WBD033 Audio Transcription
Will Bitcoin or Ethereum Become a Global Digital Reserve Asset with Multicoin's Kyle Samani
Interview Date: Wednesday 5th September, 2018
In Kyle’s second appearance on the podcast, we discuss his approach to research, his changing view on Bitcoin and proof of work v proof of stake.
“Bitcoin is not going to zero, I don’t even think Bitcoin is going to be irrelevant, I do think in 10 years time there will be at least one asset worth materially more money than Bitcoin.”
— Kyle Samani
Peter McCormack: Hi there. Good to see you again,
Kyle Samani: Peter. Good to be back on the show. It’s been, what, four or five months?
Peter McCormack: And a lot’s happened for multi coincidence then. You’ve now raised what, $75 million?
Kyle Samani: At the last time we publicly disclosed it was at about $75m.
Peter McCormack: Excellent. And obviously a lot’s happened in the market and and that’s given me a lot of questions today. The first thing I wanted to start on is just a minor observation about how I think you work and just tell me if I’ve got this right. I think you’ve got very thick skin and I think you put out as much information out there in terms of thoughts, but also questions with the ultimate goal of finding out where there is value. Am I right?
Kyle Samani: Yeah. You nailed it on the head. Either I ask questions or I post controversial opinions. There’s a famous or funny kind of law. It’s called Cunningham law. And it basically states that on the Internet, the best way to get a question answered is not to post the question, but post the wrong answer. I frequently tend to post the wrong answer and can informed from this quite quickly. I believe very strongly in the power of these kinds of feedback cycles, and leveraging the power of the crowd. I think more than any of my direct peers, I’m kind of used that, again, with being a little bit divisive and with asking very open questions and we actually very much incorporate that into our thinking at Multicoin. it’s fairly frequent that investment committee will pull up a twitter thread that has a bunch of responses and a bunch of debates and we’ll actually talk through the merits of both sides of the argument, and I think the fact that I can do that basically at a whim is actually an asymmetric advantage.
Peter McCormack: Another thing I’ve noticed is there’s a slight change in your view on bitcoin. So let’s go back. Let’s go back to your post 25th of July. “Tl;dr. I’m bought in kind of.” So you give me the background and then I’ll start throwing you some of my thoughts.
Kyle Samani: So it ended up a priority that summer to really spend more time reading and trying to empathise with the Bitcoin maximalists and understanding the literature they’ve read it and really try to kind of grow myself, in that culture, that ideology. And after spending a few months doing it, I basically come around to the view that I think the probability is higher that they are right than I had previously assigned a probability. Whereas before I’d consider the possibility a near zero. Today I signed the probability, let’s say 10 or 15 percent. And so because of that, yeah my views on Bitcoin have changed. I still am generally have the belief that the store of value hypothesis and that the prioritisation of only focusing on just being quote unquote hard money at the expense of all other forms of utility or usability or usage I still think is incorrect or I think correct. I think it’s unlikely to produce the global digital reserve asset in the long run. But I have come around more to the view that path could in fact be the path that takes.
Peter McCormack: Can I guess the three books you read were The Demonetisation of Money, Debt the First 5,000 years and The Bitcoin Standard.
Kyle Samani: Yep.
Peter McCormack: So if you’ve view has slightly changed, does this mean your investment thesis is changing and are you now hedging with bitcoin.
Kyle Samani: We are long bitcoin today and we expect we will be long Bitcoin for the foreseeable future. Relative weightings obviously matter here and today, bitcoin is a material position and it has been, it’s not like a 50 percent condition or anything of that magnitude, but it is, it is a material possession in our fund today. We have not like over the last quarter or so, we have not increased our bitcoin position as a result of internal discussions about this particular issue. I tended to lean more extreme against bitcoin inside the firm than many others and I’ve kind of come around more centralistic view if you will. But has not impacted portfolio construction in the meaningful capacity.
Peter McCormack: Okay. And then within that thread you also said Bitcoin maximalism is subjectively stupid. Will you be in a provocative on purpose?
