Today's episode is dedicated to my first son, Alexander Samuel Rochard. He was born on February 12th 2018 at 9.08 p.m. 7 pounds 19.2 inches. Welcome to the Noted podcast. This is episode 0.9.0 and today I'm joined with Jameson Lopp. Unfortunately our co-host Michael Goldstein was not able to make it today. And Jameson, how are you? Not bad. Been keeping pretty busy. I spent most of this week actually at Satoshi Roundtable. Fantastic. And usually I start these interviews by going into the beginning like when you first got into Bitcoin. But I actually want to start with the opposite now which because I just saw that you had tweeted out about a new rig that you've got and you were able to sync your full node and how long was it? I believe it was 152 minutes. That's astounding because I think that I now granted I've never messed with increasing the DB cache but just like on my own computer I think when I did it it took me a few days at least. So at this point what would you say is the bottleneck? Yeah so you know I was testing a few things. Now it's important to note that like I'm on gigabit fiber so a lot of people might be bottlenecked by their bandwidth and their ISP. I am not really bottled by that anymore. It's probably the the CPU and the actual disk storage that are going to be the bottleneck. So I maxed out and got like a new six core i7 CPU but I think the more important thing is that I got one of those NVMe you know like pure RAM solid-state drives that are like 10,000 IOPS per second or maybe a 100,000 I don't even know but they're screaming fast. So I think that plus jacking up the DB cache I put 32 gigs of RAM in it and jacked up the DB cache to like 24 gigs but even once you get all the way through the whole sync it's only using maybe around 12 or 13 gigs. So that's I think about as fast as you can probably sync on consumer hardware right now. And what is it storing in the RAM when it's doing that initial sync? Should be pretty much just the UTXO set so you know that's what pretty much all of the the database intensive write operations are just you know deleting UTXOs as they get spent and adding new ones. Gotcha and so I think that the you know this is the noted podcast so obviously this subject is very apropos and one of the contentions we saw last year with the scaling debate is on the subject of the initial sync time for nodes and that the fact that technology is advancing so quickly that really we should be increasing the block size limit to keep up with these advances in technology and what you're presenting to us here is a solid case in that direction. Kind of. This is the you know optimal best case scenario and so I'm fortunate to live in an area with gigabit speed internet and I you know have the financial ability to buy a $2,500 computer but you know on the the exact opposite end of that there's gonna be people who are still on pretty slow broadband and are not gonna be able to afford anything much more than like a mobile phone or Raspberry Pi so I run both and the the Raspberry Pi nodes are struggling right now so it's good to you know get a better idea of the entire spectrum. Right and it's interesting that I mean you got a cutting-edge rig for $2,500 I think we we heard some people saying that you would need to spend ten thousand dollars on a rig for having your gigabyte blocks it's unclear to me where that ten thousand dollars is going. Well and then the other thing is when when people say stuff like that most of the time they're talking about what would be required to stay synced at the tip of the blockchain they don't think at all about the initial sync time of what happens once you have a year worth of gigabyte blocks like it's gonna take you a long time to chew through that even if you have you know crazy like industrial server that cost you a hundred thousand dollars and uses you know highly parallelized CPUs and disks and everything so. Maybe the vision is like Amazon has a service where they ship via shipping container all of the data rather than over the public internet so maybe that's the vision for spinning up a new node is that you get a container delivered. Oh yeah I mean if you want to get really creative any sort of transportation mechanism for data is possible here it doesn't have to be over the internet and in some cases it is faster to ship via truck or I think that like Google had like a carrier pigeon parody thing they were doing or they were showing like how much bandwidth you could get like strap a USB stick to a carrier pigeon or something so you can get pretty creative with it you can even set up a satellite and beam it down from space if you want. That's right all right so let's rewind and I think that we've known each other on Twitter since like 2013 and this is our first time talking with each other we haven't met in person yet I'm sure we will at some point but yeah so did you initially get into Bitcoin in 2013 or before then? It was like summer fall of 2012. Okay and what initially like were you interested in it from the libertarian angle or the technology angle what attracted you to Bitcoin? It was about 50-50 on the libertarian philosophy and computer science stuff and you know I'm sure that I had heard of it a few times before then and just dismissed it and for some reason I think it was a Slashdot article I realized oh this thing keeps coming back and so I started looking into it more and read the white paper and realized that like it was just a very different system from anything I'd ever seen before and that it was pretty elegant in the way it was designed so that was kind of what hooked me and got me started running nodes and you know sending my first wire transfer to MT Gox and eventually led to forking Bitcoin core creating Satoshi and you know trying to offer operational insights to the developers with that. Yeah you want to get into Satoshi because I think it's a really interesting project and and shows how much one developer can can do in this space. Yeah well and I think it's interesting because I don't really spend much time on it you know I maybe spent a few dozen probably less than a hundred hours total on it and most of that was in those first few months where I was just trying to figure out how Bitcoin core worked. So Satoshi is pretty simple from a technical standpoint it's Bitcoin core plus about 500 lines of instrumentation code and a few hundred of those are the actual stats D library and then the rest is just scattered throughout the code base whenever an interesting event happens maybe a transaction comes in a block comes in a peer to peer message comes in or goes out tracking all of the different events and activities within the node then I just emit some statistics to stats D which is a local daemon that runs and aggregates all of those stats up and then those get stored in database that is hooked up to a very pretty Grafana dashboard on Satoshi.info so you get some nice eye candy and basically the idea was that you know Bitcoin crypto assets it's all about openness and transparency and while the code was very open and a lot of the other metrics of like the the network itself like the coins and whatnot were were fairly transparent through other block explorers and other projects the actual internal operations of the nodes and what they were doing was very opaque and so I wanted to understand it better I figured that other developers also wanted to understand it better and I think it's been fairly successful because a number of Bitcoin developers over the years have referenced stats and graphs from Satoshi and I think it's helped them make more informed decisions about what needs to be done to the nodes it's interesting to me that you mentioned that it's taken up so little of your time to maintain because like once you set up a piece of software correctly oftentimes you can just let it run and it'll keep doing its thing for a long time but the challenge that like I would imagine in this case is if this is a patch set on top of Bitcoin core and they are constantly refactoring the underlying code has it been a challenge to keep up with that and rebase essentially yes that is where the