then you know feed them bad information um so so ethan actually implemented or suggested a number of really interesting criteria um to help deal with this situation like for example um he he

buckets peers based on their subnet. So he ensures a distribution across the IP address space to some extent. The thinking being that it's pretty hard to obtain a diverse set of IP addresses beyond a certain point. And so that's just one way of mitigating that. But there are all kinds of heuristics that go into the selection of which peers that you want to pair up with. Yeah, I mean, it's crazy how granular you have to get and think about these attack vectors. An Eclipse attack, the thought of somebody intentionally buying space at a data center that you are using just to Eclipse attack you, that's deep down the rabbit hole thinking there. Yeah, and I think that's why I'm personally so conservative and slow moving with all this stuff is because it's sometimes a priori hard to tell which details are going to become relevant, which details of design are going to pose a risk to the system. And so I think it behooves us for this technology that we know almost nothing about to be extremely risk averse.

And I think sometimes people in the Bitcoin community beef a little bit on some of the popular choices in the Ethereum community because there's sort of a cavalier attitude about, oh, you know, we'll just kind of seek social consensus for this or, you know, we'll trust some number of our peers or, you know, whatever it is. And I think any any degree of hand waving can can get kind of risky. And I mean, not, you know, nobody can avoid like there's no there's no silver bullet. Nobody can avoid some amount of hand waving because we don't know how this stuff is going to play out over decades. But I think you need to be as as careful as you can be. i agree but i can also see the ethereum side where it's like hey we don't fucking know let's try it as much as we can i mean i i think that might that might be a little uh more off the cuff than they are to say just fucking try it no you know i think i think they're you know there

is a lot of consideration um i mean going and reading any any of their like they write very nice faqs on uh i can't remember which github repository it is but there's one one uh directory called faqs and i really like that is that the one from carl not sure i don't think so yeah it might be um but anyway you know explaining things like um obviously before i came on the show i had to buff up i knew i was gonna get slaughtered by some beefy bitcoin boys um so you know the resources are out there and and and i i was reading up and and i mean there there is a lot of consideration for these kinds of um you know game theory-esque issues um and another thing you know just to to not that eclipse is the only ever vulnerability that um has ever come up but

but that is a vulnerability that exists in one form of node software that the worst case scenario is that you run a node that you aren't certain is true. And you could say that's pretty bad. Well, it's certainly very bad if Coinbase is running a node that has fraudulent data in there. But I think there is something to be said for the most common use case for a node is for me to run a node so that I'm going through it. I think that if, you know, a company, a multimillion-dollar company wants to run a node, they can do the due diligence to adjust the way in which a node bootstraps or, you know, only, like, manually trust certain nodes that they have verified through some other auditing process.

I think that the relative costs of losing some amount of security for, like, kind of a low-risk common case is sometimes worth it. I mean, Ethereum right now is running more nodes than Bitcoin, and I think a part of that is because it's incredibly fast to bootstrap with FastSync and not very data intensive. And I know Geth 1.81 really just crushed the numbers on that. And I think the attack factors are larger for that, but I think me just running a node to do a little bit of development, just play around with Ethereum, I'm not quite as concerned about that use case as somebody else. And if somebody else cares a lot, they can do a full sync instead of a fast sync, or they can choose which nodes they bootstrap from and audit that themselves. So let's take a step back here and de-alienate between a full sync and a fast sync for our listeners. Okay, so a fast sync, I'm sure you all can imagine that after running these nodes and having this ledger go on for many years now, it's gotten quite large.

That was the main divide between Bitcoin and Bitcoin Cash, was they wanted to make each block larger, and the Bitcoin core team said, no way, it's not going to scale. It's already getting pretty big and that's only going to make it worse. So it is a problem, the idea of having to sync all of this data. And so to mitigate that, one of the options, and I believe it's the default option now, but for synchronizing your node with the entire ledger is to verify a subset of the data

on a lot of the older past transactions and then only do full verification of the last 1024 blocks. So you open up yourself to potentially being fed some bad data about very old blocks, But that would require whoever to also have been able to generate 1,000 blocks that are also fully valid, which would require an absurd amount of computation power. So the agreement there or kind of the tradeoff there is someone doesn't have to forge the entire blockchain. They only have to forge the last 1,000 ones. But then my sink is sped up by like a factor of 100 or 10 or however much it is.

I think you might be able to manipulate data before the thousand blocks, right? I think basically you're doing a headers-only sync until you get to some window. And so presumably if you can come up with a Merkle root that collides but contains the wrong data or some otherwise kind of bungled chain of headers that still passes Geth's sync, even though it's doing a full sync of the last n blocks, You still might have some bad data embedded in there somewhere. Yeah, that could be possible from what I understand. Now I'm starting to step outside of my area of expertise a little bit.

Again, I work on the ecosystem, not the core protocol. But my understanding is that it couldn't be just one node who feeds you that bad past data. I believe there is corroboration in the network, so it would require a lot of nodes to feed you that particular bad data. Or just all the nodes that you're connected to, and then you've got the- Right. Yeah, so that's where something- Eclipse attack. Like where Eclipse plus FastSync could lead to a difficult thing.

But again, Coinbase, they're not going to do FastSync. They're going to do a full sync because they know the importance of this. Well, let's talk about Coinbase. They fucked up something this week where- It was Coinbase fault, wasn't it? I actually wrote about it on Monday or Tuesday. I forget when it was. But the way it was written in the article,

the couple articles that spoke about it, they said Coinbase fucked up in configuring their smart contract and was basically allowing their customers to steal Ethereum from them. Yeah, so what would happen is you would execute a transaction against their smart contract. And I don't know the particulars, but I know that there was some way to cause that transaction to become invalid after Coinbase had accepted it as valid.

So even though your transaction did not go through on the blockchain, Coinbase's database would update with, oh, now you have however much Ethereum you just deposited. So, yeah, to your point, that was an issue with the contract that they wrote. another example of coinbase incompetence but we're not gonna dwell on that assume good faith i i yeah hopefully they're trying you know i believe they are i believe they are i just think they're a little misguided um but they're getting on the bus they implemented segwit so yeah i mean after after trying to change the protocol getting screamed at after

their poor bcash launch but let's not dwell on that let's not dwell on coinbase brian armstrong he's getting he's getting he's getting like cash traded in public right now yeah and you know as you said as you said in an early episode um we owe coinbase a lot we do it's you know they would not be where we are without them oh yeah they were my baby's first exchange yeah me too absolutely so i mean yeah never forget your first you know companies companies go through arcs and you've always got to leave the door open. But this particular bug, it really does

make me wonder, so the bounty that got paid out, $10,000, not bad, 10,000 Clambos. It should have been more. How much people could have taken them to the bank? It's a hundred grand bug at least in my mind.

