Jake Brukhman ~ Prospecting Protocols: A Discourse on Generalized Mining Coinfund

[Music] awesome thank you for coming my name is jake Brookman I'm the founder of coin fund I'm gonna be speaking to you guys today about generalized mining and the third party economy this is sort of a little update of a talk that we did it at DEFCON four in Prague this past year and tell you a little bit about coin fund we've where a blockchain investment firm our first fund came out in July of 2015 and we've been in a space before that so we have a kind of long history in the space looking at blockchain technologies and investing in digital assets as well as companies working on blockchain tech we have a multidisciplinary team of nine today based in in Brooklyn New York we have Dmitri also NSF working part time with us and we study consensus crypto economics and financial structures around blockchain in 2018 we launched our second fund backed by the partners at Ben rock and David Pakman and we consider ourselves kind of full-stack investors in a sense that probably before many other investors we decided to invest in not just the base layers or protocols of blockchain but also the middleware and also gaps and in addition to that we invest in in kind of traditional company rails that get money into crypto and help issue assets and help to be compliant and finally and most importantly we are we like to dog food we are we like to participate in the decentralized networks where we invest in just in general and that's how we come to generalized mining a couple of days ago you may have caught got some press coverage on the launch of grass fed network which is essentially a quant wing for coin fund but unlike traditional quant which tends to write quantitative strategies on trading stocks right or even digital assets we think of this more as a vehicle that plays crypto economic games with with networks for a return and so what I'd like to do in this talk let's tell you how how we conceive of generalized mining what that means and what it is and because this is a developer focused kind of conference I'll just walk you through one of these games that we spent a lot of time on a coin fund and it was a was a great experience so generalize mining what does this mean all these-- is kind of begin with some assumptions and the Assumption behind generalized mining is that decentralized networks will proliferate there will be many of them and also that decentralized networks are kind of public goods they're publicly owned they're owned by the users and inventors of those networks and when we think about what decentralized networks were in 2015 and how they might have colored our view of what the centralized networks are well in 2015 you basically had cryptocurrencies you know block chains that implemented pure currencies like Bitcoin had a couple of asset issuance platforms like NXT counterparty and then etherium came along and really lowered the cost of creating decentralized protocols with a smart contract platform so in 2018 really 28:19 if you look at the market and kind of try to point out what decentralized networks are out there in production we really see quite a bit more than that we see networks that offer storage space or computational power we see networks that are trying to implement social media like steam it and in those networks content curation is kind of the important the important metric and how well that works more recently we've had decentralized finance come onto the scene and and basically create protocols for traditional financial services and those protocols also have a lot of you know places where third parties need to participate in those protocols for them to work and finally things like TCR Zoar registries or or ENS or handshake which is kind of trying to disrupt dns is also a use case where you need many third parties to participate for those of networks to work so in short the centralized networks if you believe that we'll see many such networks and many different verticals they need services to be provided not by a central company as is usually done in traditional technology but by actual network participants let's just kind of maybe go through how this developed over time you know in the first such network a proof-of-work cryptocurrency blockchain you generally had two kinds of participants miners these are the people who use hardware to validate transactions to make the network work and users who are sending money to each other and then the miners would get compensated in the network asset like Bitcoin then we started to think about proof of stake systems and in proof of stake systems miners became validators and then we also got this additional concept of you know for using a delegated proof of stake system we might delegate some of our stake to us to another party who will vote for those validators in our behalf but then this keeps going when you get into a decentralized social media network you now have four kinds of participants you have witnesses which are the validators of steam you have the people who are actually creating content and publishing on the steam platform you have people who are curating content meaning they're up voting and down voting certain content and the theory is that the good content should go to the top and the people who have predicted that some piece of content is good by the measure that it you know went viral or something they get rewarded and interestingly this is where we start to see content curators being bots as well the steam ecosystem has a number of BOTS operating on it today we heard about scalability on the last panel and in layer two and this is a really compelling kind of case for for third-party participation you need things like hubs Watchers routers dispute resolvers and other kinds of you know liquidity providers for layer two or state channel networks to actually function how many has anyone heard of live pure life here cool for anyone who doesn't know live pure is a company based out of New York they're creating a decentralized network for video streaming and encoding and out of the gate this network has six kinds of participants you had sort of this early investors