majority of my time comes in I haven't really been spending a lot of time you know trying to add more metrics recently unless someone comes to me and says hey we really need to have this and it's easy for me to figure out because quite honestly I'm not a C++ developer before I started tinkering with Bitcoin core I hadn't done any C++ development in ten years you know since I was in college and even today very rusty on it so it's still a bit of a pain for me to get in there and make changes and then you know allow them to actually compile and get the test to pass but yeah I would say you know each major release that comes out I merge all of the changes since the last release and you know sometimes it takes me an hour to fix those conflicts sometimes it takes me an entire afternoon I think for the point 15 release it was more like an entire afternoon because they've been refactoring so much and like splitting out the main dot CPP file and moving just code around to make it more modular that it was a real pain so you know from that standpoint like if you want to talk about like altcoins and their maintainability you can go out there and look like the vast majority of altcoins are pretty much stuck at the point where they forked off years ago unless they have a very experienced maintainer who stays on top of it and keeps pulling those changes down and of course the more changes you make locally the harder it becomes as those code bases diverge right and so really though if if these altcoins had their consensus critical code that was what differentiated them from Bitcoin in an isolated manner then they could much more easily keep up but that would also prevent them from making their own changes that really like what's the point of having an altcoin if you're going if you're not going to make your own architectural changes or whatever adding features like Zcash does which involves touching a lot of different parts of the code base and making those merge conflicts increasingly challenging to reconcile so then the other let's see I lost my train of thought oh I wanted to get into your your background as you said that you weren't very familiar with C++ so what was your development background before getting into the wild world of Bitcoin so I've been a web developer of sorts ever since I got out of college and I really started more on the front end with design HTML C++ JavaScript type of stuff more user-facing and over the years moved further and further back so after a couple years I started doing more lamp stack development and then a few years after that moved even further back into doing cloud computing like large data scale processing with Hadoop clusters and other highly parallelized type of software because I was working for email marketing company that was sending out a hundred million emails a day and then ingesting just a ton of analytics that you know the customers wanted to get analyzed in as close to real-time as possible and it it was never really a passion of mine but it provided a lot of interesting technical scaling challenges and so that led to me to be a primarily Java developer and then once I went full-time at BitGo three years ago I was lucky enough that their infrastructure indexing services were also Java and Bitcoin J based and so I've pretty much stuck to Java for the past three years as well they'll also been doing a bit of Node.js and a little bit of Go. Awesome so you mentioned you joined BitGo three years ago and do you want to describe like what BitGo does and what the vision for the company is in broad terms? Yeah so BitGo doesn't get a ton of press or recognition in the sort of mainstream consumer crypto space and that's because we're really an enterprise focused company so our our customers tend to be large services that have to operate hot wallets so that tends to be exchanges and payment processors and anyone who is receiving or sending fairly significant volumes of crypto that have to be automated because security of course is one of the top concerns in this space if you screw up there's nobody to save you and if you need the ultimate security it's very simple you just take all your private keys and take them off the internet you've got your cold storage so you've now pulled all of your security concerns into the realm of physical security which is a fairly well understood problem but if you have to have automated withdrawals then you've got a hot wallet that is running code somewhere on the internet that some hacker could potentially get into and screw with the code and cause you to automatically send a transaction to them and now all your money is gone so the way that BitGo operates we provide a multi-signature security model where instead of just having a single private key on your computer that's signing the transactions there's instead three different keys and you have one key on your server BitGo has one key on our servers and then you keep a third key either stored offline yourself or stored offline with another third party which is basically a recovery mechanism now when you want to make a transaction you're gonna create the transaction and half sign it on your end and then send it to BitGo and BitGo will have to look at all of your recent activity look at various security policies that we've set up that you set up and decide whether or not to co-sign that transaction and until we co-sign the transaction it's not valid because all of the wallets are two out of three multi-sig so if anything is wrong then all we do is nothing and we alert the user and say hey you need to go check your wallet you might have a policy violation but if everything's fine we co-sign it it becomes a valid transaction and we broadcast it out onto the network and hopefully it gets confirmed in a timely manner if our fee estimation and UTXO selection and all the other stuff worked out well fascinating and so does does a BitGo customer they have their UTXOs are essentially segregated right it's not like yeah Coinbase where all you know it's kind of a layer above yeah so BitGo is non-custodial so each time someone creates a wallet they have their own UTXO set for that wallet and nobody can mess with those UTXOs unless you know you use two out of the three private keys in that setup and it's it's an interesting model from a regulatory perspective as well because it means that we're not a financial services provider we are just a software service provider and it's also interesting in the sense that if a hacker gets into the user's computer then there's not enough private key data for them to steal the money if a hacker gets into BitGo servers there's not enough private key data for them to steal from our users and even better if BitGo ceases to exist for any reason you don't lose anything you just go get your recovery key and we have open source software that you can use to create a transaction without ever talking to our servers so you can just route around us and does this use the standard Bitcoin networks multi-sig feature or is this something else yeah this is a two out of three P2SH and so that's interesting because essentially the the customer is not custodial you know assuming that their backup is put away and you're not and I think that it's funny because it just reminds me of the Bitcoin network in terms of decentralization where because it's decentralized no single entity is a money transmitter you know like some people might argue that like oh the miners are money transmitters and thus should be regulated kind of saw like crazy statements like that and we even now we're hearing this about lightning right where people are like oh lightning nodes are money transmitters but it it gets to the fact that like one of the cynical views on Bitcoin is that it's a form of regulatory arbitrage by being decentralized you can avoid these labels that come with a lot of government imposed regulations and Bitcoin like you know people you'll see politicians going hey we need to regulate Bitcoin and then they realize oh wait it's not so simple but you'll learn eventually you mentioned though that you know your software company and yet recently we saw that you guys