ZKPodcast: Diving into Cosmos with Sunny

Transcript

[Music] welcome to zero knowledge a podcast where we explore the latest in blockchain technology and the decentralized web the show is hosted by me Anna and me Fredrik [Music] in this episode anise is done with sunny to talk about his journey to cosmos a bit about that protocol and his work as a blockchain validator today I'm sitting with sunny from cosmos how's it going

sunny going well how are you doing very good so I say sunny from cosmos but do you would you say like are you sunny from cosmos or you sunny the validator are you sunny the podcast host I guess a little bit of all but you know I like to I would say my head is mostly in cosmos with and everything else I do is just you know validator I'm kind of help cosmos as much as possible and in epicenter you know it helps you learn and so you know all that learning goes back into improvement cosmos and we're sitting here actually in New York during blockchain week there is a little bit of noise in the background you may or may not hear so how was your week sunny my week was pretty great yeah New York

boxing week was a little bit you know stressful like you know lots of running around and stuff but overall I love music and I love New York I'm I grew up in New Jersey and so this is like my city nice were you here last year yeah I've been here for the past two years how does it compare way quieter like last year was like insane this year much emptier I think it was so much saner no Lambos outside this year no Lambos people have been humbled but people are optimistic so what I realized in setting up this interview with you is I don't actually know that much about your background where did you start I guess you know you could say I first learned about

blockchain like in high school I heard about dogecoin sponsoring the Jamaican bobsled team Wow and I just thought what and so I kind of looked into it I didn't really thing and then there's this guy senior year of high school because a really scammy project that people may have heard of its called veritasium it's like kind of made it into the top 40 during like peak 2017 like super weird scammy project but that guy reggie middle and i think his name was he kind of like I was at a hackathon senior year high school and he was kind of like going on about is in New York where we are now and you haven't going on about like oh Bitcoin assume he's like the next big thing and so when I got to Berkeley so I went to school at UC Berkeley I was studying

computer science and political economy and so kind of remembered what he was talking about Bitcoin and I was like okay this seems like a cool combination of these two fields and in hell maybe and you know I was generally kind of pretty libertarian Ash and stuff and so there was a small Bitcoin Club at Berkeley it only like five six people and so I went there and I had no clue what was going on because they were just like they were talking about like like something about the ring signatures and Manero and it was just like push way over my head he jumped in like at a very advanced level yeah so I have no clue who went on but somehow I felt a little bit smarter by osmosis I felt I was like felt smarter being in the room with them and so I just kind of kept going and

then what I did was the fault of sophomore next year I started teaching a class on Bitcoin me and two friends from that small club and we just said you know my strategy when I want to learn something is I'll sign up to teach it and I better learn it and so I did the same thing I taught a class on Switzerland once it's cuz I wanted to learn about Switzerland nice yeah I like teaching so through that I kind of learned about it back then we were all like very very Bitcoin maximalist like the entire club this wasn't blockchain at Berkeley yeah okay and so that keep mine this club was like five six people and mostly just a bunch of people just like hanging out like

getting dinner just like talking about like bitcoins like oh what's Roger over up to these days and like what year is that roughly 2015 okay yeah so and then what happened was after that class we started teaching our class we got about sixty students the first semester and from that we kind of had this idea like oh what if we like turn you know first of all first off we realize like okay half the stuff we talk about is in Bitcoin we started talking about like you know Manero and the Zeke asked who's come just coming out at that time and back then we were much more interest in privacy like scale abilities like oh okay whatever that's cool but like you know some reason we were all like really into like Bitcoin

mixers and stuff so see cash was coming out at that time so then we're like okay we don't talk about Bitcoin half the time but just call it rename it to blockchain at Berkeley and it's car just like scrap that old Club let's just start a brand new Club and just you know kind of let's make this like a cool you know we saw there wasn't many big boxing clubs in like in universities the biggest was probably the one at MIT Bitcoin Club other than that there really wasn't too many sore like our let's make this like the center of blockchain education on the west coast I didn't actually realize that you're a part of the sort of founding of watching at Berkeley yeah yeah there's our that's funny because it's not a pretty wasn't a program it was a club yeah but a lot of

you are now working in it yeah are you did you finish your degree no I didn't so I'm one of those yeah what happened was summer after sophomore year I was interning at consensus for a little bit I wasn't a huge fan I don't know some I just kinda want to work more protocol their stuff there were much more application layer at the time I think the Pegasus team didn't quite exist at that time it was just getting started so I worked on like their consensus Academy stuff and I worked on this culprit they called Alessio but kind of halfway through the summer I kind of just like stopped showing up and I was I told him like hey I'm gonna switch to like part-time or something and but that

summer I met this guy Nate rush who was also in tourney a consensus with me and he and I kind of just got like obsessed with proof of stake so that summer I was living in New York but I feel like didn't enjoy it because two reasons one I wasn't 21 but two I just me and Nate to spend every evening that summer just like reading white papers like we didn't do anything else and I'm like that's kind of why I want to move back to New York cause like I need to enjoy it this time well it sounds like you did something productive I did I hear you it's like you missed the experience yeah you know the classic New York life I don't know that so we started just really deep diving proof of steak together and so we started reading a lot

of Vlad's stuff and batalik stuff and J stuff and so while doing that I kind of sought Endermen and I'm like oh wow what a nice like practical implementation of proof of steak that I can think it actually like work pretty quickly and then Nate was more like oh no this like Casper cbc is this most beautiful elegant thing mankind has ever created and so and then your friendship ended we run a tiny little podcast yeah they're still yeah so he but no so he actually started reach out to Vaadin started working with vlad directly on casper CBC and I reached out to tender mint and I'm like hey I know you guys are doing some