So the thing I'm wondering is if anybody had found this bug and said, eh, 10,000 bucks or I'm just going to do this. Do we know that anybody hadn't exploited it? I don't know. Would their database be able to catch it? So I don't think, I think maybe if they

went back and corroborated their database against the blockchain, because the contract is there, you can look at every transaction that went in on it, but I wonder if anyone's out there in the wind with... Would not be surprised. Yeah, would not be surprised. And the other thing is, like, what other inconsistencies might there be

between the blockchain and any proprietary database, right? I mean, that's why... guys this is why we should stress hold your own private keys please like do not trust exchanges like the only time i've ever gotten burnt in this space is by exchanges like it's happened like in 2013 to the first excuse me 2014 2015 happened to me on two exchanges like i might like your shit jacked like it and like the one min pal was a fucking exit scam like you live you live and you learn yeah scar tissue man yeah very yeah very aggressive scar

tissue like you you literally go to log in it's like hey this website doesn't work anymore and your money's gone i think my favorite was the ico that uh after i think they only made like a thousand two thousand bucks but the guy just replaced the html file with the word penis that was his exit strategy hopping on a plane to the Cayman Islands and just commits this one last change penis.html I love it

should we embark on a penis game that's how you gotta end your episodes whenever you're done with this thing penis I am not affiliated with these two to your point about owning keys though i i feel like this this is something you work on yeah so so this is kind of where i come in so yeah i mean i i am also i'm super into controlling your own keys i know that scares a lot of people um so i just gotta throw out there if if you own

i'm gonna say more than if you have more than like 1500 into crypto you owe to yourself buy a hardware wallet? Probably even less. You know, I think if you, if you look at the management fee for any like investment, just apply that to your own investment. And whenever it exceeds the amount of a hardware wallet, that's, that's your management field fee. These things are great. I think, um, you know, I've heard some people say they're not sure about them or, or, you know, whether or not, um, but I, I have a lot of faith in them. If you're worried about holding your own keys no reason you shouldn't go out and get one i think will might be alluding to a previous mystery guest i might be i might be i might be throwing some shit myself who pierre he was a mystery

that's true he wasn't a mystery pierre does not uh he was yeah he's is he still bearish on harbor wallets he he was at the bit that meet up last night sorry for doxing you pierre um he said he bought he had he's gonna start experimenting with what i'm not gonna say which which build but it was ironic because last night we're going over uh the ledger vulnerability that came out last week that kid let's talk about this like that kid rashid what's his uh i don't know his last name but he's 15 years old and he's like literally finding all the bugs in the hardware wallets he found one in treasure a couple weeks ago he found one in ledger last week and the one in ledger was pretty pretty egregious like you could remotely like

attack a ledger like from what he found well that's what does i mean you know this this may be a nonsensical argument but it is one of these gut reactions i have like to you know to pierre's point a little bit i guess it it does worry me a bit that there's this specialized device out there that's marketed with the intent of storing a lot of value and so it's like in some ways that's that's kind of a honeypot right that's like it's a it's one thing that someone has to go out and think really hard about in order to attack a lot of value though i mean any attack would require some amount of interaction with the device itself yeah that's i mean but that's the part that sketches me out like you don't know what happens before that thing got to your doorstep like

yeah yep yeah i i totally agree well at least in the case of any wallet i've i've gotten there's usually a you know tamper proof sticker whatever i know on on on the most recent trezor model model t um great wallet by the way uh not getting paid for that um they they put a sticker on the device itself and that thing will not come off i've been trying to scrub it off it's just like it looks awful so i hate that but i love knowing that if someone were to try to take that off i mean it's like embedded in the plastic or something. I don't know. It's pretty hardcore, so I feel pretty confident

at least that nobody got their hands on that or they got their hands on those stickers. And this is another piece of advice for you freaks. Please do not buy your hardware while it's on eBay. Please. Oh, yeah, definitely not. Don't buy them on eBay. I don't even recommend Amazon.

Would you recommend Amazon? No, I've bought Trezors from Amazon. They do a pretty good job of sealing the Trezors. i'm is amazon just a distributor for like trezor or like do they buy a bunch of treasures and then distribute them from amazon factories i think these were cashed in amazon at some point so i yeah i'm i'm a i'm a mediocre cypherpunk at best no it's like that's what we're talking about with all business pete before before we started recording it's like is how can we how can we trust these things like he he was

more worried about like the fact that you can it's turtles all the way down you can just copy the code and create like all these coins but like there's so many attack vectors at this point in time where it's like we are taking a big risk like this is we are on the edge of computer science i would say i would say psychology too because this is completely going to change like the psyche of of humans i would say like if if sound digital money becomes a thing not only that but like a deflationary economy i mean i guess maybe if you go go way way way back to um you know when people you know nick zabo will tell you about when people were using shells long shells as currency and i suppose that's deflation well not even really in the same sense that bitcoin is because we know we

know that Bitcoin has a limited supply. It's a very clear feature of the system. I'm kind of with you. I don't think humanity's ever gotten to play with the deflationary store of value before. Yeah, we were talking about that at brunch a couple weeks ago for John's birthday. Sorry for doxing you, John. It's a lovely birthday. Lovely birthday, and John's a lovely person. we got into a really good conversation like are we ready for like a change this big like an

economic paradigm like it's fucking heavy shit to think about like if if this stuff comes to fruition like we're gonna have to rethink the way like we consume yeah i i agree it's a little scary i think um back you know at the turn of the year when the price was going nuts and you know it was like oh wow okay bitcoin can achieve this sort of public acknowledgement um i was kind of thinking through scenarios where it it it is say adopted internationally as some kind of de facto um store value and say oil contracts are denominated in bitcoin and not the u.s dollar and the u.s dollar crashes and um i mean that's that's like um that's a potentially it's it's it's an interesting event with very scary implications in the short term well thank you for the cue um i mean i i uh

i think that we are not ready for it i think that maybe our kids will be ready for it or their kids what about us makes us not ready for it i think um it's kind of like object permanence like it's something you either have or you don't have kind of as a as a being um and i think it's probably a very hard thing to learn object permanence i've never heard that phrase oh that's the idea that um you know how kids like when you play peekaboo and you put your hands head behind your hands like they think you just vanish from the face of the earth they're like holy shit that's the most amazing thing i've just ever seen bam you're back um you know and i kind of think that that um it's that fundamental and it's that much of a thing that you need to learn

um when you're young it's something you need to be comfortable with the idea that something isn't there but it's there yeah no i agree like we were thrown into this world of conspicuous consumption and we're addicted to it to a certain extent like but i mean will to to rebut that a little bit How is looking at the balance on your wallet any different than looking at the balance on your checking account? I would say it's because there's a lot of infrastructure around it, and certainly marketing. I mean, there's the idea that if I walk into any bank, for whatever reason, I feel secure that I can get that money. I know that that's not true. I've seen bank runs.