who who brought the team to fruition you know he eventually have users or customers this is a delegated proof of stake system to pick which transcoders are gonna be doing good work so this has a notion of delegation and delegate errs in the system you have the transcoders who are actually transcoding the video and getting paid for that and who are curated by the network you have Merkel miners which I'll talk at length about and basically these are third parties who help to distribute the live peer token supply and more recently in their proposal that they released a couple of months ago they're you're gonna have a six kind of participant Orchestrator this is for scalability so an Orchestrator will coordinate many different transcoders to enable linear scalability and we can go on and on right we can talk about lending protocols and liquidity provision we could talk about predict resolving prediction markets as it's done an augur we can talk about CDP liquidation and maker and what we what we realize is that there's actually a huge number of roles that are either in production today or are poised to be in production as these protocols hit the market so in short decentralize networks they need hardware to run on they have they need software to facilitate all of these functions they need public stakeholders to govern the network and a long list of domain-specific supply-side services to enable the functionality that that particular network or vertical represents and so generalized mining is this idea it's also been variously called mining to our supply-side services network service provision or even from the investor perspective it's been called active network participation is the idea that you know you have a supply-side service that you're providing to the centralized network and that decentralized network compensates you for that service on protocol in the asset of that network and the the market of all these opportunities is almost like its own economy sort of if you extrapolate it out in a world where many of these networks exists they're acting as public goods and they're actually providing services you know in place of Amazon AWS or something like that just a couple of more things to say like how do you make sense of all of these opportunities the way that I break it up in my mind is that you know on one side of the spectrum on the red side you have opportunities that are really hardware intensive so this would be like proof-of-work mining right you're you have a giant hundred million dollar data center and here you're competing on how much hardware can you can you run and how efficiently can you run it on the other side of the spectrum this is where I think you know as we'll see kind of funds will have a a better time participating is this idea of software or proprietary software strategies that you can run and these are things like again crypto economic arbitrage helping make stable coins stable or you know again liquidating makers CDP's these are opportunities that don't require a lot of proprietary hardware or Hardware scale you can run these on regular computers but the the competitive advantage of someone who runs this strategy is the fact that they're smart that they can come up with a proprietary strategy of how to perform these services better than others and somewhere in the middle is a mix of hardware and software is probably this idea of staking which we're most familiar with we have a market of maybe a hundred companies going out and doing staking provision today you know in staking essentially you take some of your asset you put it toward a network and you earn a rate of return and a token denominated way I think classical miners or people who already have data centers they're well positioned to go into the generalized mining opportunities that have Hardware provision that have a lot of computational intensity I think quants and crypto funds have an advantage in the software intensive side of this market and of course we see staking companies and and their customers kind of going into the middle of this market and one thing interesting thing to to consider is that there's a lot of opportunities here and they're all very very different and staking by far is not the the main opportunity might be the main opportunity today but it's not going to be the main opportunity tomorrow and even within staking not all staking not all staking crypt of economics have this idea of delegation so if they don't have that idea they're not a great candidate for for staking companies and so there's an opportunity for for other players to take advantage of them in the second part I'd like to take you through an actual crypto economic game that was actually put into production by by the live peer team how many people have heard of the live peer merkel mine okay so relatively few if you google live peer merkel mine it gives you a little text back on what that means but I'll I'll go through that here and essentially a life year Merkel mine is a crypto economic game instrumented by the live care team that implements a decentralized airdrop in other words the distribution of tokens to some set of aetherium addresses in particular they the way that life peer did is they they said let's take a snapshot of eligible addresses it turns out there were two point six million aetherium addresses that were eligible eligibility just meant you had maybe 0.1 ether in there or more and if you were the owner of one of these eligible addresses then you could claim about 2.4 for LPT it didn't matter if you had a lot of ether or little ether in the address your reward was always the same but obviously this is quite hard many people don't know how to use smart contracts or just simply weren't aware of it and so live peer created a crypto economic game where third parties can go and redeem that LPT on your behalf and essentially when a minor would mine your address that 2.44 or so LPT would be dropped into your address 63% of the token supply of live peer was air dropped in this fashion and the way that the game set up its incentives it said okay we're gonna do this for 15 months when this begins basically miners can go and claim LPT on behalf of