cool proof of stake stuff how can I get involved with you there when was that roughly 2017 mid mid yeah so summer 2017 so June 2017 right before that the whole cosmos fundraiser like it wasn't part of that or anything I max Fang pointed out the fundraiser to me once and like I said keep my back then we were all super Bitcoin maximalist and so I looked at it and I'm like Oh proof of stake nothing doesn't work we had like a talk about like I think the class we had a lecture on scalability yeah and you know 90% of it wasn't lightning and stuff and we're like oh you know increasing block time it's decreasing the block time that's

that's insane how could you ever even think of doing that so from there you join cosmos what was cosmos like at that time they had they just done the fundraise yeah so it was like a month or two after the fundraiser just happened and so we were kind of back then we had told people we would launch the cosmos hub meaning that by q4 2017 easy my goodness it turns out that didn't quite happen and so but about thing it was pretty small I think I was like maybe that person to join the team I think so just much smaller the team was much smaller

than it is today and it was pretty cool like you know I just felt like I was learning a lot I mean Jim yang who was like the Seaton that had a strategy there right now he he's kind of the one who brought me on and you know I was at teaching me and I kind of learned a lot from Jay and Ethan and then come the end of the summer I was like wow this is just such a great learning experience I think I learned more in the last month than I didn't like while in school so it's like all right I'm just gonna drop out and join this full-time do you think that so I I mean I've met a bunch of people in the space who have dropped out yeah I don't think it's I personally don't think it's the greatest thing for everyone but in

your case like do you regret leaving school um no I mean I think school universities have three purposes which is education community and degree or signaling I think education is free if you want it like I've been taking a lot of EDX and Coursera courses taking on like game theory kapag Rafi taking a great one called um economics of money and banking to give me so much more sympathy for fiat currencies like I feel like no one should be able to allow to talk about like economics of cryptocurrencies until they take this course cool what's where is that it's on Coursera from Columbia University so great course and so yeah education is

free then the community I think community is probably those valuable aspect of University this is why I continue to still live in the City of Berkeley even after dropping out and so I'd still keep attending blocks in a Berkley event I'd still go like sit in on lectures that go hang yeah like you know we have an office in a cosmos office in Berkeley but often times I'll just go to the library to like the school library to work cuz so I can meet people there and stuff but also it sounds like because you were a part of that group you kind of had the benefit of the community already yeah so I would say I couldn't have done that unless I at least was there for that first two years right so join get like get bootstrap your community there and then leave unless unless the degree is

necessary for what you want to do yeah so like you know there are some people who some feels where like a degree is really nice or I mean the other thing is it sounds like you found out what you wanted to do very young yeah in my experience I didn't know what I would wanted to do so it took me a lot longer and I'm actually pretty happy I have a degree because you're my search I was able to do a lot of jobs because I had a degree mm-hmm and that definitely allowed me to then find what I wanted to do so but I think in your case if you know you want to do then it makes sense right I mean you know that's a good point like I don't know I know what I want to do right now who knows how the chain is in five years

own you decide to become a professional gardener yeah I need to go back to school for horticulture there's a really good horticulture program by the way it's like a state school so it actually has like the government mandate that it has to teach agriculture and mining like we could probably I wonder if we can uh convince them to switch it to like you know learning blockchain mining instead might be more profitable yes yes now I think what we understand of your story is like where you started what it started to get you interested started working at cosmos mm-hmm already we hear kind of like your philosophy had changed like you had been

sort of a maximalist and then you started to be open to proof of steak stuff yeah would you say like did that switch in your philosophy happen before cosmos or did it happen during cosmos like did you join at all skeptical and then she changed or like what yeah I think what happened was when I so at some point near the end of my sophomore year and 20 like you know maybe spring 2017 I kind of started to be more open to aetherium someone from digits Dow came and gave a talk about aetherium and that was the first time so fun fact here's the my first experience with aetherium there's a hackathon in summer 2016 that I was at with a friend and I

told him that hey I heard about this aetherium thing like I heard you can it's a blockchain but I heard you can build applications on it let's you know let's look at it and so after hackathon the first thing we do is pull-up reddit.com slash are slash aetherium and the entire front page is oMG the Dow just got hacked at the world is burning we got a Ford so I found out about theorem during the Dow attack and I just like these guys are talking with like hard forking and stuff I'm like alright these people are insane I turned it off and like didn't look at a theory I'm again for like another nine months so then finally when that did it though guy came he kind of like it was finally exposed me to theory I'm after that and

so I'm like okay this theorem thing is really cool and so then that's like he said that I started interning at consensus that summer and so I was getting really into or thorium then so I think at consensus 2017 I met David vorak from SIA and he I think I asked him about like oh why is SIA still running on his own block train while storge is like kind of switching to using a theory I mean like using everyone's switching to your theorem nowadays like that's that's just the hot thing and so he kind of pitched me on why application-specific blockchains make sense and why Asics are actually good and so you know that just got me really thinking out you know he kind of

put the seeds of doubt of a theory I'm in me and like this whole I know we talked about this last time I was on the package podcast but like mysterium Empire and stuff and so he kind of showed those seeds for me and so when so like I said I you know I was thinking about this multi-chain world but then I reached after tender moment with the focus of this most who's working on proof of steak I didn't even know that the tenorman team was the same as the cosmos team but when I reached out to them they said oh yeah we we work on this thing called cosmos and so when they explained it to me two things cross online like one okay that sounds exactly in line with what like David was saying and it seems to make sense with me and two I'm like okay

finally this is like the application layer for Bitcoin I want know I got into because I want Bitcoin to be the money chain that just and then BTC flows out of that chain and into all of these other chains and we can act as the primary currency on many different applications and then we don't have to pollute the Bitcoin chain with all of these applications which is what kind of happens with aetherium and that kind of how I first started approaching cosmos might be a little bit of still how I approached a little bit as well but I think I've come up way less Bitcoin actually so I do see that I do see the neccessity or you know benefits of a little bit more feet more base currencies and stuff than just Bitcoin