acquired a financial services company do you want to describe that transaction and I don't know if there was kind of a public like vision put out for why why that transaction occurred or if you're allowed to speak to that but curious um I don't think that we've done any huge like press releases regarding the long-term vision but I mean it's pretty obvious you know we raised a series B of something like 43 million dollars and we then use some amount of that I don't even know how much to purchase Kingdom Trust and Kingdom Trust is actually one of our larger customers we've been working with them for a number of years they run the Bitcoin IRA and they have something like 12 billion dollars worth of crypto assets that they are a custodian of on behalf of people who want to hold various crypto assets inside of their tax advantaged retirement account and the the main reason for this is that you know we do see a shift in interest in the ecosystem where a lot of institutional investors are becoming more interested in crypto assets but and this is highly ironic I believe I don't know if the irony is lost on them but due to various regulations around like financial compliance in the way that these institutional investors and various funds operate is that there has to be a qualified custodian of assets in order for the investors to hold the assets or more specifically the custodian holds the assets for the investors and so basically we were having to pass up a lot of deals from these type of institutions and clients because we were not a custodian and even if we were technically a custodian we were not you know falling under the like financially compliant you know regulated custodian model and I know we looked into a number of different things and like trying to set up trusts and all of that and at the end of the day the most efficient way for us to be able to acquire that status was to just acquire Kingdom Trust or some other entity that was already you know reputable and licensed and basically performing those services so we're just folding them under the arm of BitGo and we're gonna continue to expand our offerings into the custodial front but we are by no means becoming a like fully custodial company we'll continue to operate you know hot wallets for people in a non-custodial fashion. That's really interesting yeah and I think that it's it speaks to the increasing realization that crypto but even Bitcoin specifically is a long-term investment and it's kind of it's caused I think a bit of a philosophical rift within the community because whenever people talk about Bitcoin as an investment there's always at least one comment that's like guys this can't grow this Bitcoin can't you know appeal to more people if people are not spending their bitcoins and you know using their bitcoins so that they can buy goods and services. What's your view on that like tension between hodling and using quote-unquote Bitcoin? Yeah I mean well first of all hodling is using so that's one of the big disconnects but it's been interesting especially you know as we've had so many bubbles over the years and the exchange rate keeps going up by orders of magnitude and and sometimes I try to look back on it and think like why do I still have almost all of my coins like I didn't I didn't sell any crypto assets before this past year in 2017 and the only ones I sold were all these altcoin air drops so that I could just buy you know get more bitcoins from those of course that resulted in me having a big tax bill which is not fun but I I can see why you know some people are in situations where they need to have a censorship resistant form of money or they need to be able to make payments globally quickly without going through traditional systems but I think the vast majority of people that I meet and talk to don't really fall into that camp like almost everybody I know and talk to has credit cards and they can do wire transfers and use Venmo and PayPal and blah blah blah and you know for the vast majority of their day-to-day transactions it's fine for them to use those systems like I I don't think I need to worry about you know Visa canceling my payment to Amazon for buying some books or software or whatever so like personally I see the majority of the utility and value in these systems is in holding my wealth in a fashion that is not confiscatable by anyone and this has actually become even more clear to me over the past few months as I was going through the process of selling these forks and converting them into bitcoins and converting a fair amount into dollars to pay the tax bill because I ended up spending well over several days in aggregate in December trying to get dollars from all these different exchanges into my bank accounts so that I could pay the IRS and I had problems at Coinbase and G Dax and Kraken and Gemini is like every exchange as soon as I started to try to use the traditional financial system again was just screwing with me and made my life a lot harder it would been and I hate to say this but my life would have been much easier if I could have just sent a crypto asset to the IRS and you know I think that the person who would hate to hear that is Roger Ver who is essentially through through his activism providing more funds to the IRS's people have to pay taxes on on his quote-unquote airdrop so all right and so I I think that we've heard of other companies trying to get into the like institutional custodian game like I think the Coinbase launched a product where to sign up you have to pay a hundred grand which to me is just that's the market right if that and that indicates how much demand there is for this for this kind of product and so but at the same time like back in 2014 I feel like I would always hear this is the year of Wall Street this is the year that like you know big institutions are going to start buying or this is the year that an ETF comes out and those narratives always seem to get pushed back six months or a year and the flip side is that I've these are concrete steps towards that and it seems to be getting more and more realistic yeah I think that you know 2018 really will be the year of the institutional investor getting in and the main reason that I say that is that anecdotally I have started to get contacted by wealth management firms and and all kinds of institutional investors who they they want basically to fast forward on the Bitcoin crypto education you know and so they're willing to pay me a premium for my knowledge dump to help them get a better understanding of this without spending months and months trying to understand you know the entire history of how we got here and it's funny to me that you know it's so political right because if they ask you for a knowledge dump it's going to be a very different knowledge dump than if they asked someone who works at R3 or you know who's really like in the oh it's blockchain technology not Bitcoin or distributed ledger technology and I was listening to a podcast put out by Goldman Sachs this morning and the lady was saying that oh Bitcoin and the current generation of crypto currencies they're inherently flawed because they're too volatile and they are not mediums of exchange they're not stores of value because they just went down by 60% and no one's using them as a unit of account and so you know we're we think that the value is going to drift towards zero over the long run and I like obviously you know I was raging but it did provoke some of my own tweeting and so someone who has that let's call it an economic ideology they're going to have like a very hard time having a conversation with people who are in our camp yeah and well a lot of these calls that I've been having are interesting because they're basically blind calls where it's myself and a coordinator who understands the space somewhat well enough to come up with some good questions and and we basically have a conversation and I don't even know who the various investors are you know there's like a hundred people just listening passively to this conversation so you know I'm sure that they are getting a diverse set of opinions and doing these type of listening