Kyle Samani: Always, of course. I mean the fact that over such a long period of time I tend to play games with them as well as ultimately caused me to go deep into bitcoin. Will I end up actually ultimately with a more nuanced understanding of a bitcoin maximalist view? And also, I believe I have a more nuanced understanding of, of why I think it is unlikely to produce the global digital reserve asset, but I do understand more completely why they think the way they think.
Peter McCormack: Yeah. So I found that kind of interesting. I’ve been on a similar path, probably not as deeply researching as you, but struggling still sometimes to maybe find value in a number of the projects that are out there. It kind of feels like the fat protocol thesis has mainly been debunked recently. I know you’re not completely against it, but there’s been some criticism of it and there’s a been a struggle to find value in a number of the projects. Not only because the projects themselves are struggling for users but also the ability to release capital by investors. So I’ve kind of gone a similar path to you, but I think we should explore this. So I’ve taken some of your notes from that Tweetstorm and you made have three or four interesting points. The first one was it is very likely that after the 2024 having that this creates a problem for a layer two. So can you explain what you meant by that?
Kyle Samani: So basically it’s actually been proven that with bitcoin it is game theoretically optimal, that whenever a minor finds a block that they should immediately propagate that block to as many people as possible, as fast as they can to maximise the probability that their block is ultimately included in the canonical chain. Giving the way that Bitcoin finality works probabilistically, that today is game theoretically correct. The reason that works is because you can look at minor compensation, right? There’s two forms of minor compensation, one is inflation, and the other is transaction fees. And today something like 90 to 95 percent of a miners revenue is inflation, and I think even when bitcoin fees were at their highest, like in the December timeframe, I think transaction fees were 40 percent or something of miner revenue, but today it represents 90 percent plus of revenue, maybe even more.
Kyle Samani: And so when you consider this, right as inflation decreases, excuse me, is the rate of inflation, you know, decreases in the system, there’s literally just less money left to reward the miners. As the ratio of fees paid, the miners for inflation versus transaction fees is that ratio flips it becomes game theoretically possible to not propagate blocks immediately and when he start compounding that across multiple miners and multiple blocks, and then if you’re relying on blocks for things like entering into a lightening agreement, right? The HTLC for lightening that starts to create very strange behaviour. It just makes the system overall, I don’t want to say the word brittle but it, but certainly it makes it a little more flimsy. And so that can create a little number of strange compounding problems.
Peter McCormack: And you also said that there’s lots of reasons why layer two technology might not be adopted.
Kyle Samani: With Layer 2 technologies I’ve come to the conclusion that they will be adopted at least to some capacity. It’s still very much tbd. The number of people that need to really buy into lightning and agree to use lightning is actually relatively small for lightning to work. But conversely, if they don’t buy into lightning and then I think it’s unlikely to work in a meaningful scale. In those hubs, very obviously are the exchanges. So Coinbase, Binance, Bitfinex, Bittrex, Poloniex, etc. Like these are the people that need to do this, that fiat on ramp to crypto. These people sit on the exchanges, with so much Bitcoin actively, and if they choose not to do lightning for any reason with the other exchanges, I think it’s unlikely that you’re going to see kind of meaningful adoption. I think the idea of Liquidity providers emerging who are just liquidity providers that are not attached to existing fiat on ramps, I just find that to be extremely unlikely. I don’t see where all that capital will accumulate from, again, at a meaningful scale, right? I’m talking $10 million, $15 million worth $50 million, like real amounts of money, so if those people don’t sign on then I don’t see how the lightning vision really works in a meaningful sense.
Peter McCormack: And then you said there’s no fundamental logic that the global digital reserve asset must be a store of value first before medium of exchange. And then you said at ETH or EOS current scale, they are enough of a store value to power meaningful commerce. Does the order really matter?
Kyle Samani: I think my issue is more with the, the bitcoin maximalism view that it must become a global store of value before anyone would ever use it for any commerce. And I think that is materially debunked today. We have a lot of empirical evidence that that is false today. Even if you hold the EOS today, you likely to believe EOS, for example, will be worth much more in the future. If you hold the Ethereum today, you’d like to believe Ethereum will be worth much more in the future. Despite that people continue to transact, do commerce, engage in betting markets on auger, whatever, all kinds of different things. And I’m like at the level of scale, whether it EOS is worth $5 billion and Ethereum today is worth, let’s say $30 billion, that is not global store of value scale, that’s multiple orders of magnitude away, but it is still being used to power meaningful forms of commerce.