I was in Columbia last year, and the line went out the block. Really? In the new year, yeah. What happened? I don't know the particulars, but what I know is that, you know, I was sitting out in a cafe and I watched the same person get in the line and like two hours later, they finally were able to make it out with their money. You think it was Venezuelans going over the border? Honestly, no clue. I'm painfully ignorant. But what I do know is that for whatever reason, you know, my friends, my family, they trust that that exists.

They don't trust cryptocurrency, which is ironic because I know I own my cryptocurrency. i know that i can take it that's the one thing that i do actually know is true so i think it will take time for that trust to shift um because right now people always ask me like yeah but what's it worth what's it worth like what you know can you hold a bitcoin yeah what's the intrinsic value is all business pete was just growing yeah it's the censorship resistance i mean that's like that it's a power that we've never had before i totally agree i think when it comes down to it that's the real distinguishing characteristic of of bitcoining and cryptocurrencies generally i mean the programmable aspect of this programmable money idea is a little

bit overstated i mean we we have programmable money right now it's just that a lot of it resides in a private database. But what we didn't have before is this way of transacting value that can't be stopped, where any one actor can't stop you from sending a payment to somebody. That's the thing that we have to guard and preserve

above all else, and that has to kind of guide our decision-making around this stuff. And so what I think about is how can we keep the system as small? How can we keep the chain as small as possible while still facilitating that aspect of it, but without exposing us to other risks? Completely agree. I mean, it would be a non-starter for me if it wasn't censorship resistance.

That's the beauty of it. The fact that, again, we're going to get back to the sketchy area of embedding data. We always knew we were coming back. Are we going to embed beefy Bitcoin boys into the data? If we want to, we can. They're going to make them illegal. You better go now. No, but I was talking about the one thing that fascinated me last year

was somebody embedded information about the Tiananmen Square massacre into the blockchain. The fact that if people were running nodes in China, or not even running nodes, they had access to a block explorer and could decode, decrypt that encoded text about Tiananmen Square, they could get access to something that is very, very, very illegal in China. And something they should probably know about.

So, I'm sorry, just to get some clarification here, because I, too, Marty, am an idiot. And maybe James could answer this, but when people talk about encoding text about Tiananmen Square, sounds, you know, not that large, but when people are talking about, you know, trying to embed imagery and stuff like that into the blockchain, they're not just talking about committing a hash to it, right? They're talking about the actual content. What is the vehicle? Yeah, I was reading a white paper about it, and it has to do with pay-to-fake-key hash.

Is that correct? I don't know about fake-key hash, but basically there is an instruction called push data, where you push an arbitrary piece of data onto the script execution stack. And so people, I guess, use or misuse that in various ways to just get some arbitrary data into the blockchain. So I think you need a parser that knows a certain way to unpack whatever data that you've packed into the blockchain. There's no, like, you know, your Bitcoin core wallet isn't going to render the JPEG that you've just thrown in the chain. But some program that's maybe doing both the serializing and the de-serializing will know how to do that. Not a recommended use of Bitcoin blockchain, by the way.

Very expensive. But that actually does, I mean, that does make me wonder, is there a use case for getting around this by not using typical standards like JPEG or PNG or whatever? I mean, could you find whole ways of communicating these things in proprietary formats that are for, you know, some sort of like Chinese underground anti-censorship network? and likewise could i write a client that could kind of um try to read the tea leaves of random data out there and generate bad stuff and say oh someone is trying to encode something in in the blockchain but actually it's just my particular client um is is you know trying to make a mess of the noise yeah i just think with this stuff we already have existing means that work well enough for transmitting um verboten data you know we have tor um we have torrents we have vpns and so the people of china can get it a lot of information that's the the blockchain doesn't

really uniquely enable that um didn't know that oh yeah yeah i didn't know like i thought i thought like vpns and torrents in china would be like would not make it past the firewall no no it's if if you go over to china somebody can get you set up with a vpn pretty quickly really yeah again i'm stupid he's not stupid um but in you know and and again like hosting and retrieving um any kind of data uh like a picture or or a piece of text is fairly like it's it's it's not of the same criticality as like value transmission is yeah and that's like so let's let's get on this like so where i feel like people are so misguided in this space is that like they're like

wow this new technology is going to enable us like rework like everything that we know now we're going to be able to make it turn into a blockchain where like like i said on the last episode of this podcast when you compare bitcoin to things stop comparing it to technology you have to compare it to like macro things like salt like like salt was the base currency of spices for for a while and still is arguably like that is such a a fulcrum of like humanity that that that we sort of revolve around that when we're comparing these blockchains like facebook and myspace i'm just like you're completely missing the fucking point uh in my mind like this is something like paradigm so paradigm shifting it's not social it's not like a social network it's orders of

magnitude more it has orders of magnitude more gravitas than the sort of social media internet technology boom that we've experienced yeah yeah another um kind of false analogy that i hear a lot is uh vhs versus betamax right and betamax was the superior technology but it lost out um to vhs in favor of or because of you know distribution mechanisms or deals or whatever and i think that's a really faulty analogy because again the marginal difference between those two is is pretty slim um and cryptocurrency systems are something that that are are totally predicated on the technical details and the design of these systems so i i think it's it's a really hard it's hard to draw analogy for this stuff because the

the parameters of what makes a system resilient or valuable um are different than i think anything we've seen before um i mean i i definitely agree that the the blockchain everything movement is probably one of the most undermining things right now. Let me just throw out a quick Tales from the Crypt investment advice. We're going to start a new segment here. We're not financial advisors, but we will give some investment advice. Okay. If you see a hot ICO out there and you're skimming that white paper,

as we all pretend to do but none of us do, there's only one section you need to read, which is why is this a token or why is this on the blockchain and if you see a lot of words that basically say so we can get the money or because we think it needs to be decentralized or whatever if they can't explain to you why it should be tokenized um it's a terrible idea the question i always like to ask people when they say oh is is this a good use for the blockchain is could you do that with venmo would it be it would everything be exactly the same if i could just transfer the value using venmo and if you're okay using venmo you you should not be on the blockchain because

you're already okay mitigating trust to some other party the only things that should be on the blockchain are things in which you are completely unwilling to trust any entity and that is a very small subset of things yeah the the other shoe i'm waiting for to drop is there's kind of a lot of companies right now looking at trying to put physical assets on the blockchain, so property, cars, you know, whatever. I'm waiting for the first judge to rule once somebody loses their wallet