so far we've never done an episode really on cosmos what we've done is we've done an episode with Ethan on test nets when you guys were running or I don't know if you'd really announced yet but you're about to do the game of stakes yeah and we were talking about test nets and value bearing test nets and I'm pretty sure he would have shared with us a very brief description of cosmos but now that it's launched yeah how do you describe cosmos cosmos is I think the best way to describe it is it's a somewhat of a rejection of the etherium development frame style where it says let's put all applications on a single chain cosmos basically says okay you know smart contracting platforms are

great for what they sound like smart contracting I'm making a short term a smart contract with you I'm like you know a bet or an escrow contract or icos or the perfect use case for smart contracts because they're short term use and they need high levels of customizability that's what aetherium was meant for in my opinion when you start to build more complex applications we're talking Dex's prediction markets like maker Dow like these kind of much more complex complexity app based applications I think that's where our theorem starts to kind of fall apart a little bit it can't scale both technically but also socially when it comes to who decides

what upgrades to make to a theorem and you know there's so many conflicting vips and stuff and it's like how you upgraded through and become some technical is a social challenge and you know I don't think the Bitcoin people's way of solving it is correct either which is oh we're never going to change anything so our strategy is sort of say okay if we can put all of these applications on their own chain and they kind of have sovereignty over their own train and the sound of what David was telling me about he was saying like oh there was like this fear of a six in the SIA community and their community was able to coordinate to fork out those Asics and that's something they could do because there were a smaller community with heavily aligned interests and so that's

kind of where cosmos comes out it where we say okay these larger applications but whether it's Dex or something you know we should put it on its own specialized chain and but as long as it's still able to interoperate with the rest of ecosystem that's the nice two things that I think did really well and nailed one is the composability where it said you know I can go buy a crypto kitty using my make or die on the 0x tax right I think that's something you couldn't do before etherium and so that's where cosmos is whole interoperability comes and I want to be able to compose different applications together and then the other thing if theorem did was I made it really easy you know solidity wasn't it's not the

nicest language in the world but you know I've done a lot of it it's not that bad anymore it's getting better and better and it's way better than like you know dealing with the spaghetti code like that is the C++ Bitcoin core client and so it made it way easier for developers to write more complex application that because of that we got much more diversity in applications pre aetherium all box and applications were like Bitcoin or like coin or any tower send private money exactly yeah they want the few that tried to do something different was like named coin right but like it was a fork of the Bitcoin core codebase and it just doesn't it's very badly architected for its use case so if there were made that easier and so

that's kind of the two things cosmos is kind of do is basically say okay let's we want that social and technical and technical scalability that was somewhat president in generation 1 but that's with it while maintaining those nice benefits from generation 2 which was decomposability and ease of development do you I mean the early eath aetherium idea was the world computer and they'd be able to run also of applications on it is that more in line with what you guys have planned now that you would run applications I mean you mentioned these sort of different chains yeah but is that is the overarching goal this sort of idea of a world computer so I don't like the term

world computer that's actually probably the main issue saying that we can have a singular world computer that sounds unscalable to me what I think sorry yeah we actually did an epicenter episode with Mark Miller from a gorrik just last week and so I think he nailed it when you know he starts talking about like what we really need is bft computers that doesn't mean the BFT is world scale but you know at the end of the day these block trans are just third parties but they're highly decentralized third parties and therefore somewhat more secure maybe but you know if we're trying to settle something between us we're kind of saying okay instead of having a third party escrow will use this blockchain third party as an escrow

and so you know does that mean and so what you really are going to have is a world of many bft computers that are all connected to each other that's kind of really the Internet of blockchains catch phrase the cosmos comes in where it's like no we don't know we don't want a single world computer that's that's crazy we want many BFT computers that are connected you know world wide network when you think of like the sort of earth 2.0 and the shards is that it all related to what you're thinking about what's the difference between cosmos neath 2.0 well one thing with e 2.0 is it still continues this idea of still using the e vm or whether it's

webassembly but some common vm that's really designed for smart contracting and so and then it also kind of says eath 2.0 says all of the shards in e 2.0 kind of have common security where they have to use the saint's they all have the same security requirements which make sense in the etherium world I think it makes actually a little bit less sense in polka-dot but in aetherium and makes sense because all of these shards you know we if you want a highly secure smart contracting platform you know you want and smart contract or interacting with other smart contracts and other chains on other shards yeah you know maybe they should have somewhat

equivalent security because they're supposed to be homogeneous cosmos on the other hand and polka dot as well are kind of really focused more on heterogeneous chains and shards right where they're application-specific one one shard might be tripped or KDS and the other side might be maker Dowell I think we'd agree that those probably don't need equivalent security and the crypto kitties people are overpaying for security that's why you know their fees are higher when they don't need should be paying that much for security they're overpaying and so that's where cosmos kind of approaches it where we say no these different chains don't need equivalent security and so we allow them to either figure out their own security model with by doing sovereign chains

whether they want to use permission validator sets or proof of stake or even proof of work if they wanted to they can have their own they can figure out their own security or they can ask the cosmos hub to lease security but that's also done in a more kind of diversified way design fred rottenness change where and even in polka dot where they kind of have the same a similar idea but they're all the pair of chains kind of have to have equivalent security which i think is kind of missing the point when you're getting into a world of heterogeneous chains in cosmos we can have different chains even that are doing some level of shared security still have different levels of security

how are the how are the chains connected actually what are they called in cosmos what are the different they chains we call them zones a lot of biggest fan of the term I mean I just call them change her zones okay side chains okay so that how do you connect the chains in cosmos so what we've done is we've created a protocol called inter blockchain communication or IBC and essentially at the end of the day it's a fancy way of saying side chains side chains are not a new concept that you know I think block stream pose them back in 2014 it's just the idea of I can do SPV or like client proofs of one chain in another chain so