Kyle Samani: And so, what I take issue with is the idea that it has to be a store of value, you know until something scales to $5 trillion, and at that point they can start to be used as a medium of exchange. We have quite a bit of empirical evidence that that’s just not correct. And moreover, I think there’s real merit to the idea that money has multiple dimensions, there’s obviously the hard moneyness of the Bitcoin and digital gold and there’s obviously very rational libertarian Austrian appeals to that. But there’s actually much more, money acts in many ways, like just yesterday we were talking about this in our investment committee, one of our employees said his father in law, he asked him what his money is, that money is whatever he can go to the grocery store and buy groceries with.
Kyle Samani: Right? So to him money is the ability to send it, and the other person who to accept it, and so you look at all the kind of financial primitives being built with things like Maker, Auger and Dharma and these other kinds of tools and it seems pretty clear to me that it’s very myopic to interpret money as just fixed supply, scarcity, can’t hack it, censorship resistant. All of those are all certainly strong benefits of a money. I think it’s rather myopic. Basically those four to five traits are the only ones that matter and everything else is stupid. It seems pretty clear to me that in order for money to be money, it needs to use this money. I think the idea of, you know, get to $5 trillion and then people will spend it. I think that’s incorrect.
Peter McCormack: And then the last point was that you find it quite likely that proof of stake will deliver stronger security guarantees at scale, why is that?
Kyle Samani: This is probably contrarian within crypto and aggregate? But I’ve spent enough time on proof of stake systems understanding them, and I’ve developed some pretty coherent mental models to compare the two. The key thing with proof of work systems is that the system is a function of inflation times the market cap, so $100 billion, and let’s say the inflation is one percent. Then you can actually quantify the security of the network as basically a billion dollars, which is that rational market actors will spend up to a billion dollars acting honestly, to secure the network because it will generate some profits accordingly.
Kyle Samani: And so when you look at proof of stake systems, conversely, let’s say it’s $100 billion network with one percent inflation. With that model, buying up enough of the asset in question such that you could manipulate human history and basically execute censorship resistance or saturated attacks or 50% attacks, or some other weird forms of selfish mining in order to do that is very likely to cost you substantially more than a billion dollars because you would have to buy up that much more Ether and have that much more capital at risk.
Kyle Samani: I’m just going to give it a kind of market dynamics that is in border books etc. Moreover, if you were going to do that kind of an attack, with proof of work, it’s impossible for you to be forked out of the system. With proof of stake systems, you can actually get forked out if the network can detect an aggregate, if you’re acting dishonestly, even if you are the majority of the network. If the minority of the network and can detect your acting dishonestly, you can literally get forked out, which would be catastrophic. You would have zero return at that point. You’d lose all of your capital at stake. And Vitalik has talked publicly about auto enabling this by default in Ethereum and it’s very controversial view, but I’m inclined to agree with, with Vitalik on this issue, because if you set that threat that if the network and detect you doing this at scale and you could get forked out and get forked out to a hard zero, I think that risk reward ratio of attacking a proof of stake system, that risk tolerance becomes so high because you have a chance of just getting hit with a zero, and instantly.
Kyle Samani: With proof of work that’s impossible. If you have mining equipment, no one could stop you from mining and you can always stay mining, and then literally no one can take that away from you.
Peter McCormack: Totally true but is it because, for example, with Siacoin, they were going to change the algorithm to fork out Bitmain’s miners. So it’s not strictly true. You can’t on an individual level, it’s on an equipment level, right?
Kyle Samani: So for ASIC miners that is more true than not, but lIke specifically in the case of Bitcoin and like maximum security, I find it extremely improbable that the Bitcoin is ever going to fork to change the hashing algorithm because the security of the bitcoin network, as the largest coin is explicitly a function of the fact that it has so much hash power. Like if you were to fork away from that, you would literally be throwing away all existing hash power, and I just, I find that highly improbable.