and loses their Ferrari coin for the judge to say, no, that guy doesn't own your Ferrari like you do, you know? Because you're trusting property law at that point. Yeah, there's a lot of confusion around the intersection of meat space and digital space

like there's like this is the whole oracle problem like what the fuck like there is what you have built

into these protocols the rules that you have built in these protocols it goes back to the Mike Tyson philosophy everybody has a plan to get punched in the face fucking

like you show up in the real world it's like come on you're really gonna seize my house because I lost my private key like something like that like

right it's not gonna happen I mean at least it's not gonna happen today yeah And maybe way in the future, I mean, I don't know, I'm sure back in settler times, if you had the deed to this plot of land, maybe the government would, like, kick the other guy out, even if you stole it. But that's just not really the world we live in today, right now. And it's, I don't know, it's really going to take the wind out of somebody's sails when, you know, some judge rules against their entire platform. Yeah. And so let's go back to heuristics, and let's go back to raising money.

like you were saying, how much money ICOs have raised. It's been egregious, like Tezos, what, $230 million? A couple other projects. Tron's still hanging in at number five? Yeah. I don't know. EOS is definitely a money laundering scheme. They have an untapped ICO that's still going on,

and they're still making millions of dollars a day, which leads me to believe that somebody's laundering money through that ICO. But I digress. So let's talk about heuristics here. So Tezos, Banker, EOS raised hundreds of millions of dollars, zero product. Lightning Network, Lightning Labs, which Elizabeth Stark is the CEO of. Shout out, Elizabeth. Shout out, Lalu.

Shout out, Yoa. Shout out, Roast Beef. They just went to get a funding round, and they only raised $2.5 million, and they did it the traditional way, and they did it with a product. like so to me they like they have something built it's on main net it's working it's not vaporware it was like the whole bcash uh fud around like networks that's going to be vaporware now it's on main net it is something that is going to bring immense value to the bitcoin protocol network overall not the protocol level it's going to bring to the bitcoin network and ecosystem

um and they only raised 2.5 million dollars so like you look at that you say this is something with that already has utility that has been proven they didn't go the ico route and it's like why why are people dependent like do you think as somebody's working on ethereum with ethereum that like a lot of the ico hype is is doing you guys harm in any way you know actually i think it's by and large dying out um i think of the icos that i have somewhat paid attention to or know people who are working on them um all of them are they're extending their windows they're taking longer to reach their cap if they even reach their cap most of them aren't um and they're they're delaying it or or you know otherwise um i i think the ladder's been pulled up on that i

think everybody got caught up in the craze for a little while and it was a craze but it was a craze yeah i mean you know i think the network has chilled out it's you no longer get dosed where you just can't send a transaction for four hours unless you're willing to spend like three bucks yeah um you know that that time seems to have gone by um i you know i think a lot of people are still holding the bag on a couple of those uh comes in waves though it comes in waves there's gonna be another wave at some point it's like that's another thing like how long is this uh this consolidation slash correction gonna last do you think is it we're gonna completely shift phases here though like so comparing this bear market which is three months at most right now

um to like 2014 2015 do you think obviously the ecosystem is more mature do you think we have a similar like 18 month drawdown or do you think the hype got too high last year or do you think like the i gotta stop saying like man it's the curse of the 20 something right you're stuck with it it's your accent i think it's relatable marty it is it is um but is this time is it different this time this is a terrible question to ask because it will probably be proven wrong in a long time but is this time different you know i'll take a whack at it um people people tend to shy away from this kind of stuff i um i guess the caveat is i have no clue um i i think though that this stuff is kind of in the public eye more than it was like there's this been this kind of

critical mass of of oh okay you know you ask your man on the street what bitcoin is and instead of him saying get away from me what the hell are you talking about he's like oh yeah but have you heard about nano blocks you know um so uh but but at the same time i think the markets are extraordinarily irrational right now i mean when you have tron at a multi-million dollar market cap that means that prices might mean nothing so i don't really know how to how to interpret that stuff um i'm i'm fine waiting out for a while uh i'm you know i i would love to kind of get i've been trying to for the past week to get out of the news cycles and just kind of put my head down and think about development um think about how we can make this the system last um and

preserve those characteristics of censorship resistance even through you know bear and bull markets um but i i don't know i i think that this this concept um has been kind of sewed widely at this point across a lot of smart people and people are going to keep thinking about it and they're just going to keep a pretty steady pace of infrastructure development around it. So I'd be surprised if it was, like, a low period for a really long time. But, like, you know, as people are quick to point out, these low periods are nice

because they kind of prune out the fair weather fans. Yeah, exactly. These are, like, my favorite. I mean, my favorite. I mean, fucking millennial repeating, like, the same phrase. i mean i mean um um no i love this period as well because people focus on the nitty-gritty details there's not too much like it's impossible not to get caught up in the price hype when it's running that high when you go from a thousand dollars to twenty thousand you know yeah my my attention span

went to nil you know it was it was horrible it was like no human is meant to engage in this kind of like speculation yeah it's just mental whiplash constantly yeah but um so just let's segue into this there's things getting built out right now there's a lot of very intense conversations going on and like what is the next battle so last night we basically came to the conclusion for bitcoin specifically the next battle is going to be fungibility in my mind like that's going to be the next big like what was the segwit battle in the future is going to be the fungibility battle and some people argue fungibility doesn't matter because of plausible deniability uh at the protocol level some people say it doesn't matter at the protocol level because you can get it on

second layers like lightning and get fungibility there um i would argue though that ideally you would want fungibility at the protocol level um so for those of you who don't know fungibility talked about this before matt corallo fungibility with the u.s dollar sense is you go to a bodega you give a dollar and they give that dollar as change to somebody else they have no idea that that dollar came from you right now in the bitcoin network specifically uh you can sort of track transactions on the blockchain and sort of know which utxo came from where um so when you're spending bitcoin you're really spending on unspent transaction outputs which is utxo uh james in particular what are your thoughts yeah i think the operative word and what you just said

is ideally, you know, fungibility is ideally a characteristic that you want at the core protocol level. And, you know, let's take Monero, for example. Monero has implemented an early iteration of confidential transactions. And they have a problem where the privacy of the chain is contingent on maintaining this notion of an anonymity set. And the upshot of their implementation is that the UTXO set is just unbounded in growth and you can't really prune it because if you prune it, then you reduce your anonymity set and you compromise the privacy of your users. So that's a case, I think, where you have to weigh fungibility against scalability and you have to make sure that, again, you're not compromising kind of the core function of

the system um which is this censorship resistant value transfer but then i guess that's sort of a circular feedback loop where it's like well to be censorship resistant you have to be you have to be sufficiently fungible um but this is this is again another argument for keeping the chain really small so um uh you know i'd encourage all the freaks out there to go and listen to andrew polstra's talks on what he calls scriptless scripts and um the main idea of scriptless scripts is that you can have this smart contract-like behavior live as a property of the public key and signatures of your transactions due to something called linearity that signature aggregation buys you.