I can prove something about the state of another chain and so there have been many side chains that that have existed right you know one is a BTC relay which is a one-way side chain from Bitcoin to aetherium there's parity bridges which us to talk to other EVM chains what IBC is really trying to do two things one this like line protocol becomes really expensive when you're doing two proof-of-work change the peripheral work chains I think the BTC relay because what happens is when per for you have to transmit every single block header to sync the chain and I think the BTC relay contract hasn't been I think no one's been transmitting headers in like months

and so you can't actually use BTC relay because it hasn't synced up to the main net and to sync it up to the main neck will take so much gas right now on a theory I'm that I don't think it's worth it for anyone to do it altruistic Li and so one thing that tender mint or BFD protocols do in general but I'm actually more specifically tender mint is it provides a very efficient light klein proof where you can quickly skip blocks as long as the validator set doesn't change by more than a third in a block if that means we could probably jump like thousands of blocks at a time because if the validator set like isn't going through Oh too much of churn but so one of the reasons I'm not a huge fan

of like sortition based stuff that changed the validator set every block because now you made the light client very expensive again so because of this much more efficient like clients we are able to actually do this IBC sidechains protocol in a much more efficient way the second thing that we're trying to do here is oh wait before you go on the light client so what you're what you just described it's like you you you don't have to verify each one you can skip for the light client what does that mean exactly like what are you know maybe describe what a light client is in this capacity sure so a light client is basically saying I'm not running a full node I'm only syncing the headers of a chain and then I can use Merkel proofs to know that something

happened in that chain that I wanted to happen yeah we've we've actually covered like clients generally on the podcast but what you just described is like your are you even skipping you're not even registering all the headers yes so this is a this is a unique light client yes okay so because what happens in proof of work let's say I want it if let's say I went home at night and I turned off my phone and I turn it on in the morning again overnight there's been what six times eight so 48 Bitcoin blocks that came overnight my phone takes about a tested it I can do about three to four shot 56 hashes per second so that means it'll

take my phone still at least over ten seconds just to sync the Bitcoin block headers over that came overnight that's Bitcoin which has very infrequent blocks and a very easy hashing out for them I actually haven't taken the time to measure it for aetherium yet but you know it hashes are much more complex hashing element that's more expensive and this blocks every 15 seconds so it'll it might I don't know I don't want to say this is true but you know it might take my phone a minute just to sync that what tenorman does is be quite happens is every block the validator set from this block signs on the validator set of the next block and so wait and so at a certain time you may have a trusted validator set who you know you know that

you synced up the chain and they got to this point legitimately then what you but then you go offline and come back like five days later if the validator set hasn't changed by more than a third that means at least two-thirds of your last trusted validator set is still in the current validator set and they you can see their signatures on the next validator set as well and so you can say okay at least two-thirds of my last trusted validator set is approving of the current validator set so that means I can trust it I can trust it and so I can skip all the blocks that came in the last five days I don't have to sync all of those headers so how does a light client and the eye

see actually you interact what's so connection what you'll be doing is on in the state machine of chain B you'll be running a light client of chain a C and so that's why what will happen is I can take a token or something on chain a and lock it up in a special I'll call it a smart contract a special spot in the smart contract in the in that in the state machine and then I submit a light client proof to chain B's saying hey look here's proof cryptographic proof based on the header and a Merkel proof that I've locked my token away and then chain B will mint you a token saying oh

this is a claim to the underlying locked asset on chain a and I can do whatever I want with this claim I can send it to someone I can do whatever and whoever has access to it they can then go back and what they can do is they can burn it unchain be in a special burning contract or something and they can submit like client proof of that burning to chain a and it will unlock that lock token and give it back to them and what you just described is a bridge yes we had P way network on we eagerly on we actually had to be conversation about punches yes and it sounded very similar he said there was two ways of doing it you could either mention burn yeah or what was the other thing he did there was like lock

and unlock I mean if it's lock and unlock that's necessary when two chains maybe are relatively separate with each other mint and burn makes more sense when there's people because you don't want to burn the original asset you only want to lock it if you burn it that might make sense in an aetherium like system where like all the shards are kind of you know the master shard we just assumed all of them have similar security and that like you know the entire system is like secure in the same way so it's okay to like you know allow the accounting to be a little bit more loose in cosmos where chains might not truck my BFT computer might

not trust a PFT computer that you're using and so we need more gear tee's and like accountability keeping track of like that yeah so making sure we lock it rather than burning in this important so you guys do the lock you don't but didn't you just say you mentioned burn on the loss we lock the asset on chain a uh-huh minted nation' on be okay burn on be unlock on icy and so that was what I think her head said something like that he was like you can either do mint and burn mint and burn mint and burn lock and unlock yeah unlock unlock and lock and unlock yes we're like the options yeah and you guys are doing so is there one is it always like directional it's like if you are moving funds from one to another

it's always gonna be lock and unlock on the chain that is moving from and then mint and burn on the one it goes to yeah when we're talking about assets here token transfers IVC is pretty you know generalize knows the side chains paper from Bitcoin you know from block stream they were really focused just on asset and that's where we started as well so that's kind of why we still use that side chains terminology rather than British because we came from the Bitcoin mindset there but IBC is like I said all you're doing is proving some skate to another so token transfers and asset transfers are the easiest thing to start with but you can start doing more complex things for example you know let's say you have two EVMs on two

different chains you can in theory do a smart contract call from one chain to another I personally am NOT a fan of that one because you know no one's really figured out what the design of that will look like and you know the e to point our people are kind of working through that stuff right now but I think there's you know maybe better design VMs for that purpose and so one of the ones that you know we're really I'm really interested in and we're working closely with is a project called goreck and so they kind of have this very fascinating smart contracting vm that allow it's basically keep object capability based security which is super interesting to me and I think it is the right way of doing a