Peter McCormack: With proof of work, why would you want to fork out a number of the miners? It’s a 51 percent consensus rule, right? Like either the network’s under attack or it isn’t. So what reason would you have to fork out a percentage of miners? What could they do dishonesty?
Kyle Samani: I mean, they can just change your transactions, right? They could do block reorgs. There’s all kinds of weird forms of selfish mining that they can take or do that would be suboptimal for the network as a whole, even outright censor transactions. Like if you had 55 percent of the network and if you really wanted to screw with it you could literally…
Peter McCormack: But outside of having 51 percent of the network, what can anyone actually do?
Kyle Samani: I mean, like I’m not getting into all the nuances of like all the other interesting forms of selfish mining, but you can censor people. You can do all kinds of other interesting ways to tweak the system in your favour, but more importantly, if you have that level of capacity in the network, no one else can get rid of you. There Is no way to detect which percentage of the hash power is coming from the attacker and prevent them from mining, like that is actually impossible in proof of work systems as they are designed today. In proof of stake system, if you can detect the person you can fork them out.
Peter McCormack: So how would a decision be made to fork someone out or the proof of stake system?
Kyle Samani: So if you can detect, so let’s say there’s 100 validating nodes in a committee, and let’s say 48 of them come to an agreement that the other 52 are sort of a cartel, some sort of organisation, and they are systematically censoring transactions for any reason or for no reason at all. It doesn’t really matter as long as they can detect some pattern of censorship or some other form of behaviour. Then those 48 can literally fork out, because they know who the 52 nodes are that have the coins staked to be those 52 and they’re going to slash them to zero and for off and just go in a new direction. And then those 52 have no weight anymore.
Peter McCormack: Is that a fork in that there would then be two coins?
Kyle Samani: Yes, there would be two coins to this point.
Peter McCormack: So that doesn’t sound an optimal solution then because you run the risk of having a disagreement on the coins and having a similar situation to say Ethereum and Ethereum Classic.
Kyle Samani: Yes, correct. I mean this would not be trivial, this is the kind of thing you don’t exercise unless you have proof that this is happening, the way that proof would happen, would be if there are users all over the world who are submitting transactions to the network and let’s say those 48 in the minority systematically see that they all receiving this set of transactions and those transactions are not being included in blocks produced by the other 52. And if you can see that pattern systematically over an extended period of time, right? You can actually prove this. It’s in an objective way to the rest of the world. And so yeah, of course there’s going to be like divisiveness and fighting and like that’s never going to stop. You can actually produce a proof of malevolence and cite all the reasons and demonstrate that publicly.
Kyle Samani: And so all the little always be contentious, you can actually still pull it off in my view in a much easier way than say the DAO fork, which was a function of values. The other nice thing here is defaults in social consensus. So if everyone opts into use the Ethereum network in which default fork in the event of 51 percent censorship and let’s say that behaviour is a opt in default, right? Well then everyone has already agreed by definition that we’re all agreeing to do that. So in the event of that fork happening, it’s like, okay, well we already agreed that we would act in exactly this way and then the thing happened and now we’re acting in a way we said we would. So it’s actually very logically consistent.
Peter McCormack: So from my side, look I’m not the most technical I’ve told you that before, but I always try and do little experiments just to test myself. So I mine, I’ve got 70 S9s and 90 Dragonmints and a Sia miner. Every month I’m having to balance my power cost and against my mining rewards. And at the moment, the last few months I’ve been minding at a loss Now I’m moving to a new centre, which is going to take me back to just below break even, but then I also stake with Zcoin, just out of interest, and with Zcoin I don’t have to do anything. I’m staking and every 15 days I get 15 more Zcoin sent to me.
Peter McCormack: But with Bitcoin, with the proof of work, I’m, I’m actually having to consider the economics the whole time and, and that to me makes it, it feels like a much more secure way of a securing the protocol because the proof of work is in, is in the name, right? You have a cost risk associated of not following the rules. Whereas I don’t really have a cost risk was Zcoin. Therefore I’m struggling to see and other people are struggling to see why a proof of stake is preferred over proof of work.