So I think there are some really exciting avenues out there for maintaining fungibility in that sense, and that basically amounts to moving data off of the main chain. So, you know, a criticism that you might be able to throw at Ethereum is that if you have your smart contracts living on-chain, it's very evident what an address does or, you know, if you have value at some address where that, you know, like what function that has been acted upon by. Whereas if you have a scriptless script-esque system or a method like graft root or tap root, one address looks like any other address regardless of the mechanisms behind it or what quote-unquote contract acted upon it. So, I do think that there are some really promising avenues out there, but I think we have to be really conscious about the scaling implications of those means. Like, for example, you know, Benedict Bunce and Polstra and a few other guys have come up with these bulletproofs, which are a significant reduction in validation time for confidential transactions. But it's still, I think it's still widely considered to be too slow for Bitcoin.

um so you know it's an outstanding question as to whether we're going to get to a point where the bitcoin community considers confidential transactions to be safe to implement um so yeah we'll see if we get there um i do think the second layer is is pretty promising um but admittedly i haven't thought a lot about this stuff so i haven't really so so with the with the second layer scripts um how how is the the trust mitigation handled where there is a layer outside of the protocol that i now have to trust that it's going to do the right thing to some capacity there are ways of mitigating that but what's kind of the angle with with like um i mean i've listened to the the scriptless script talk and i understand how there is some sort of exchange

of of hashes that that sort of proves intent um namely as it relates to like atomic swaps i think was the example given there but but it didn't really speak to like one of one of the things that that you pointed out about ethereum is that an address is an address you know what's on it you know if it's a contract you can grab that bytecode if you want um but part of that allows people to not have to trust what they're sending to is going to work they can actually you know validate that and verify that um obviously you're going to get you're still going to get the people who are responding to send me 0.5 ether and get five ether back i'm vitalik here's free stuff um but you know anybody who's concerned with their money could could directly go to um you know a

blockchain explorer see that okay this is a legitimate address and here's the the code that it's running um like what what exists for a second chain solution or uh sorry a level two solution so a lot of the difficulty in designing these layer two protocols is is thinking about the fraud case um and if you hear a guy like taj outline how lightning works or how discrete log contracts work you'll see that that they think about almost every case where somebody lies um you know somebody tried to slip the wrong signature in in um you know various places um and like lightning i think it basically you're you're you know lightning is a trustless protocol and that's because um if anything goes screwy you can you arbitrate with the main chain so

at every step within the lightning protocol you're swapping a signed transaction right and you can take that signed transaction and then go broadcast it on the main chain and get out um so i i think that's the key and the same thing with um with you know atomic swaps basically the idea is that this action that you've opted into consciously happens atomically or it doesn't um and then the idea with um scriptless scripts is that basically um you know you're you're committing to a certain operation within the public key and um you know in say sending someone a bitcoin you get revealed to you um the elements of something that'll that'll give you what you want provably so um i don't think anybody's proposing second layer protocols that that

require um any kind of mitigation of trust or trustlessness um i think it's more a matter of convenience like you know you might have to pull and watch for fraud or you might have to make sure that you're able to broadcast um your your get out of dodge transaction fast enough um i think it's more of a convenience aspect versus like a theoretical um shift in trust model yeah and i i mean i guess um to to some extent i i sort of understood that um you know going into that question i guess what i want to maybe clear up is is um why is it that that trade-off is superior to various other trade-offs to make things scalable at a and still be trustless yeah that's a good question i i guess because um it's opt-in so um your base case is always

bitcoin's trust model which is really strong and conservative um and you know if you want to you can decide to allocate some some percentage of value to to whatever trust model sounds good um but there's kind of limited risk there if if that trust model goes awry somehow i think in in a startling event we agree um because yeah i i agree that um you should opt into as much security as you want and i think you know one of the common um attacks against like sharding is um oh well you've just you've just halved or whatever percentage cut your network security but you know there are ways of orchestrating between shards to to validate each other and i think that's like another opt-in situation let's dive into sharding because this is something that

befuddles me like so the transition from pow to pos how do you see that playing out this is on ethereum in particular so for those of you who don't know ethereum has plans to transition from proof of work to proof of stake uh to a casper implementation there's two competing casper implementations right now right so from my understanding casper will ship with both implementations okay and there is some mechanism by which you you indicate what you're on honestly that that is a little outside of my understanding um just to clarify though proof of stake and sharding are are separate they are features i don't i think you can get one without the other like they're they're being developed in tandem rather than because sharding you can do sharding

on like typical databases right now correct or yeah yeah definitely um yeah the the the name sharding comes from the common technique most websites you know obviously facebook they have a billion users they can't keep that all on one computer so it's just the idea of splitting work and data across multiple machines sometimes with redundancy to ensure that either the data is correct or that if you lose a machine you don't lose all of its data but and what's interesting to point about point out about sharding is that it's it's only an effective technique because typically as a data designer you can choose what you're sharding on so you you choose some feature of the data that you think is going to mean that you're not you're not going to have kind of cross

interactions between you know separate shards or or maybe you want some kind of uniform distribution of data access across shards but the point is that you're designing with with some kind of heuristic in mind that is going to tell you whether the sharding scheme makes sense. And I think doing that in a general case, to me, sounds kind of dubious. Yeah, it's difficult. I mean, one of the things that we as engineers have to kind of shift our mentality about is we don't know the use case for 100 years from now. You know, if you want to build software that's going to be around, you have to think pretty

agnostically about this stuff um and and i will admit you know having read a lot about sharding there there are a lot of if not answered but like a lot of you know unknowns about exactly what the most effective way for uh shards to communicate with each other but it is planned for and and you know i think um these things are are open-ended still i think um there would be a level of hubris in anybody to say i have the solution that will work forever and i think these protocols are they're ever shifting as much as you know i think uh the bitcoin core team would like to pretend that it's going to be soft forks from here on out there will be occasions that rely on hard forks like what i don't know like zk snarks or something you can do that as a soft

oh really all right yeah see this is where james rocks me so i'm not gonna i'm not gonna duke this out especially not while we're recording because i'm just gonna make myself a fool but um yeah that's what i'm here for yeah uh marty's just gonna hit the edit button on that one for me penis no somebody out there on the internet can correct me if i'm wrong um but basically because of uh The way that SegWit works, it kind of virtualizes the scripting system. And so I think conceivably you can do zk-snarks without a hard fork. So what happens, because as I understand, hard forks are when you loosen constraints and soft forks are when you tighten constraints.