cross chain like contract and stuff when you talk about so you kind of mentioned the comparison to ooh eath - you hinted at a comparison to poke it on what you've described I mean in in in what you described as the difference you said that the difference has to do with this do you need generalised security or can you have chains of different security levels like you sort of described so like some chains like the crypto kitties Jane maybe doesn't need the same levels of security as the maker chain but what happens if you're interacting between these two like does like what if you were able to sort of corrupt the crypto kitty done and then you would basically

lock what would you do in that case you'd lock a crypto kitty token and you'd send it to a more secure one and it would mint and on the new on the more secure chain it would it still mint it yeah it's like is there any way like wouldn't that then mean that the the bridge itself could be an attack vector yes and so that's why the what happens is you have to have trust it so what a token moves cross the system it kind of you have to trust all of the chains and which goes there and that's why I like to avoid saying that you mint the asset on the same asset on the other side I like to say you meant a claim to the asset on the other chain so what happens

is when I locked a crypto kitty on one chain and I minted that crypto kitty on the other chain I didn't mean to cryptic ready for that chain I'm in tin a claim to that underlying crypto kitties and so that by using the terminology of a claim I think it's those people that you all still have to trust on a virginal crypto kitty chain because if there's no underlying crypto kitty then your claim is also like useless but like what if there's a chain that I don't have any control over which is bridge to like my chain yeah can i it the the bridge already exists can I prevent them from from doing that from minting on my chain if you can prove that a chain has gotten corrupted you can submit that proof to the bridge and

it will close the bridge but no I mean you you in ghent it let's say that hasn't happened yet and you know let's still continue to mint assets on your chain and just say your users have to be aware that that's not you know you did that has a security requirement there but one of the things the cosmos hub helps with is it keeps some sorts of invariants between chains so it prevents change from double spending other chains so let's say the hub is connected to many zones as well as the Bitcoin blockchain BTC might flow into the cosmos solvent into chain a chain a what it should do if people users want to go to chain B instead of sending BTC to chain B directly now you have a path

that's like chain Beach and a hub and Bitcoin you trust all those four chains well you should instead do is withdraw the claim from chain a back to the hub and then then send it to chain B so you don't you try to shorten the path whenever possible so there's less of trust and the hub keeps track saying that oh I know that there's only five BTC on chain a right now so even if chain a gets corrupted then only you know chain trainee can't send out it can be DC and n will spend yeah only has five BTC so you know the users who have launched an a maybe they might get defrauded but the overall network doesn't collapse just because one chain

gets de frauded you just mentioned the hub what we should do here is explain to the listeners what the pieces of the cosmos network are as we mentioned the hub yeah you mentioned the IBC mm-hmm so let's let's try to create a map like a visual map for people or actually it would be an auditory map because they don't have any vision so essentially what will happen is you'll have chains that can set up an IBC connection to any other chain they want that also supports IBC so ID see you have to a chain have some natively support IBC it's something you have to build into the system and this in other context would be the bridges yes okay

now just like any airport in the world can theoretically fly to any other Airport in the world for practical matters what we end up seeing in basically any district network system is you see hub and spork hub and spoke like architectures begin to arise and so this is like hub and spork so this is what the purpose of hubs is in this system is it helps reduce the number of IBC connections or bridges in the network from order of N squared to n right where and that doesn't mean that there's a singular hub that means that there will be a couple of hubs that are

connected to each other and then are connected to many chains that kind of you that use the hub so these cotton these hubs in the system that there's already two hubs that are in the cosmos ecosystem there's one called the cosmos hub which is I agree a terrible and confusing name and then there's another one called iris iris hub and so what these hubs should be are they're also application specific block chains and their application is being optimized for transmitting IBC packets and that's like what they are focused on and other features that they can provide as hubs such as providing these shared security features and whatnot to their patron

chains well let's call them cool I really like I like your use of the metaphor of the airport's cuz that definitely you can sort of visualize that somehow yeah um right now you said there's two mm-hmm are they different like can they be unique in their construction or are they sort of replicas of one another yeah they can be unique right now the iris team has it's pretty similar to the cosmos problem I mean they do have some big changes one of the main changes is that they really wanted automatic software upgrades and so like we discussed last time when I was on the cosmos hub I generally don't want the ability to have automatic software upgrades so that's already one of the

main existing differences between these two hubs and then iris is really focused more you know I would say they're a little bit more on the enterprise-e side well the Kosmos hub is maybe a little bit more cyberpunk II I don't know if I know I like to see it that way at least although iris is really heavily focused on in China and so you know people often think like oh you know Neal was the Chinese aetherium and you know the Chinese blank right I think it's actually really interesting because I'm worried about that Great Firewall of China and once if that fire will start censoring packets going over it then I do not want to do tenorman consensus over that firewall because yeah you know maybe you could do you could do it over

tour or something but it'll be really slow what I'd rather have is a major hub in site China and a major hub outside of China that they own have their own little eco systems and in your drink tenement consensus separately and then they can send IBC packets to each other over the fire over the firewall through whether it's tor or you know some of the big corners are working on like these cool software you can like transmit over like long-range radio networks and stuff so but this is a different like that would be much smaller I guess what you'd have to like between the hubs the communication could be a lot smaller yeah that's kind of one of the premises where we come from where is a notion of locality like that's kind of one of the

premises of the whole application-specific blockchains is that in an application most of the transactions stay within that application and then the minority are elsewhere that's one of the problems with like I think the youth Tohono kind of design is where let's say you're randomly interacting with another contract on aetherium on East 2.0 and let's say the for example is ten shards there's a 90% chance that the contract you're interacting with is on another shard and the inter train communication is the bottleneck so when you have application base specific systems it's like you know let's say add X most of the transactions are probably trading on the decks and the minority are probably deposits and withdrawals from the decks