Kyle Samani: The flip side of this is honest and dishonest behaviour. So if you’re a mining honestly, the fact that you are mining honestly is actually purely wasteful, right? And that’s reflective in the fact that you’re actually losing money as a miner even though you’re mining honestly. You’re trying to do any shenanigans and you’re following the protocol honestly. The key with proof of work is that on order to disincentives dishonest behaviour, it creates this tax, this cost, on all honest behaviour as well. The key with proof of stake is you’re not creating all of this exogenous cost in the event that honest behaviour, but in the event of dishonest behaviour, it can be detected and it can be systematically managed accordingly. So that’s kind of the fundamental, like you have to split that and kind of consider the two. I mean, there’s different forms of dishonest behaviour, but at a high level, that’s the way to think about it.
Peter McCormack: So if Jimmy song was on this call, he would disagree with you, right? That’s her.
Kyle Samani: That’s very probable.
Peter McCormack: So let me ask you a different question then. So I’ve read all your articles, I’ve read them a few times, they take a while to get my head around them, but for those things you’re invested in you are constantly looking at different ideas to improve them or different theories, different thesis and that is all very interesting. But with Bitcoin, I find your writings tend to be more about debunking Bitcoin. So let’s flip it. If you were get to make changes to Bitcoin and you wanted Bitcoin to be a successful global store of value, what would you do to improve it? Rather than debunk it?
Kyle Samani: Bitcoin maximalism is rooted in the idea that no one can change it, no one can control it. As a result of that, you cannot evolve in the status quo, all changes have to be soft forks or have to be like layer two kinds of things and have the base protocol basically cannot change. Other than, like some sort of catastrophe. My sense is that although that feels like the right story to tell, digital gold, and gold doesn’t change, as people want hard sound money. I understand the appeal to that argument, I think in the long run, if you don’t evolve, you die. And so you have this kind of fundamental value set of how frequently do you have an answer for Bitcoin is at the core layer.
Kyle Samani: The answer is not at all under any circumstances. And obviously the value set for basically everyone else’s at some frequency higher than Bitcoin. I kind of disagree with that value set on a long enough timescale. Like, I don’t know what I would do to try and make Bitcoin more successful because my ideas would be fundamentally contrary to what makes Bitcoin Bitcoin. Within that framework, working on lightning, working on interesting new forms of privacy, like Schnorr signatures, this as a soft fork. Given the framework and the bounds in which the Bitcoin community tends to operate. I think they’re generally being quite rational and the operating reasonably I just disagree with in the bounds that have been placed.
Peter McCormack: But I guess up until now there’s a strong argument to say that Bitcoin maximalists are being proven regularly, right? So Bitcoin cash as a way to scale has pretty much proved itself as a failure, people don’t want it, people haven’t adopted it, Ethereum has also had it’s scaling issues. EOS has had its issues, like plenty of other protocols and ideas have had issues yet Bitcoin, out of everything, has managed to probably retain value the most outside of like obvious, rare cases like I don’t know Binance Coin. It’s retained most value over this period of time. So maybe the theory of sound money is right?
Kyle Samani: I mean bitcoin has retained more value than most altcoins over the last downturn. I don’t think that’s in any way reflective of it’s efforts to do scaling correctly or not correctly or anything else. It’s just, it’s the brand. It’s flight to safety. Like it’s the bear market. I don’t think you can read much into that faggot. It’s the largest generally is obviously a functional functional being first and being conservative, which is reasonable. I do think the bitcoin story is by far the most coherence in the market today. I mean, I think it’s far and away has the most coherent story to tell today if I had spent time today talking to a lot of institutional investors and you don’t have. The conversations are like crypto stupid that another half hour. Okay, look, we’re really, we get bitcoin like this makes sense. What’s all the rest of the nonsense?
Kyle Samani: That’s a pretty consistent level of understanding and the reality is the Bitcoin story is easier to understand, it is fundamentally coherent. It makes sense. There are a lot of people who are politically predisposed to buy into that, who may not be the Bitcoin bugs or Bitcoin maximalists today, but just given the current political leanings once they learn about Bitcoin they are like, oh yeah, I agree with ideologically. I do find that the Bitcoin story is by far the most coherent, but I think that really no way reflects longterm opportunity to become a global digital reserve asset.