Yes, that's correct. So what happens if I'm running a node and I hit one of these newer, you know, like an opcode or something like that, and I don't know what to do with it? So, if you're running an old node, I mean, the way that SegWit worked is that the transaction, if it's seen by a non-upgraded node, sorry, we were motioning at another wine bottle, which is exactly the direction this conversation needs to go in. If an unupgraded node... 2015 CaliCab. That's right, LeaseFitch. CaliCabs all day.

if an unupgraded node encountered a segwit transaction it was considered anyone can spend and so in that way um segwit was it was a tightening of the rules basically um so yeah i mean there there may be some factor uh of zk snarks that i'm unaware of but i don't think i mean it's i don't i don't know that it would um well maybe the setup aspect might require But generally speaking, I think validation of a transaction is now kind of virtual because of the way that Segwit works. So there's a lot that we can do without a hard fork necessary. All right. You heard it here first. Bitcoin is done with hard forks.

No more hard forks ever. James O'Byrne certified. You can follow him at JamesOB on Twitter and yell at him the next time there's a hard fork. Well, let's get into this. I'm not going to say whether or not there's not or is going to be hard forks in the future, but it doesn't matter if you own Bitcoin right now. You own the UTXO set. If it does have a fork.

I think hard forks are a healthy thing to happen. Yeah, I would agree. The majority of the people involved in keeping this thing running agree that it's a good thing to have happen. Let's dive into it. Let's dive into why most people hate Ethereum and why I was an Ethereum believer before the DAO hard fork. Oh, man. I knew it. I knew it was coming.

It had to come. What did you expect? Oh, no, no. In the same way that James does not shy away from price talk, which, by the way, I will never comment on price in a recorded format. you know at your own risk ladies and gentlemen um but i i have no i have no concern about that because it wasn't just the ethereum foundation going yo we screwed up or somebody screwed up like it pretty much was though it pretty much was like there's chat logs of vitalik going to all the exchanges like all right you guys can stop trading now we're about the hard fork and then the vote

that they had where they reached consensus was like 1% of the coins in circulation. I think it was maybe more than 1%, but it was like on that order. Yeah. But I mean, if there were, and there were, because obviously we have Ethereum. In their defense, most people that owned Ethereum

probably weren't technically competent enough to participate in that, but it was probably speculators who were not technically savvy. I mean, I also think that is probably true if Bitcoin came up with a contentious thing that most people would not be technically sad i mean i mean this is the truth of any vote right i mean who you know who among us is well versed enough with like agriculture and you know whatever

to to to be able to vote on like certain i'm actually very tariffs or culture oh yeah i'm sorry i'm sorry fanquake uh the bitcoin core um this this guy is an unsung hero fanquake is a longtime core contributor and um uh this guy i met him a few weeks ago but he's super super interesting he lives in western australia and he is a farmer and he runs an autonomous farming operation where all of their machinery is is autonomously operated but this guy is a hero because for every pull request that comes into bitcoin core he tags it with relevant labels and he does a lot of categorization work on the build system and it's like it's almost instantaneous is every time you open a pull request he's somehow like awake and conscious and looking at your work

despite the fact that he's running a friggin farm in australia that's fucking cool shit um going back to the dalhack though yeah sorry to get a soft talk no no he almost saved me we almost stopped talking about it no but like going back to it like it goes back to this is what irks me about ethereum like the way they marketed it like again like it's heavy marketing on the front end and then tech on the back end of my mind and the whole unstoppable code is law meme that they sold ethereum on and as soon as the dow hack happened and unfortunately solidity was it a shitty programming language that allowed for a leak in the contract on the dow and specific specifically like code is supposed to be long like from an economic perspective like

if you're going to market that like if you're you're going to like market your system as code is law what code is written is going to happen and then the first time some shit hits the fan you're like actually code's not law we're going to change the code well i so i i still believe in the code is law thing but law exists to serve the people that it governs this is where we get into like social right right yeah i mean there's a lot of semantics that we could get into but basically you know the laws that this you know that any country started with are not the laws that are there today laws exist to serve the people and if they no longer serve the public interest we amend laws all the time you know and i don't see why what potentially could be the platform

you know of of many things of the future is to be immutable that would be ridiculous you know if i If I were going to build my product on top of something and I were to hear that there would be, you know, no room for change even if everybody else wanted it, I wouldn't build on top of that. You know, things need to be dynamic, especially in software. Well, that's my whole point is push that shit to the second layer. Like, the protocol layer should not – like, code should be law. You should, for certain, be able to tell at some point in the future that these things are for certain. There's going to be 12.5 Bitcoin traded every 10 minutes until block 630,000. We're not going to reverse a transaction on the blockchain because somebody got money stolen from them.

Again, I just lost my train of thought. well let me all right so let me take us on a brief digression here into um something a little cosmic oh yeah a little cosmic that's good cosmic john we're getting cosmic in memory rip john um this i think highlights why uh culture is really important in these systems and so you can say okay so the dao hard fork happened um from a technical standpoint it was a hard fork So, if you don't agree with the DAO, you can just mosey on over to Ethereum Classic and continue on with your chain. Sitting at a solid, I don't know, $30 or something? Yeah, yeah, yeah, yeah.

Exactly, right? So, technically, that was your outlet if you disagreed with that change. But in actuality, the system moved a certain way, even though technically it was capable that you could opt into the other system. So I think that highlights, like, a lot of the more nuanced discussions that we have around Bitcoin design are in, culturally, what is this going to do? So a great example of this is a little-known feature of the Bitcoin system right now, something called checkpoints. Marty, you ever read up on this stuff? I have heard about checkpoints. I have not read deep enough into them to speak with any authority.

So basically, the gist behind checkpoints is there are hard-coded hashes in the Bitcoin Core code base. And we say, basically, at height x, we expect the block to have hash y. And if it doesn't have that hash, then something has gone wrong, and you're downloading from the wrong peer. So this opens up a pretty interesting discussion. because you think to yourself, okay, well, that's pretty, you know, that seems reasonable. That seems practical. If we have a reorg, basically, if someone reveals a fork of the blockchain that is longer than the current blockchain that wipes out, say, you know, two years' worth of history,

well, then we should pack up our bags and go home because this little experiment is over. So the idea of checkpoint sounds totally reasonable, But the contention is that it introduces this cultural element of the developers basically codifying what the right chain is. And is that a road that you want to go down? You know, people in Bitcoin are very averse to that. I think often the culture kind of dictates that consensus is emergent and that the developers don't define which chain is the right chain. other than the fact that the right chain is the longest valid, I'm sorry, the most work valid chain. Another kind of issue on the technical horizon for us is thinking about some kind of fast sync equivalent.