conference calls with all kinds of different perspectives but you know if we can get at least a few of them to understand our perspective I think we'll capture their attention and I think that the most interesting question that gets posed by investor types is what do you see as the biggest weakness for Bitcoin like what is Bitcoin's Achilles heel and I would be really interested in hearing your answer for that one well I mean the the biggest weakness is apathy if if people lose interest that is how Bitcoin dies and that's also why I think that it's incredibly difficult if not impossible to kill a crypto asset that is decentralized by any fashion I mean you only need to have a handful of people that are interested enough in one of these protocols to run the nodes and secure them in some manner of consensus and and you have a quote-unquote working crypto asset now then you know the sort of range of what working really means can vary widely but as it as long as there are a sufficient number of people who are devoting their resources to continually maintaining and improving the systems I don't see how it can die if a problem comes up then we fix it and move on and that's where a lot of the anti-fragility of these systems comes into play where the system it could crash you could have a market crash it could have a technical crash where the network goes down but the anti-fragility is due to all the different actors who make up the system who then come together and say okay we we make this tweak and we bring it back online and that's something where I think I said recently if you know the network might take a nap for a little while but eventually it's gonna come back and the the heartbeat of you know blocks and transactions coming in will resume right all right and I I think that when I got this question my answer was that it really would be I was you you're thinking of it at a meta level which I think is the right approach to to reasoning about what would actually be a problem for Bitcoin and I was kind of thinking of it a little too concretely in that my answer was there well if if a government controls their internet then they could conceivably block Bitcoin and have a firewall of sorts that prevents Bitcoin from you know being a viable like granted someone could always use a satellite dish although even that I was thinking like couldn't a government be flying drones over and making sure that your satellite dish is not pointed at block stream satellite I mean I'm sure they can jam you know signals at different frequencies you know there probably is a decent point to be made that you know the Bitcoin protocol itself like the message passing is not encrypted it is you know clear text I mean technically it's like highly compressed you know binary but it could be very easily you know sniffed and filtered and blocked and that's why I guess you can make a strong argument for you know small blocks and compact blockchain size because that makes it easier to use other forms of more privacy available communication you know like a tour so if you wanted to you know run a blockchain with hundred megabyte blocks or gigabyte blocks over tour that's probably not going to be feasible but at least now with the current block size and block weight you can still do that right and on that note like what what do you see as what excites you about Bitcoin in 2018 I think it's it's all gonna be about lightning and lapse and actually like connecting all of these decentralized networks together so you know the the ecosystem has grown significantly over the past few years there's far more crypto assets than anyone can even possibly keep track of but we still have some systemic issues mainly you know these centralized exchanges which we had a pretty good run for a little while where there weren't any major exchange hacks but now we seem to have another wave of exchange hacks coming in probably just as more new entrants are coming into crypto starting exchanges and failing to learn from their predecessors so I'm interested in you know decentralized exchange protocols and atomic swapping for privacy reasons and also for like health of the crypto ecosystem basically because if you're just using software and a protocol to have economic exchanges and changing assets and trading with other people peer-to-peer then you've removed these honey pots that you know everybody puts their money into and attracts all the attention from hackers and results in a lot of people losing money due to one really stupid exploit that you know somebody accidentally put in and if you're just doing it in a decentralized fashion sure you might get hacked but if you get hacked it's not like everybody else also gets hacked so I think it makes things more robust and it also is just kind of smooth the process of exchanging so even from a not a security level but just a scalability level a lot of exchanges have had problems with their multiple orders of magnitude of growth and the exchanges have been falling over or just been so slow that they're unusable and you can't even trade on them and that's just once again because they're centralized and everybody is going into one little point so I think that you'll also see you know fewer types of congestion there if you're able to actually do your trades across a more decentralized network yeah my hang up with decentralized exchanges has always been the fiat interface which seems to rely on like me giving my bank info to someone so that they can go deposit cash and in my account which and that aspect of it doesn't seem particularly scalable to me but in terms of crypto to crypto there I can see that that's definitely much more robust yeah and maybe we'll see a rise of a stable coin that can be used on these networks you know tether has been a hot topic recently there's there's no reason I guess why you couldn't have a lightning network compatible tether now of course like any of these stable coins that are you know ultimately backed by fiat are gonna have a trusted third party single point of failure at some point but you could at least you know create this gateway where people can use that single trusted point to get into a crypto system where you have you know more liquidity and you're just able to move around a lot faster without permission from everyone so when I saw you tweet about decentralized exchanges I thought to myself well is that's interesting because aren't bit goes most you know most of their customers are exchanges like I don't want to get you into hot water or anything but it seems like you're you're suggesting that big go would be obsoleted somehow because now we don't need these big honey pots with a bunch of coins that need to be secured yeah I mean I think the ecosystem is going to continue to change in a number of different ways and people are gonna adapt and you know some of these exchanges are going to innovate and they're going to find out ways to to pull these technologies in and make their exchange even better and you know probably offer hybrid models but you know bit go we have been pretty good about staying on the forefront of innovation in this space we got you know segwit active on our platform within a few weeks of it being activated on the network and you know we are big fans of payment channel technology so we've got a team of guys that are are working on various ways that we could use payment channels both internal to bit go and eventually of course external with a lightning network compatibility so the the general idea I guess of bit go is that we're going to serve whatever market demand there is and so while you know bitco did start out as a Bitcoin only thing you know as we got a lot of exchanges coming on to our platform and then the exchanges were basically saying hey we we want this same security model for coins X Y & Z we were like okay you know we'll do the work and we'll add those and that's a lot of how we got to where we are today and now it's more turned into like aetherium and you know adding support for all these different tokens and you know hopefully in 2018 it'll be lightning network and adding support for various types of