Peter McCormack: Would you bet against Bitcoin over a 10 year time frame?
Kyle Samani: On a 10 year time scale, again, Bitcoin is not going to zero. I don’t even think Bitcoin is going to be irrelevant in 10 years. I do think on a 10 years time frame, there will be at least one assets worth materially more money than Bitcoin.
Peter McCormack: Okay, interesting. And are you still hedging across multiple different protocols?
Kyle Samani: Yes, we’re long and short in quite a few positions.
Peter McCormack: Okay. So say during your research, reading the various books and going down the Bitcoin maximalist rabbit hole, if you were to come out of that, a Bitcoin maximalist, would that have presented quite a difficult decision for you within the fund?
Kyle Samani: No, not at all.
Peter McCormack: What would happen with the rest of your investments?
Kyle Samani: I mean, I run a hedge fund and I have no mandate to buy things and come back in 10 years. Like I manage capital every day. I can move it around every day. So my mandate is to simply return capita. I mean obviously going to return more capital, but the market moves short term and there’s dIfferent things that happen. So I don’t think that’s a fundamental problem. There are quite a few Bitcoin maximalist who run hedge funds. There are a number of other folks in space who either are publicly or privately are our maximalists and they run high profile funds in the space.
Peter McCormack: I guess purely on a return level, I guess it’s harder to create a return from Bitcoin than it is from these alternative projects because if we look at say the dfinity returns for early investors, there’s still a lot of opportunity for funds to make a lot more money on new projects than with Bitcoin?
Kyle Samani: I mean generally, yes. That’s why not everyone is 100 percent bitcoin. Like that by definition is true.
Peter McCormack: I also want to talk about your most recent post, Aggregation Theory, Thin Protocols, Recentralisation, the Auger Edition. I’ve read it three times now. I still don’t understand it all, but let’s talk about it. And we’re going to return back to the previous point, you said “I believe with extremely high conviction that our global digital reserve asset will be worth tens of trillions of dollars” you’ve written before about the path to trillions of dollars. So tell me what this global digital reserve asset looks like in your mind.
Kyle Samani: Today, we have I think somewhere on the order of $250 trillion of assets around the world, which are real estate, equities, debt, etc. plus cash money in mutual funds and other kinds of money substitutes. The vast majority of those assets today are equity bonds and real estate by far, those are the three largest asset classes on the planet. Today, all of those are managed using basic paper stocks, a paper stock certificates or paper, right? You know, land titles, deeds almost exclusively. I believe on a long enough time scale, 100% of that will be managed on a blockchain. The idea of having a private key to do value transfer, like to sign a message and then broadcast that message publicly, that is how value transfer will be done in the future. There is no question that will happen, that is the right way to solve the problem. and if you assume that that $250 trillion of assets will live on a number of chains, I think it’s likely ultimately to converge towards one chain and that whatever chain that is, I expect that ultimately, the chain itself that supports all of the other assets will also be the chain that’s worth tens of trillions.
Peter McCormack: And is that an aggregated value? So for example, with Ethereum, would it be the aggregate value of each and every single sub token that represents an asset? Or are you talking about he asset of the protocol itself?
Kyle Samani: I think the base chain can be worth, let’s say $50 trillion and they it can secure assets worth, let’s say $215 trillion.
Peter McCormack: So right now outside of bitcoin, who are the contenders for you in that?
Kyle Samani: So there’s a few, there’s four-ish major categories on things that could have become the global digital reserve assets, I think two kind of sort of fall under the store of value hypothesis and then two fall under the utility hypothesis. Within with the store of value hypothesis you’ve got two subcomponents. So I think those two components are, let’s call it decentralisation maximalism and let’s say privacy. So Bitcoin obviously falls under decentralisation maximalism and there’s, I think two contenders kind of starting to come out there, one is Chia, which is basically Bitcoin, but with proof of spacetime instead of proof for work and the other is a new protocol called Coda. Within the privacy bucket you’ve got ZCash, Monero, Grin, maybe Mobilecoin kind of fits in there, those seem to be of the real players.