So one option for doing fast sync in Bitcoin land is that you can commit to the hash of the UTXO set within every block. So that when you're doing your initial block download, instead of having to download and validate everything, you can just download headers and then say, okay, this header chain is valid. And I'm going to take the hash of the UTXO set for the last header that I downloaded and download that entire UTXO set, which is roughly three gigs right now from my nearest peer, you know, and that would allow you to sync much more quickly than downloading the entire blockchain and validating it. But that's slightly changing the cultural and trust model. You know, you're now talking about having miners include this hash that's relevant for IBD.

So I think culture is really, really important. And I think a lot of what weirds me out about Ethereum is that there are these very vague elements of their culture about the trust model and about things like long-term viability that... What's vague about it? I think the idea that, like, proof of stake uh has demonstrable drawbacks um and concerns and that are kind of fundamental

to the nature of proof of stake and people still kind of go on with it not to go too far in the rabbit holes but would you mind elaborating for audience no this is an important discussion to have because this is i mean ethereum the ethereum project is banking its future on proof of stake which is unproven up to this point right so the classic one is called the nothing at stake problem and that's this idea where you can attempt to submit fraudulent data to the chain and get penalized for it but if you succeed then you can basically roll back those dings that that you were hit with and thereby kind of removing the negative incentive to cheat And to my knowledge, this is why Ethereum is right now either entirely proof of work or a combination of proof of work, proof of stake.

So in this case of rolling back, I mean, so if you have some number of nodes who have already accepted that you got slashed, slashing is when you say, hey, you tried to lie to us and we're going to take your Ethereum now, the Ethereum that you staked. But the rolling back, how is it that you roll back further than what block is on your plate now to validate? Now, does that require, like, a mass takeover of the network where you're feeding bad data from multiple blocks? I don't think so. I mean, maybe it's along the lines of a reorg where you just introduce another chain that's longer and still valid and then feed that to your peers. So to my understanding, proof of stake, it's sort of this round robin consensus thing where a handful of people are told, hey, divine us the next block. Those who have staked money on it and then and then kind of among them, the largest consensus is the next block. And anyone who diverged from that, if they were breaking rules, get slashed.

so if you get a few blocks deep i don't understand how you can propose something i mean i think that's that's a basically 51 attack right you have to have enough people not slashing you in order to allow for this longer chain which to me you know i i mean it's like anything it's levers that you pull um Um, one of the things that I think Bitcoin has relied on is, is, is hash rate. And that's always, you know, kind of the, for lack of a better word, this thing that's advertised, it's got the highest hash rate. You know, we, we're so strong. Um, Bitcoin beefcake.

Bitcoin beefcakes. Um, so, so, uh, where was I going? I've lost my hash rate. Yeah. You're talking about reorgs and 51% tax. Right. So there your risk is 51 percent of hash rate, which, you know, it's like an attack that people talk about sometimes is obviously the quantum computing thing. I think that if you were to have picked Satoshi's brain way back when, he would have never guessed that like A6 would have been a huge thing and that China would have been like, you know, one of the hugest miners and that three of the largest miners control more than half of the network. Like there there is a lot of friction to working in meat space. And so I think the intuition to move more of this into a frictionless environment kind of sits well with me. Wait, what are we moving into a frictionless? And what is frictionless? Does frictionless exist?

So I would call proof of stake frictionless because it deals with fewer real world possibilities. One of my favorite things is the Bitcoin Dyson Sphere, surrounding the sun with solar panels to power some ultra-powerful computers. We have to get the Thorium reactors first before we get the Dyson Sphere. It's Marty's part-time project. Yeah, while everyone is throwing economics books on your reading list, I'm going to throw out Three-Body Problem because that's my jams. I'm taking us into sci-fi literal cosmic. cosmic getting the three body problems who's that by um that's uh shu uh shu lin or shu lu maybe uh shin lu shin lu man i'm i don't know yeah i'm getting three body problems getting right

that here three body problem it's a trilogy greatest sci-fi i've ever read chinese author talks a lot about um uh the game theory of the universe yeah game theory of the universe holy shit i've never heard it's a deeply terrifying book to read i feel like we're about to dive into that with the glass bead game too just read the introduction to that like oh man yeah we haven't even told you guys about the glass bead all right i'm gonna let you lay that down hard after i went to three body and i i agree i agree i want to i want to keep going with the three body problem because this is another one of those topics that we completely agree with across the sorry it's just the dyson sphere made me think of it there's there's a lot of crazy theoretical

science going on in that book it's a really yeah it's a really interesting book because and it kind of mirrors cryptocurrency a little bit because the author basically really investigates the incentives of of life across the universe and makes it into an incredibly hostile and dark place um and plays out those implications in in um fairly vicious and inspiring ways um so highly recommend the three body yeah surprisingly large amount of analogies to cryptocurrency space yeah anyway so uh what i was describing with a frictionless environment is one in which you know so like i was saying satoshi probably could not have imagined a6 probably could not imagine a dyson sphere you know around the sun but when you put something into software

um a lot of the unknowns become known because you sort of define the space now there are unknowns in there that's basically what a bug is or at least a good bug a bad bug is somebody made a mistake. A good bug is there is an unknown in the way something operates and someone exploits that. So proof of stake largely moves your staking out of meat space. So, you know, having a server farm that mines into virtual space and removes a lot of the unknowns. You know, what if someone in an attempt to do a 51% attack were to fire missiles at all of the, you know, physical mining devices like you just you can account for that on a network whereas you know in in in a proof of situation like most of those kinds of physical possibilities are accounted for all right so

a few thoughts here um the first is that when you when you you know say like you were to move to proof of stake and shift your trust model in that way fun fact um meltdown inspector could be used to steal from hot wallets pretty trivially so you now meltdown inspector for those that you don't those of you that don't know is the inherent uh backdoor in intel chips correct not even just intel i think um i think it may be any chip going back to 1995 or something like that basically so so when you know programs have what we call branches and that's like you know one of n ways a program's execution can go and what what chips do to speed up execution is something called branch prediction so sometimes they'll execute something kind of in the quote-unquote future of a program

even if they're not sure that it's going to happen and they then cache the result so that they can serve it up to you if it actually does happen and this has certain performance benefits So anyway, Meltdown and Spectre can be used to read arbitrary data out of memory from a process that isn't yours. So if you're staking and you have your funds tied up in a hot wallet, Meltdown or Spectre could have resulted in a confiscation of everything on the network, potentially. When you have a SHA-256 miner, that's an incredibly dumb machine. SHA-256 is an incredibly simple thing. And I think that's what I love about it is that, in a way, the simpler and more straightforward your hash function, not only are the incentives easier to reason about, but the mining hardware itself can democratize.