applications on there so it's going to be a fast-moving space yeah and it's a lightning network use case that I've heard of that rang as being particularly useful to me is the inter exchange arbitrage where you would be able to instantly send bitcoins from one exchange to another whereas currently it's kind of a it can be a haphazard process especially when the mempool gets backed up you don't really know how long you're going to be in limbo for yeah there's been a number of products you know just targeting nothing more than arbitrage you know we've that's really what blockstreams liquid sidechain is about and I learned at Satoshi roundtable that is the main reason why tether came about is that it was well tether was created you know on master coin slash Omni and it was then actually a bit Phoenix that was like you know what we have a lot of traders who would love to use some sort of dollar backed coin in order to be able to arbitrage between exchanges because we're having a lot of banking problems and you can see if you read any of these things about tether that it was in fact serving a very big market and the market grew to be so big that there's something like two billion dollars worth of tether now that have been issued Wow that's it and so they're like everyone obviously is concerned about I think that the going theory is that tethers get printed to pump up the price of Bitcoin yeah kind of like I think back in 2013 well we only found out about it after the fact but this thing called Willie bought on Mount Gox that supposedly was pumping up the price and so there's always kind of a manipulation narrative in every Bitcoin run-up that we see and so I guess like what are your thoughts on that and I know that things were said at the Satoshi roundtable that seemed to be reassuring in that regard yeah so the the simple explanation for why so many tethers came about was like I said they were being used for arbitrage and basically the process is that traders on Bitfinex who want to arbitrage will sell their bitcoins for the USDT the tether and then send that to another exchange and then buy crypto on that exchange and actually you know benefit from the arbitrage when they do that they're sending tether out of Bitfinex's tether wallet to the tether wallet on another exchange so the balance and Bitfinex's tether wallet goes down and as more and more traders do that eventually the balance reaches a critical level where they're not going to be able to do any more withdrawals and so at that point Bitfinex sends actual dollars to the bank account of Tether Inc and Tether Inc then goes through a process to issue more tethers at which point they send those to Bitfinex's tether wallet and the process repeats so basically like as those tethers move out and spread through the exchange ecosystem more of them end up getting issued so that Bitfinex can continue to offer the ability to arbitrage with them what raises my eyebrow though is that the exchange who which is receiving the UST is then it seems as though it's not going to Tether Inc to get US dollars they seem to just be holding on to it which seems unusual from like a risk management perspective yeah so it was asked you know like what is your normal level of you know tethers that you want to hold and you know why why is it just growing and growing and actually they were saying you know after the recent correction that they were actually expecting to decommission to you know deissue a fair number of tethers because a lot of that demand has dropped off for now so hopefully we will see it fluctuating some as the markets fluctuate as well yeah and then the other aspect of it is just like the regulatory like KYC AML aspect of it where you know this this would make sense to me if it was an actual bank issuing this although even there like they they would need to know who the tethers are being sent to right and if tethers they use the same kind of blockchain the Bitcoin does right so it's it's pseudonymous and I I don't I don't want to drag on this conversation for too long but that's that was on my mind as well yeah that question came up as well and I think the response was that well there's AML KYC coming in and going out and no one controls what's going on in once you're in just inside the system itself but yeah I mean that makes sense from a technical standpoint from a regulatory and financial compliance standpoint I have no idea yeah I am NOT a lawyer it's interesting well okay so what would you say other interesting conversations you had at this Toshi roundtable really caught your attention I know there's been a couple of write-ups now about it but I'd be curious to hear your thoughts um let's see which ones can I actually talk about though um so we there was a lot of talk about security especially physical security and you know I got to hear some specific examples from some of the more prominent people in the space and stuff that they do and you know some of them have you know armed guards and armored vehicles and you know really go all out because they're in a position where a they can afford to spend the money on that type of thing but B it makes sense for them to do that this is especially the the people who live in countries that have you know higher rates of crime so you know it's nice in America I think we we don't have as many you know kidnappings and extortions and like organized crime of that nature so I'm hopeful that I don't have to worry about that at least anytime soon but you know as this whole ecosystem continues to grow if there's there are more like orders of magnitude price appreciation then that's just going to make people bigger and bigger targets and the the more interesting thing that I think we talked about was the difference between like the traditionally wealthy people and the crypto wealthy people because you don't really hear about you know billionaires or folks on the Forbes top 500 you don't hear about them you know getting kidnaps and extortions and other types of stuff like this and I think the main reason for that is that their assets are not liquid and you can't really steal large percentages of their wealth because they're just tied up in all types of different things that can't really be stolen so I think that we should actually learn a lesson from that and figure out you know how to make our own crypto wealth almost impossible to be stolen you know even under duress you know if a gun is pointed at your head you should not be able to transfer 100% of your assets or even a significant portion of your assets right reducing the liquidity of your bitcoins suddenly becomes an important aspect which generally people think of it in the reverse yeah that is troublesome and I think that we are fortunate here in the US and at the same time we can't be too lackadaisical in that regard thankfully I'm still crypto poor so I don't have but the problem is that someone doesn't even necessarily know that right like they just see oh Pierre comments on Bitcoin a lot he must be like Brock Pierce I'm like no I'm nowhere near Brock Pierce okay have you tried tweeting out that you're not a crypto millionaire yeah I had too much student loan debt to pay off to be in that league but they did in their minds it's not really it's not really relevant to the crime right and these people might not even be necessarily the most intelligent people around and then the other thing too though I I try to make this point as often as possible because I saw it firsthand when I was working at BitPay and now this is sensitive information but it's public too because BitPay decided to sue their insurer and so all of this information became public but social engineering is a lot less risky than actually physically physically engineering with a wrench or a rubber hose I think is the the common trope because at the end of the day if you can be sitting you know in your basement in Eastern Europe sending out phishing emails and trying to get people to click on things and at BitPay specifically you know they lost a couple million dollars worth of bitcoins in 2014 which today obviously is orders magnitude more and likewise so that's kind of