Kyle Samani: With the utility hypothesis you’ve got what’s called a medium of exchange tokens and smart contract platforms. So medium of exchange tokens would include the likes a Bitcoin Cash, Dasg, maybe Mobilecoin fits here. And the smart contract platforms, is Ethereum, EOS, dfinity, Algorand, Solana, Tezos etc. Whatever wins will be one of those four sub buckets.
Peter McCormack: Can multiple win?
Kyle Samani: On a long enough timescale I expect you trend towards one but a long enough time scale could be 30 years, like five years from now, I expect there will be a handful, like five-ish, five to 10 that will be much larger than the rest, but I don’t think in five years we’re going have one chain that’s bigger than that, like the dominant gene.
Peter McCormack: And could there be a scenario whereby the assets are issued on this protocol, but as Bitcoin is seen as hard money, we discussed that and it has a bitcoin is trusted around the world that still becomes the base currency that people buy these assets with.
Kyle Samani: That’s the kind of the status quo today. That’s actually much less true today than it was six months ago. Today USDT pairs account for quite a big of part of it, as trading pairs. So I think generally correct today, but, but it’s trending away from being correct. I expect 24 months from now, a majority of altcoin volume will not be denominated in Bitcoin.
Peter McCormack: But you say because it’s trending towards a stablecoin?
Kyle Samani: We’re getting more exchanges or adding more non Bitcoin pairs.
Peter McCormack: So also within that article it was brought up because of Augur, something I’ve taken a look at, and you identified a number of problems with it. I think you listed them clunky, not inexpensive, it’s not a seamless web application and runs slowly, etc. Why do you think it was such a failed launch?
Kyle Samani: So if you’re going to say it’s failed, meaning that uses growth is not accelerating every day and then the reasons are basically the ones that I put in the article. I mean it’s slow, it’s clunky, like it’s hard to, pretty pretty obvious things and the reason it’s all of those is basically because the Auger app is designed to be as maximally decentralised as possible and so there’s a lot of compromises that were made to achieve that trait and as such you’ve got a pretty poor user experience.
Peter McCormack: So do you not think there will be an Auger app itself in the future? as in there won’t be a native app.
Kyle Samani: No, there will always be a native Augur app, there is no reason there won’t be. I just expect that a centralised private company built on top of the Augur protocol will deliver user experiences that are an order of magnitude better.
Kyle Samani: I mean obviously like stablecoin, right? Like whether it’s Dai, whether to TrueUSD, whether it’s something else doesn’t really matter, but obviously you need a stable asset dominated, and I expect that the Augur team, I’m sure is working on this. This will come in time and then you know, we’ll see centralised companies on top of Augur deliver everything else. So push notifications, the ability to monitor for accounts or monitor for new types of markets, categorisation, marketing, mobile support, fast loading, just like everything that this needs to be good can be delivered as a centralised service.
Peter McCormack: So do you think it needs to as an app in it just look like a traditional website or app experience? No one would know the difference.
Kyle Samani: Correct.
Peter McCormack: So why does decentralisation matter then?
Kyle Samani: Modern user experiences are all rooted in first principles, given the context in which your a consumer is thinking about solving some problem. All modern, successful apps are designed on a first principles basis to optimise for that problem. In doing a betting market today, that decentralised has all these implications for performance and infrastructure, etc. Augur is fine at let’s say the user interface level. It’s reasonable, it’s not perfect, but it’s good enough. The problem is not just the shape of the buttons on the page, their problem is the entire experience end to end of downloading it, using it, etc. And so that’s kind of the problem with Auger, which is well beyond just UI and like as all of these other problems get solved, then it work just like a regular consumer application.
Peter McCormack: Why does it need to be decentralised? What are the benefits?
Kyle Samani: For Augur, censorship resistance and the ability to discover new markets, indicates specifically a prediction market it seems quite clear to me that because these markets are today so heavily regulated by governments that they’ve actually stifled the opportunities for prediction markets generally. I think the ability for anyone to create any market and we will be able to bet on that. I am quite certain that will unlock all forms of new liquidity and new types of ways to hedge risk, etc. on a long enough time scale.