And I think, as Marty points out a lot on Twitter, which I love, is that ultimately when mining hardware commoditizes and people kind of saturate the technical innovations that go into making a SHA-256 squared chip, you're going to get to this point where the capex on running a mining farm is is kind of like dwarfed by the opex and all of a sudden you don't care if your mining hardware is only utilized half the day so all of a sudden now you've got a solar array in the desert that's mining bitcoin for you i don't know if it's exactly going to pan out that way but the point is that it's it's an incentive mechanism to encourage development of alternate energy that isn't just kind of burning burning oil which is something really cool and you're you're hitting all of my you know

love of sci-fi potential future outcomes but i think that the reality right now is that that is not what's happening what's happening is that cheap electricity means hash rate and that's why you have places like china that have an unbelievably large hold on hashing power and while i want to believe that future um i feel like reality has a way of always twisting your expectation and so you know i i kind of like proof of stake as a rejection of reality and it's saying we're going to build our own reality in which we can define the rules by which we we hold people accountable for their actions um i don't pretend to know you know all the ins and outs of the implementation um and whether you know certain vulnerabilities are going to hit it

i think one thing that i believe will happen is we will see specialized staking hardware that will you know it's not going to be an operating system it's not going to be uh you know also like playing video games and browsing the internet it's going to be a piece hardware that is is a little roi machine you throw a hundred eth on that thing and it gives you back five percent a year something like that yeah and that's so that's my main beef with pos so like full disclosure like i said this on the last podcast like just for shits and gigs like i staked a coin a pos coin for two years and it's just like again going back to heuristics like just it doesn't it's too easy to an extent in my mind like it's too easy like i gotta say marty

That's the weakest insult I've ever heard. Oh, it's just too convenient for me. No, it's like, it's just, again, like, again, going back to, like, heuristics, like, proof of work. Like, I think you need to work. Like, this is, like, going back to, like, existential, like, being a human, like, and how do you, like, attain what you assume is self-actualization, which is, takes hard work and determination. and again going back to like proof of like this is i don't know if this is a good analogy or not but like proof of work the fact that we're expending so much energy and putting so much value into a proof of work system gives it a lot of value whereas proof of stake you just

flip up like your macbook and just let it run for two years and you make money doing nothing but hoarding like and again proof of stake incentivizes hoarding in my mind which and like that's a big that's the big meme and bitcoin is hodl which i would argue is different than hoarding which proof of stake enables and encourages well so i mean you could call it hoarding or you could call it investing which is you know something our modern economy really values a lot is is investing and that's investing in the security of the network and you you get yields from that you know there is still liquidity in the market by way of fees and those fees go to you they don't go like directly back into staking you can do whatever you want with it likewise

when you talk about the ease of access yeah it's easy for you to flip open a macbook and and start you know staking a shit coin but we're talking about thousands of dollars here which there is still a barrier to entry though i like that it is not a physical barrier i don't have to go you know in new york it's impossible to mine electricity is expensive space is expensive and like my apartment gets hot enough in the summer you know so i'm not i'm not gonna go mining but i can i can stake um regardless of where i am if i have the investment in the network you know that's really the main benefit to me is it just feels like it levels the playing field on a lot of things i hate the idea that i would be mining at a loss in america

i got a mid washington state yeah yeah no i and so another interesting factor here is in proof of work um quickly your your your limiting characteristic becomes um heat diffusion uh so so like to to get back to what marty was kind of getting at um when you're doing proof of stake all you need is a lot of ethereum but to do a giant proof of work operation you actually like there's not that economy of scale there's an anti-economy of scale because when heat diffusion is your limiting factor you need a giant space you need to go out and get a bunch of land and so that ceases to become just having a lot of bitcoin and actually that anti-economy of scale becomes a forcing function for decentralization so as soon as as soon as mining commoditizes i'm gonna go

out and buy a how long miner or you know whatever it is that comes around and and and run my little miner because that will be reasonably profitable yeah but you're also the guy who quits his job to go work on bitcoin core like you're pretty hardcore you know i it doesn't matter what the investment opportunity is for somebody even if there are gains on it people are you know um they're into the value that they can see and if you tell somebody plug in this computer um and it'll make you money i don't think that's going to reach mass adoption i mean there was a there was a wi-fi startup or an isp startup a while ago i can't remember the name but their idea was plug in your router and set it to be open and by setting it to be open we'll subsidize

it because what they were doing was charging people to access these open routers and the idea was... You mean like open some ports or something else? No, no, no. Like, you know, no WEP key thing. Oh, okay. Basically everybody was a Boingo hotspot. Yeah. Right? And that was

their idea. And nobody did it, of course, because there was just this inherent like, oh, I don't want to open my network up. Which, don't get me wrong, it's a good attitude to have. Shoutouts to WEP2. It's good stuff. Shoutout WEP2. I guess we'll shout out

Marty's not in on that one usually I feel like I can I can rally a shout out to pretty much anything but that was a pull let's go web to

is that bar still Louie that that originated from is that is that a Louie meme I don't know I'm gonna give it to Louie alright

shout outs to Louie shout out deep web Tim deep web Tim he's probably loving this conversation deep web when are you coming

on the pod we've been asking for a while yeah man uh but anyway um i i don't know i i think i hear a lot of these like lofty answers to to really a lot of the questions that i have about not just proof of work but a lot of like bitcoin things and there is always an answer and i i would almost call it a canonical answer because i hear it from more people more than once but they don't always feel like realistic answers to me

they feel like like coming up with an idea in a frictionless environment but interacting with a very frictionful environment which involves you know real people well that that's literally my argument against proof of stake i think human systems are more complicated than thermodynamic systems and proof of stake is inherently a more social thing than having an asic chip that proves that you've done something but we can agree it goes beyond thermodynamics it goes on you know if if the president decides to put tariffs and taxes on on on foreign imports well now all of my hardware gets more expensive and now the decentralization fails because it's cheaper to run this hardware everywhere else in the world that's fleeting to an extent because it's only

for i would argue that's like that that could be temporary well network security is not a temporary issue if you have if you open yourself up to a single 51 attack as james puts it you might as well go home and i'm going to cut it there for right now i don't want to bore you guys with uh too much information at once we'll be back tomorrow with the rest of this conversation