like the the easy like social engineering phishing attack but then there's I think there's more sophisticated ones which I broadly call scams where basically someone is a sold on an ICO or something else like that and they voluntarily part with their money without ever having a gun put to their head instead they have a dollar sign put in front of them you know it's it's kind of the carrot and stick in in the nefarious way that's a it's the angler fish I think but yeah I mean a couple thoughts on that that's why you know you can't just protect your your crypto assets and your private keys you have to protect your phone you have to protect your email you have to put two-factor hardware authentication on every online account that is associated with your identity that supports it and especially from the phone standpoint at least in America apparently there's something stupid about America that makes it so that it's easy for someone to just walk into a phone store and claim to be you with a few pieces of information and and take over your phone numbers steal your identity and and now you know be able to use like 2FA reset functionality to get into a lot of your accounts so I actually recommend Google Fi for phone service because they literally have no stores that anyone can walk into to reset that they require a pin in order to do any phone porting so I had that exact thing happen to me with Verizon where my phone got ported thankfully like as you said I had everything on Google authenticator so there wasn't really any SMS things for them to get to and what was really intriguing was the conversation I had a couple of months later with the guy at Verizon who was like the forensic investigator into it because you know you were saying you know someone walked into a store and that was the initial story I got from Verizon when when I called them about this and the investigator was like that is not what happened and probably just called the phone number right well he's he didn't say it directly but it seemed to imply that this was an inside job by a Verizon employee who had a list and was just working the list and you know that's the other concern is that even with Google Fi now I'm sure like this is much lower risk at Google Fi than anywhere else and my my concern were not my concern my issue with Google Fi is that I'm an ample Apple fanboy and they only have Android so that disqualifies me from being more secure but yeah the inside job issue is real and it's kind of like that's part of crypto anarchy that we live in yeah and I guess going back to the scams and ICOs and tokens and whatnot I think one of the more fascinating things of this seeing how it all plays out is and obviously there's a lot of easy money in the system mainly because the the hodlers who have been around who have more money than they thought they ever would are now looking to you know get on the next thing and then also just mainstream media coverage has brought in new entrants who want to get in on the next thing that has resulted in these you know crazy valuations and people throwing money at projects that there's like there's no sane reason that you should entrust the people behind them with that much money and I think that what we're gonna see happen is a starvation due to lack of of technical skills where another thing that I talked about in my Satoshi roundtable post was the general lack of developer supply in the crypto ecosystem and so as a result like all this money has come into the crypto ecosystem and there's a lag time for developers to get up to speed so as a result we've got the you know salaries and consulting rates and everything for experienced developers basically in a bubble right now and that will eventually correct but not until enough supply of talented developers comes into the space and I think that the various ICOs and projects that are nothing more than a white paper but have tens of millions of dollars are gonna find that tens of millions of dollars are not helpful when there's literally no one with the skills available to give the money to to get them to build what you want and I think that there's an additional like piece to this which is that there's a difference between someone who's working with an existing blockchain and who is you know whether it's maintaining a node or writing implementing a node and really just trying to work with existing networks and then someone who's being tasked with architecting and building an entirely new network which was sold in a white paper and it doesn't exist yet and so that that level of developer talent I would argue like so you've got your like standard web developer and then you've got your blockchain developer who you know they could contribute to Bitcoin core and then you've got someone 10x that who could build Bitcoin core from scratch without reference to anything else right like and so that person's basically on the level of Satoshi if if what they're trying to accomplish is that innovative which or you know that that seems far-fetched to me that any more than like six of those people exist within the foreseeable future but maybe maybe I'm too skeptical about that yeah I mean I think it takes many years of experience building these types of protocols to get a knack for it and be able to do it in a conservative way where you're not opening up any exploits and of course it's it's not a single developer either like every developer is gonna make mistakes you have to have a team of developers that are watching each other's backs and preventing them from you know getting the mistakes any further than the initial code commit you know preventing that from getting merged and deployed into the production software because once it gets out there in the wild you can't take it back so yeah those those level of protocol developers are really far in a few between and have probably been in the crypto space for so long that they don't need a salary and so you you're not gonna be able to convince them to work for you probably for any level of money it's it's gonna be more ideological you have to capture their interest right and so for these ICOs that are like healthcare records on the blockchain I mean good luck finding a cypherpunk who's ideologically motivated to create a system that it you know advertises itself as being decentralized but at the end of the day it's probably unlikely to be and is it would probably be an anti pattern if it were so it's really like dentocoin like I don't know who's excited about improving the world of dentistry but they got to find a developer and pay them to at least put up some effort some level of effort and make it appear as though they're trying to deliver on on these promises that they sold investors on but yeah I I think that your your your argument about this being death by starvation is probably accurate if they even get to that right like if they even intended to ever hire anyone rather than just walking away and you know taking the money with them which I think is it's got to be at least a third of the ICOs the intent was just marketing and then we ghost yeah you know or pump and dump scams or who knows what but you know I can't even keep up with them anymore so I'm focusing on on Bitcoin and you know trying to build on it as a platform rather than trying to make the crypto universe even more complicated so I are you are do you have some lap ideas in mind that what kind of side projects do you have in the pipeline side projects I mean the last project that I did was the Bitcoin core config generator and that was mainly because I was keeping track of how many different configuration options there are and it keeps going up and up and I kept having to answer questions for people of like how do I set up a node that can run on a Raspberry Pi or they can run on tour or whatever I do have another project that I started and has language languished for many months and that actually is a throwback it combines some of my earlier skills with cloud computing and MapReduce job writing where I would like to create a MapReduce framework where you can throw the raw block data files into this MapReduce job and have