other addresses and when they do you know most of it goes to the to the address but a little fee goes to the minor to economically incentivize them to participate over the course of the 15 months the fee goes from zero to a hundred percent so somewhere in the middle you know we expect that you know 50% of that 2.44 goes to the address 50% goes to the minor in the beginning all of it goes to the address and none goes to the minor and at the end all of it goes to the minor so that's what I mean by every block the minor fee goes up in order to claim the LPT a proof of work has to be done and this is what a merkel proof is is essentially a computation which is takes as an input that particular address does a little bit a proof-of-work it takes about 30 seconds to do this proof of work on unlike a regular computer and then the miner can submit this proof of work to the smart contract resulting in the in the vending of the of the LPT and the airdrop so why would people want to do this well the the core intention of an airdrop is to create a wider token distribution right it's the idea that if you can give many stakeholders some of your asset this is ultimately going to be better for building and bootstrapping the network effect of your of your network and system the other pros is that it's um it's it's hard work you know doing it doing a an airdrop if anyone has done that it is not easy so from the team's perspective it's actually great to incentivize third parties you know kind of to spend their own money and to make a speculative bet on your asset and do the airdrop themselves and finally this is I think a very democratic activity because it's radically open it's open to literally any third-party participant that takes the time to understand the system and to implement kind of a mining software or just use some commodified mining software to do it the cons of Merkle mining is that in general it's more costly than a centralized airdrop if the team were to spend their own money on the on the airdrop they might be able to get away with with a smaller cheque that they would have to pay merkel proofs in themselves just like most proof of work are not ipso facto like useful they're just there to create security and finally in in the Merkel mine implementations that were launched to date that they have been susceptible to front-running meaning someone could come in and if they see you submitting a transaction with the proof they could sort of snatch that proof resubmit it with a higher gas price maybe get in and this has actually happened during the mind though didn't have a huge impact on the outcome so these rules set up a pretty interesting crypto economic game put yourself in the in the shoes of a miner and you have a bunch of trade offs so for one if you mind too fast you won't get a lot of fees because remember the fees are very small in the beginning you have to sort of wait along that 15 months timeline in order to get a bigger chunk of the reward but if you mind too slowly your competitors will come in perhaps and take the fees for themselves and leave you with nothing so you have to make a decision of what is the optimal speed with which I mind if you pay too much gas in other words you're very eager then you will end up overpaying for LPG this is actually a way of converting ether to LPT tokens because that's in effect what you're doing but if you pay too little gas then your transactions won't be confirmed in time so the question becomes what is the optimal gas for me to pay and how do I calculate them finally you know the merkel proofs are pretty expensive to compute in fact if you were to do it one at a time at 30 seconds a pop you would end up with about nine hundred and two days of computation so you have to decide do i mmm do i compute these proofs one at a time as I'm mining and if you do that then you have placed limits on how fast yourself can mine or do I mine them all upfront and if I mine them all upfront maybe I can sort of preempt other miners by redeeming all of the proofs if I know if I notice that my competition is sort of ramping up and of course you can create a parallel computation which mines this whole thing but it's not necessarily cheap to do that so here are some of the results this was written up in multi coins post about new models for token distribution which talks about merkel mining instead of 15 months this minded uptaking 68 days which is actually kind of bullish it means a lot of people were willing to take the risk of holding this kind of asset and also spending ether and redeeming it Merkel miners spent about 2,000 if they're about half a million dollars worth of worth of ether at that time in order to execute this mine and there were a lot of them we were actually the single biggest miner which you see reflected there and the LPG mined in that column is just the is just the rewards that went to the miner doesn't count the the actual LPT that got air dropped to to the folks who were eligible and if you're interested in in in diving deep into the economics of Merkel mining or thinking about instrumenting a Merkel mine for your own token supplier Network check out this post on the case study it's you know a lot of learnings here in this process they were good and bad and certainly we would love to see more teams taking on this model and trying it out and innovating on it so in conclusion maybe decentralized networks will take over the world we think that generalized mining presents new opportunities for all the participants in these networks to first of all have a say in the governments of those networks second of all to make a return for providing something useful to the network and unlike let's say traditional quantitative trading that makes money for the sake of making money when you do generalize mining you're actually adding value and helping these networks to be sustainable and to exist in perpetuity and if you if you're thinking about doing stuff like this come chat with us we have some experience and consider the role of these kinds of crypto economic games or third-party services within your decentralized network thank you if there are any questions I'd be happy to answer them