Peter McCormack: So you’re quite bullish on this being a potentially multibillion dollar industry?
Kyle Samani: Absolutely. Already trillions of dollars of value flow through prediction markets in some form today. I expect that for Auger based markets to achieve volumes certainly in the billions to trillions is still very much TBD, but billions at a minimum.
Peter McCormack: I actually find the concept of recentralisation quite interesting. I know that wasn’t what the article was about, but you’ve obviously considered that people will take centralised concepts and recentralise them for a number of reasons. One, to have to take a larger share of the value, but possibly also for technical reasons.
Kyle Samani: I think we’re going to see people flirting with it. I mean, we’ve already seen this idea. It’s played out in technology time and time again with the idea of bundling software and unbundling software with the idea, right? Like the internet was an open protocol. Then you got this netscape thing and then AOL came along and all kind of recentralised the web. All the search portals before google were trying to centralise user attention, where google said no, just search and we’re going to kick you out and go somewhere else. So, we’ve always had these kind of competing forces between centralisation and decentralisation over time and depending on the state of the market and such. Within crypto, we’re seeing the same thing happen again with EOS. Kind of an interesting experiment along that spectrum versus something like Ethereum or Bitcoin is just one simple example.
Kyle Samani: The idea of, proof of stake systems though, even more generally, the kind of base idea of tendermint, which is power and cosmos. It’s called the PDFT or practical byzantine fault tolerance. PDFT was invented in 1999 and it wasn’t very hard kind of technically to make the jump from a PDFT basically what tendermint is. This is really the first kind of reference implementation of PDFT in a proof of stake setting. But despite the fact that PDFT was amended in 1999, there was literally no proof of stake systems at all until 2013 and it took Bitcoin emerging for that whole segment of research to get restarted. And now we’re seeing this massive wave of research around proof of stake systems. So my point to you is that your question originally was are things going to recentralise and my answer is we’re going to see lots of experiments play out where different people view different layers of centralisation, different amounts of centralisation; those are all open experiments right now
Peter McCormack: Where have you had your mind changed in the last, say four or five months since we spoke and what things do you think you’ve got wrong and where’s your position changing?
Kyle Samani: So I think we had a little bit too much hubris going into EOS launch. We still maintain the same level of general, medium term prediction we held on EOS before the chain launched. We’ve actually spent quite a bit of time internally reviewing EOS and what could we have forecasted, what couldn’t we have forecasted, how could we have sized or not not sized things differently given the information we had at the time. But we’ve tried and maybe we’ve come to the conclusion that we didn’t spend enough time reviewing the history specifically of the Ethereum launch and really looking at kind of the mechanics of what happened in the six to 12 months after Ethereum months on a week by week basis, understanding the challenges that faced and try and map those over as best as possible to EOS.
Kyle Samani: And so since then, we’ve spent quite a bit of energy internally reviewing what does it really take to launch a chain and what are all the things you can kind of expect to go wrong in the early days of a chain. We can be more rigorous around risk management going into a launch of a new chain in the future. We’ve spent quite a bit of energy there in the last few months. I’d say I have become more bullish on the size of the total market for crypto. I think now that 100 trillion is doable. Whereas before I thought that was a stretch, I now 100 trillion is doable. I’ve become more bullish on zero knowledge proofs, to the point where I think that Zero Knowledge Proofs will be one of the defining technologies of my lifetime. I’m quite a bit more bullish there. We’re making two investments there right now. I don’t know when we will publicly disclose them but two material investments in that space. I’ve become a little bit more bullish on layer 2 working. I don’t think Layer 2 is the saving grace that everyone wants to be.
Peter McCormack: I know you’re going to ETH Berlin, but what else is coming up for you?
Kyle Samani: I go to ETH Berlin. I’ve got a number of fundraising events in late September. I’m going to be speaking at ESTF in early October. I’ll be a Defcon in November. I’ll be a consensus invest in November. Those are the ones that I can remember off the top of my head.
Peter McCormack: And just let everyone know how they can get hold of you.
Kyle Samani: I’m pretty accessible on twitter. My twitter name is @kylesamani.