it basically build the entire UTXO set and build a fully indexed you know Bitcoin database because that's what I've been working on for the past few years at BitGo is you know indexing services and and those there are ways to make it multi-threaded but you still basically have to start from the beginning and go the way to the end so that you're tracking all of the UTXOs in and out but I believe that it would be possible to create a multi-phase MapReduce job where with some of my initial testing that I was doing a few months ago I think that you could get it to a point where you could build an entire indexed perspective of the Bitcoin blockchain in less than 20 minutes and you could just throw more mappers and reducers at the job as the blockchain gets larger and larger and so you know we're at the point now where the blockchain is around like 170 gigs and so it's still manageable but if we're looking at this from a very long-term perspective then we are going to want to have ways to destroy and rebuild you know fully indexed perspectives of the blockchain in a much more rapid pace Wow so okay so let's let's dig into that because you know currently Bitcoin Core you there's a flag for a transaction index right to have that index be built and so essentially for our audience out there what that entails is that you can then query Bitcoin Core for a specific transaction ID and it will bring back the data for that and are you what you're talking about is an address index yes I'm talking about slicing and dicing it up in as many different ways as possible you know UTXO age you know the date fields size attributes fees like every aspect of a transaction that you can figure out programatically and then have that indexed and easily queryable okay awesome and so this but this is a right now you've got a proof of concept for that and then would you be able to create a like I know Satoshi uses Grafana but what kind of interface then would you be able to use for the UI yeah well I mean if you were doing an interface it would probably be similar to any block explorer but I think that this type of project would be more interesting to anyone who runs a large scale wallet or other service or block explorer like really anyone who needs to have the ability to query for Bitcoin data by things that are more interesting than just the hash of a block or a transaction and that happens a lot if you're running a wallet so for example at Bicco we have a lot of different indexes on UTXOs and that's because we have fairly complex UTXO selection algorithms and they're looking at a number of different things this reminds me of the block side project which I think was built by the guys out of Princeton and they they presented on this at scaling Bitcoin but basically doing that with the Python library called pandas to allow scientists to dig into the blockchain and query things differently so this sounds like a and I think they're using a graph database for that if I'm correct so so yeah my my my my I guess idea is even more on the like the first stage of that where then you could plug and play like what type of database you want to actually put the data into the hard part is just ingesting the the data quickly and then constructing all the transactions and UTXOs and then once you get to that like current state of the Bitcoin system then you can say okay I want to store this in a graph database or my sequel or Mongo or H base or you know whatever floats your boat for your particular application and I okay so now just when you were describing at the beginning of the episode that you were able to get your node to sync in what was that like 20 minutes not quite that fast 152 minutes of okay sorry yeah so two hours let's call it two and a half hours it couldn't it also during those two and a half hours be making calls to a database and dumping the data as it's working through it and doing things in real time like that or would that slow it down way too much yeah well I mean there's there's two different ways I guess to think about it of like would you want the node itself to be throwing data out somewhere or do you want something else to be ingesting the data and so all of the indexer services that we have at BitGo and we have a couple different versions that we've written over the years they're all you know listening to nodes and and querying them for data and pulling data out of them though you can you know set up a node where every time a block or a transaction comes in it you know makes a call out to some other script so you know you can trigger or you can listen and ingest either way like that's that's still a fairly linear progression though where Bitcoin core will go out and and grab blocks asynchronously and sort of cue them up but it still has to go through each block and each transaction in order in order to actually make sure that all of the rules have been followed so I don't think that Bitcoin core or any of the nodes like at the that particular level would would ever really be fully parallelized it just becomes a very different problem gotcha because so the reason you can do out of order is because you're not doing validation at the same time you're you're assuming valid okay very interesting and so you're operating a wide variety of nodes at big go how many different altcoins are you guys supporting is that public let's see so Bitcoin the only forks that we have are be cash and be gold we we do have litecoin which was easy and then ripple and then aetherium and various ERC 20 tokens and the bulk of my time recently has actually been spent babysitting the aetherium nodes they have been just stalling and and falling over and having a lot of operational issues and I think I wrote about that recently some of the infrastructure issues there I've also had a number of issues with ripple nose like just from a performance standpoint but both ripple and aetherium on an operational standpoint tend to have issues with disk IO and you know you could say well that's because they're getting popular right so there there is a good argument to be made for having these like global limits on the networks because at the end of the day it turns into like real resource consumption for someone who is running the nodes and it just seems from my experience that amongst all of these really popular networks that are having scaling issues the only one that isn't you know having suffered from like performance degradation at the node level is Bitcoin so well let's let's get political then I mean wouldn't it be the case that everyone can just run an SPV wallet and we only need very few services that will have well a twenty five hundred dollar piece of hardware with a solid-state drive that's cutting edge and we'll just connect our SPV well I mean that'll that'll get you you know maybe 10x growth but I do have an article I wrote last year where I did a lot of the math and let's see that article was called like can SPV scale Bitcoin to a billion users and what I was basically showing there is that one the the math that most of people gloss over when they're saying that SPV can scale is they're only doing the math from the perspective of the SPV client itself and yes those are highly efficient you know they've used on the order of you know kilobytes of bandwidth and and you know hundreds of CPU cycles which is basically nothing you know in order to verify a transaction but if you switch around and you look at it from the nodes perspective that has to serve those SPV requests it's actually having to go through and read all of the data in the blocks in the blockchain that the SPV clients are requesting and saying hey do I have any transactions here do I have any transactions here do I have any transactions here and so it's a very you know asymmetric performance where it scales very well on the SPV client end it scales very terribly on the node slash server end gotcha and so the the point there then is that either you need to have web wallets where you've got a trusted third party right or that there would be another way to scale where you just sidestep that issue and you just have coinbase run one node and you rely on