and so that you know that helps and then the other kind of locality is geographical locality you know I I'm a big believer in and so as Ethan Buckman he's the really one who got this idea into me was about like in a localized ball change most of the financial trends are you do probably every day are people in your close geographical proximity you know I go out buy lunch and you know the most of it from in a single town or maybe no neighbor like even sometimes the neighborhood and so you can get way more locality that way I one thing I'm not really clear on is like since these two mealy like these are independent block chains here talking about the hub are also block chains right so how are they actually localized like are they

not decentralized as well like do they not have their own nodes and like node operators that are spread out all over I just I don't quite understand how there's like a Chinese node or a Chinese hub yeah so that we could be talking on two layers we could be talking on the user layer and we can be talking about the consensus layer and so first from the user layer let's say that the entry you know that's kind of where that locality of transactions comes in where if there's you know yeah let's say we're New York City right now right we have NYC chain right we're which is four payments within New York City if all the users people who are actually making transactions and you know we're all have

accounts on the New York City chain we're just moving money quickly in the New York City chain and not and very few transactions are going to other chains so that kind of what I mean by user locality maybe the blockchain nodes are could be geographically distributed for example but user the users of the chain everyone who has a count on that might still be localized now you can also take it a step further and localize the consensus layer like say all the validator nodes have to be somewhat localized as well so you know tenorman it becomes faster you know it's limited by the speed of light right and how fast nodes can communicate so if we can move stuff closer and all the chains in a localized area then we can actually that

helps as well and you know I'm a fan of this idea of what a nice thing about cosmos in the sovereign chains is you can have additional requirements beyond just proof of stakes so let's say a town could say oh the validators of the chain have to be New York City residents or help this project called librarian it's just like little joke country look at micro nation it's like a strip of land next to a river yeah yeah exactly right between Serbia and Croatia I actually visited it once to school and so you know I help them out with some like blockchain stuff and you know I was talking like you know what you could do is have your system be proof of stake but it also to be a volunteer you also have to be a

librarian citizen that seems to make sense and you can you can and I think it actually gives you higher security than if you didn't have it because I think sometimes like crypto econ people think that economics are security is the only thing that matters when in reality people have political and social motivations right and so if all the people are labelling citizens maybe they're less likely to want to attack the system I'm trying to picture so like say you have these a small you have a tiny localized blockchain I mean the thing that's a bit scarier there's like what if you're so localized down to like five of your friends you all ran the same nodes I mean this is kind of going

back to that idea of like a corrupted yeah I one thing I'm curious is like how does an IBC get built and like do you prevent some of those from being built like do just cuz like what if I me and five friends are like me alone my local computer I have the node I have what one node it's my blockchain and I do all sorts of double spending stuff and I try to like interact with it you sort of mentioned that there's like almost a good like a court system or something that would knock it down but like before before you know that I'm malicious can I can I build an IBC can I build a bridge yeah and but you know all the only funds that would be at risk there are funds

that users willingly put onto that chain so if it's just a chain with you and five friends it's probably only people who are probably gonna be using that is you and your five friends and that's okay that makes sense like you know people say that like you know we're gonna use lightning to settle transaction between our friends and lightning just makes like it seems that there is zero trust that it exists the entire world and thus makes it actually very expensive transact and you know if I just have a group of my friends like you know I can make a little side chain for ourselves and we can just do all of our transactions there and you know I have trust in the majority of my friends so how do my 10 friends I you know I assume at least five of them are honest right and you

know if that's not the case I should probably go find better friends but what would happen if it's like it's not I'm not worried about the chain itself where I do the double spending yep or the five friends do the double spending I'm more like when you start pushing it kind of going back to that other question when you push it towards these other chains yeah but this is what I want to understand this like so say I do it let's go through the steps I create an IBC I'm allowed by bridge to somewhere I do the lock and unlock on my chain I meant on another what what token are locking and unlocking chain let's say let's say I make it I make a token that's super generalized brand new token the brand new token yep I guess this trick is here like I need

to have some value in it yeah I can't I mean a DC on my chain okay okay okay so I have to have a chain so then I see what you're saying it's like I get five friends we have a little mini economy that's very minor value yeah but I lock it on my chain I double spend I lock and unlock right but in this case a lock I meant on a new chain get that money I do something with it yeah and then it turns out it's faulty it's total I like disappear original five person chain faulted somehow yes yes and so when you're using a token on another chain that came out at over IBC you should be able to see where this token originated from and if you don't trust

the origin chain then shouldn't be using this token so I can if I don't trust bitcoins like root chain I should not be using BTC on any other chain because the original chain might just fault and destroying my value same thing here with you know that those that hope that token I was issued by me and my four friends you know the people who are gonna be trusting it is probably gonna be pretty small and the only people who should be using this people who have some notion of trust in that initial issue and so that's why there's different types of BFD computers there's some that are gonna be like large-scale highly decentralized global computers like the cosmos hub and then

there'll be some which is just me and my friends but let's use it maybe a different example let's use an example of one of these chains where it's super centralized we know it's centralized and like so and they're in it is value bearing and it yes for some reason due to speculation and good marketing people have like bought into it they own these tokens and what if like that central power starts to play around and they double spend like you know the question is like could like if they don't mind destroying their own network could they all of a sudden access a ton of funds in another chain and and like what you said there it's like it's traceable sure but like if somebody if it's sort of BOTS trading like nobody will notice nobody will care if somebody's selling yeah any

different than if like you know even today let's say Ripoll decided to inflate the hell out of XRP right and you know they could just dump it on the open market and like you know I don't see how it's any different today it's sort of like it if they could do that on the decks right now yeah it's like you know yeah why didn't maybe the attack surface of Dax's and so maybe it requires a little bit more education of users that are you should be wary of the risks that you're taking but we should already be doing that today and I think we might be doing a little bit less of that today that we're supposed to be would you say say that say that sort of say it is like a major chain a chain and then we we learned they've acted