that but that doesn't what would be your problem with that solution yeah I mean it's pretty easy to you know scale up web applications we've we've had you know decades of experience doing that and there's a lot of technologies that are built for that but you're kind of you know you're regressing the model once again where like we were talking about earlier the centralized exchanges are creating systemic risk both from a security standpoint and just from a usability standpoint of you know everybody going trying to get through you know one little bottleneck that you've created so we want to see these systems be as dispersed as possible we also want people to be able to get the best security model that Bitcoin can offer if they want it sure I mean the vast majority of people are not going to run full nodes and may not even ever know what a full node is if we especially if we go mainstream I mean they may not even be understanding that they're using Bitcoin in certain cases it'll it'll be shoved so far down under the hood but we want the the technology to be accessible to as many people as possible and give them the option to run a node and to have the best security model if that's really what interests them right and the security model in terms of verifying transactions coming in that that they are valid per the consensus rules that you are a subscriber to right essentially and so in that regard if Coinbase was the one running their own node they get to decide what Bitcoin is for you on your behalf with that you really ever having a say in it whereas if you're running your own full node you are a member of the consensus that you want to be a member of which hopefully is called Bitcoin but I have nothing against people who want to call it be cash or bit gold or whatever so I wanted to jump back actually as my last question you mentioned that you're running these different nodes and you have to run a node for each erc20 token no actually we don't and we we managed to set up our indexer for aetherium in such a way where it's it's pretty efficient on the indexer side of things where we were ingesting every transaction that comes into an aetherium block but we're just looking for fingerprints basically we're looking for fingerprints of bit goes multi-sig smart contract we're looking for fingerprints of erc20 tokens and then if we see something that looks like it might be interesting it might belong to a bit go wallet then we're actually making various API calls to our own service to say hey is this actually a bit go wallet but from the token standpoint we have the erc20 functionality built into the bit go multi-sig smart contract so it's it's all like technically in the same wallet you know same private keys it's just using different functionality within the aetherium protocol so as a result we use the same indexer and we just you know set a new balance that says you know this token balances X got it and is it the case that like it's having all of these erc20 tokens inside of aetherium is that contributing to the performance issues for the aetherium node definitely just from the standpoint that the aetherium network is a lot busier now and you know I don't I don't know the protocol quite at the level to say like what the performance of you know running and validating an erc20 smart contract is versus other stuff though my understanding is that like our multi-sig smart contract in general is pretty heavyweight and we are looking at a few ways to to try to make it you know less pricey from gas terms you know in order to execute various functionality within it but the the nodes themselves they've just been having a lot of performance issues and part of that you know is because we are running like full archive nodes we're not doing the like warp sync or anything else like that you know we want to make sure that we're doing full validation and and also it's just that the aetherium nodes they're updating so much state you know blocks are coming in every 17 seconds or so in in these full blocks have you know a hundred plus transactions and they're they're just doing so much reads and writes from the the disk that you know spinning disk doesn't really cut it anymore you need to have preferably solid-state disk so that you have enough IOPS on it and it becomes kind of ironically a problem because IOPS are cheap at a consumer level where you can just buy a solid-state disk and get a lot of IOPS but at the the virtual private server level which is what a lot of enterprises are going to be using their you know cloud computing disk IOPS are insanely expensive and I'm not entirely sure why that is but you know as a result we end up spending like thousands of dollars a month to run an aetherium node that is performant on like Amazon Web Services for example yeah maybe they price it like a restaurant prices alcohol you know it's like they're where they make their money but they make everything else look a little less expensive so yeah you're gonna have to be running your own bare metal if you want to do that be an aetherium full node in a cost-effective manner all right this has been a lot of fun Lopp thanks for coming on where can people reach you where can people find you on the internet well I'm most prolific on Twitter and my handle is Lopp L-O-P-P and once you get there you can easily get to my other various properties but I have a website at Lopp.net where I have all of the content I've ever created videos interviews presentations and also educational resources of basically everything that I think is interesting and worth reading within the ecosystem yeah so if you go to Lopp.net slash Bitcoin I think it is and you are you might be a developer or honestly you might even just you might not even be a developer yet but you want to learn how to program go on Codecademy learn some programming languages go to Lopp.net slash Bitcoin start reading as much as you can because there is a real developer shortage out there so I would highly encourage people to start getting involved and you got to start somewhere granted you know the the best time to start is yesterday second best time is today by by no means is it too late I think that people might have like this misconception that it's quote-unquote too late like crypto is gonna continue growing the the fundamentals haven't changed even if the price is down 60% over the past month or whatever it may be it's still a very strong wave that's occurring so anyway thanks Jameson it was a lot of fun and I hope to have you back on at some point great thanks for having me chairman Giancarlo thank you chairman Crapo ranking member Brown and distinguished members of the committee I've submitted a written statement for the record that details the CFTC's work and authority over virtual currencies but with your permission I'd like to get begin briefly with a slightly different perspective and that is as a dad I'm the father of three college-aged children a senior a junior and a freshman during their high school years we tried to interest them in financial markets my wife and I set up small brokerage accounts with a few hundred dollars that they could use to buy stocks yet other than my youngest son who owned shares in a video game company we haven't been able to pique their interest in the stock market I guess they're not much different than most kids their age well something changed in the last year suddenly they were all talking about Bitcoin they were asking me what I thought and should they buy it one of their older cousins who owns Bitcoin was telling them about it and they got all excited and I imagine that maybe members of this committee may have some had some similar experiences in your own families of late it strikes me that we owe it to this new generation to respect their enthusiasm about virtual currencies with a thoughtful and balanced response not a dismissive one thank you chairman Giancarlo and I'll begin the questioning first I will say I've had those dinner conversations with my own children and you're right this is an incredibly interesting but growing new area of financial challenge particularly among our least my children and yours