maliciously what would happen like what is the what is the repercussions what can the network do like is there a way to go you cancel the bridge yeah so that's right so the IBC yeah so when the if you can get proof that there is that chain has forked or something right then yeah you can just cut off that bridge and say okay there's no more I we see connections happening no more packets are gonna flow back and forth from here you can have governance on the chain try to figure out what to do with it or you can say okay you know chain let's say it was like a legitimate fork like you know what classic style fork well that case that there in classic is the real aetherium but you know if you ever have the improper state transition but you know let's say you had a fork like that right

and then you know you can have say okay both of them can register I knew I BC connections and governance can figure out which one gets to act as the originator of the old one kind of like you know when China broke up right the the UN had to decide okay who gets its seat in the Security Council hey I'm sorry you yes that in that kind of those kind of situations you do get a little bit fuzzy and have to kind of rely on human social coordination speaking of I want to sort of switch gears and talk to you a little bit about validators and being a validator and being a Staker so cosmos is a proof of stake system meaning the cosmos hub ah because you

can you different chains in the cosmos system can actually choose to use whatever security system they want like I said you might have a chain that wants to be a permission validator set by Nance Dex which is a chain in that's building on cosmos right now and call on our software and content connect they're currently running a completely permission validator set and that's okay you can have proof of state systems interacting with permission systems you can even have a proof of work system that interact with the state system so the cosmos hub is ain't no we I mean I really like proof of Stegen so I encourage people if they're launching a chain and cosmos to use your state but there's no requirement do that so but the cosmos hub blockchain that is currently running on for mistake and the

cosmos hub is the one we talked about before it's also acting as sort of like a okay the goal is to make it the most secure proof of stake blockchain in the world okay but it's proof of stake so when you talk about this is actually really helpful when you talk about being a validator on cosmos you're a validator on cosmos hub yes that it's it's an unfortunate naming collision there got it so you're a validator on cosmos hub yes are there also validators then on that other one you just mentioned the one in Asia you're cool and it's a different like there's some overlap I actually have a some Irish tokens myself and I might my vodka company a sicko we like we have

our nodes synched up and so as soon as I get back to San Francisco we're gonna start routing on that chain as well cool back to that though you have atoms on the cosmos hub blockchain yeah this is not only acting as a place for transaction to be relayed through or tokens to be checked it's also action I guess from what I'm understanding as the governance area the decision-making body for change that want to delegate decision-making to the cosmos some level of decision making to of the cosmos hub got it did you know when you started that you would be a validator um no so what actually happened was so the company that I work for which is tender mint it's a for-profit company

that is contracted by the inter chain foundation to write the software for cosmos hub with the cosmos hub yes as well as some work around IBC and whatnot so being a for-profit company we did have the intention to so as we're kind of approaching the end of like you know our contract with the interchange foundation basically the stuff that was laid out in the white paper we're about to finish that all and so we were tournament was actually considering running a validator itself and then we decided that I think that tenorman as a company has too much social reputation that we might over centralize the network what if what happens if we

accidentally get over 50% of the delegation that would not be good and so we kind of said okay TENORM it's not gonna run our value here we're actually gonna allow a bunch of a different any employee that wants to can go out and create a new legal entity completely distinct from tenorman at the company and in front of a leader themselves and so now we have about there's about five validators that are created by um tenement employees you started a validator business yes and what you would do is a validator stakes their tokens and that secures these basically creates a security of the system and then pays you a small fee validator fee yes rewards and those rewards get split

sent to delegate your delegate errs as well and the validator can charge a additional Commission rate if they so wish and the delegate errs would be other token holders that don't want to do the validator setup yes because actually so this is what's kind of I'm curious about this so you're if you're a validator you have a node I imagine you set you download some piece of software that allows you to validate on the Kosmos hub and you are kind of like locking your tokens into this every time you do it so you cannot be day trading with them and they're sort of like lockdown how long how long does it take to set it up so running a node as fast but the part that's interesting is you

have to run a suit as about it what you're doing is you're signing on blocks and if you sign something incorrectly whether you double sign or something you can get slashed and if you fail to sign blocks you can get slashed and so what you're really doing is you're running a hyper secure hotkey and you know you can go ask any exchange or custodian that's not always a fun thing to do so in order to set it up you'd have to like do a small or some sort of hardware investment yep some sort of software setup mm-hmm get a security expert on it maybe or your annex security expert yourself make sure it works very well make sure that

your internet connection is really good because I guess you have to be there all the time yep and then you I'm guessing stop me if I'm wrong you push a button that says validate now and it sort of automatically will do these signatures yep consistently and the way that a validator is so ok so let's let's go really basic imagine there are 10 tokens that exist and you have to and Zacky has three and someone else has four what happens for every round like are you getting a chance to sign 20% of the time you I was I will be signing every block okay but I will be proposing the block 20% of the time

and it's only the block proposer who gets the reward no we split the reward even not evenly but uh amongst all the Baldur's the proposal gets a slight benefit and there's a algorithm that we use to figure that out based off of other factors but for the most part it's also but evenly amongst all the valid interest I think in my example I've listed three we'll add a fourth to finish it off so there's four validators as individuals with different amounts of token that they're staking how is it what's the breakdown like you mentioned so this the proposer will get a small advantage but then is it like equal by four it's the proposal we'll get a maximum 5% that depends on some other factors and then the rest is split by

stake so just by stakes on it but it would be like the person who has the 40% would actually get more reward yes yes being a validator you have all the architecture in place you can stay you don't get delegates as you just mentioned people with tokens who want to they don't necessarily have that set up so they're like oh I'd rather you stake for me they're kind of like your supporters for yeah yeah you you are not necessarily beholden to them but they're trusting you maybe your constituents yes and every time you get your block rewards whatever percentage that is you then we'll just you'll give it back to them and you might take a small fee yeah but

it automatically gets distributed like for example in Tasos that doesn't happen and there's been some validators that have run away with their delegate class in the cosmos sdk in the cosmos sdk what we've done is we designed a system where it automatically does it so i can't do that we've minimized the risk to delegator could someone take that SDK and change it though yep and but then they would just have written like bad code that is malicious and yeah but it's like what you guys are providing yeah and like anyone who delegates could actually check to make sure that whatever is being used is in line with that okay right what it sounds like right now is so those four validators that I mentioned they are already token

holders they're already atom holders and they're validating and gonna make some money on that the question that automatically comes up is like well then wouldn't the rich just be getting richer like if the one who has four mm-hmm out of the ten she's always gonna get a larger portion of the reward of the validation reward so her share her general wealth will be growing faster than yours will I mean there will be growing at the same rate right like as long as we're all getting the same she still has 40% ten like next block is she has this block those who only stake sometimes yeah they will be losing out yes and so this is why we say atoms are not money

there are staking Tobin said I think it's a new type of asset in the crypto space where atoms are not money and what we say to people is if you're not staking your atoms like what easy yeah you're losing out like you know maybe we want some liquid just for price discovery but otherwise the atom should be stick and we actually design an inflation system that fewer people that are staked the inflation ramps up so it's like oh here's you should be staking what are you doing go stake you look how much money you're losing out on that's interesting and so that is it's it's basically preventing people from holding yep and sinning and this is what you want yes that's a motive like that's an

incentive that you wanted to put in yeah no like I said for us cause it for me in cosmos like you know I think Bitcoin is money and I want Bitcoin to be the money across many chains or maybe other tokens and grin or something right but atoms are not supposed to be the currency that people use they should be staking do you are you a little bit worried though about like creating a barrier to entry like right now like what about the groups that want to enter later I mean they're the price to participate will get higher higher we met like in order to actually get anything that would be usable you need to be there early um that's true maybe to an extent but you know if someone can come with a personal there is enough atoms available on the public open market that you can

buy them and the other is that yeah like you know if you're able to provide a really good product like you know bison trails for example you know it's an industry grade custodian and like you know I think I know there's a lot of excitement around them because and so if they're able to come in and provide a really good product then they can attract delegate errs too delegating to them and then there's other levers that you know it one of the other things with cosmos is we really believe that validators are somewhat and differentiated we don't want there to be non differentiated validators and so which polkadot kind of has this assumption that like oh all that it is or homogeneous and like you know they can just be shuffled across chains you know one of the things that

cosmos hub will do is like like I said once you can start stating on different chains but valid can choose which change they want to do shared security with and so one of the things about the delegate is a look at what chains are is this guy like shared security with and other things that you could do is like you know one thing that sikka is planning on doing is currently recharging zero percent commission rate just to you know help out people who like but what we're going to do is when we start charging Commission rate we're going to put half of it into sin half the Commission rate into a Sicard out that will be controlled by our delegate errs and then they can use it to fund

you know one of the reasons so I run this validator with dave oda who he used to work at Endermen but now he focuses on doing some user knowledge research with alessandro chesa on this new user knowledge proof system but and he so he does most of the security stuff he's been doing like you know security hacking and stuff for like since high school he and I got into this because we can one of the things that we wanted to do was we wanted to take half of our commission rate and use it to fund open source public infrastructure that maybe isn't incentivized so like we were thinking about like tor relay nodes and ipfs nodes and like Bitcoin public public nodes and stuff and so that's but we're like oh instead of us doing that

like you know we're trying to build this brand it's like the people's validator and so we were like oh let's give this to our delegate and let them vote on how what they want us to fund and like you know if they want to just pay themselves dividends back from that commission rate all right but you know I hope they do some cool stuff with it do you think of is a long-term play or is this sort of like short-term because right now there's not that many validators and it's like a fast money maker because I I have heard that that is how some validators are thinking they're cool there's an amazing opportunity but soon margins will become super small and like my percentage won't make sense it won't be worth it so let's just do it now and get out so Sikkim we're taking a long-term view

on this where we want to be like you know I've been building proof of stake system for the past two years like you know I I design a lot of this cosmos versus taken and you know I really believe in this as a long-term thing and so we want to continue being valid leaders in the long term and you know we are scared that like you know I know Polonia XRD announced that they're gonna start listing staking their atoms that are on their exchanges coin base is going to start doing it soon - and so we're worried about this but sicka our plan is finding a good niche where you know that people's validator like mantra that we're trying to push that like you know these are cutesy little things like like the sick of Dow and I actually minted these like like little gold sick

of coins I'm sending out to be sending out Fidelia to sue and it's like these are too small cutesy things but then like when you start doing other stuff like you know new features like you know pride maybe derivatives on staked atoms for example and when you do more complex stuff like this then you can I I guess it really really comes down to finding a good nice for your validator cool so one thing we didn't get a chance to talk about that much here was governance and how the whole thing is maintained but I don't think we have time so that might we might need to do that another time one other thing I'd love to do some time is bring you together with some of the players and the other spaces so we can understand a little bit better how cosmos is different from some of the

other kind of interoperability plays out there so the last thing I want to talk about is podcasting yes because you are co-host on a podcast that's in the same space how have you liked it it's been great I mean I love it so the name of the podcast is epicenter as many people probably know but it's a lot of fun like you know I spent I have my head all in like proof of stake and interoperability and so what the podcast does is every week it forces me to go learn something new every week and stuff like I said teaching people is how I learn and so you know the part this podcast is sort of an educational experience and so doing all the research for that and for him it's great way to teach people but also learn myself well listen thanks so much for being on the

podcast thank you for having me and to our listeners thanks for listening [Music]