# 1inch | DeFi Value Flows - Anton Bukov | ETHDam 2024

- Channel: [CryptoCanal](https://streameth.org/cryptocanal)
- Date: 2024-10-07
- Duration: 28:07
- Watch: https://streameth.org/watch/yt-AdbG8zcvd7k
- YouTube: https://www.youtube.com/watch?v=AdbG8zcvd7k

## Description

“DeFi Value Flows” with Anton Bukov from 1inch at ETHDam 2024. In his talk, Anton, will discuss infinitely scalable algorithms for value distribution between a large number of DeFi participants.
https://x.com/k06a https://x.com/1inch https://x.com/1inch 

Jonathan Knegtel - MC of ETHDam, data aficionado and your favourite wooden bowtie. 

ETHDam - a conference and hackathon held in the heart of Amsterdam, Netherlands from April 12th to 14th, 2024, celebrated its second edition, gathering more than 600 participants. 

In the dynamic space of ETHDam, privacy and security took center stage, featuring groundbreaking discussions on hacks, recovery, and the revolutionary work of figures like Pertsev. Privacy is dead in crypto, people that know, know. People who don’t know, should know. 

ETHDam is powered by CryptoCanal, an education and events platform growing in Amsterdam, spreading its roots to Rotterdam and Zürich.
Keep up with us to see updates on future events: https://www.cryptocanal.org/ 
Follow CryptoCanal on X: https://twitter.com/CryptoCanal
Join CryptoCanal TG Community: https://t.me/CryptoCanalCommunity 
Join CryptoCanal Discord: https://discord.com/invite/XJVjpCqQBz

We would like to thank our partners that made this event possible. 🌷
Battleship Partner 
🛳Oasis Network https://oasisprotocol.org/

Jet Ski Partner
🛩⛷  NEAR https://near.org/

Canoe Partners
🛶WAKU https://waku.org/
🛶Trail of Bits https://www.trailofbits.com/
🛶Avalanche https://www.avax.network/
🛶Privacy + Scaling Explorations https://pse.dev/en
🛶Threshold https://threshold.network/

Our Canoe Partner & Official Node Provider
🛶dRPC https://drpc.org/

Sponsor
🤝EF Ecosystem Support Program https://esp.ethereum.foundation/

Paddle Partners
🚣ChainSecurity https://chainsecurity.com/
🚣Lido https://lido.fi/
🚣Cyber Capital https://www.cyber.capital/
🚣Diva https://www.divastaking.net/
🚣Firn Protocol https://firn.cash/
🚣Beefy https://beefy.com/
🚣0xbow https://www.0xbow.io/
🚣Obscura https://obscura.build/
🚣Panther https://www.pantherprotocol.io/
🚣Maven 11 https://www.maven11.com/
🚣Zama https://www.zama.ai/
🚣zkSync https://zksync.io/
🚣Secret Network https://scrt.network/

ETHDam AfterParty Fren
🥳Bitvavo https://bitvavo.com/en

## Transcript

[Music] okay yeah that's my presentation so hi everyone I'm going to talk today about algorithms in existing defi protocols defi projects and uh what's uh I consider the most important about them so three important parts here is first that as as far as you can see that we experience that computations in defi they are very limited and uh I in defi development for more than seven years and uh I honestly doesn't believe it will be changed dramatically in few next like 5 10 years so computations will be eventually very limited and uh there's no any chance that if you are building any algorithm which should involve multiple participants that if you will have a linear complexity it will ever work in defi so you should use algorithms which have constant complexity or logarithmic complexity so some of defy algorithms so defy protocols they already have algorithms inside of them which works with the potentially infinite amount of participants like thousands millions or billions and the costs of these operations of these algorithms are constant and the third part is that using those algorithms which I will uh classify today is like a new way to describe how exactly your protocol works on like a a bit higher level than people usually describe it so I came up to some classification of these algorithms all these algorithms are infinite scalable and they are classified by the asset is it the same asset as as the deposit or nominal of users or it's different asset is it happening instant based on some event like transaction or like anything else or it's gradual over time and uh direction is it dispensing or collecting it depends if you're collecting from multiple participants to one participant some values it's it's collection if you're dispersing from one participant to multiple participants potentially infinite number of participants it's dispensing so here I puted some algorith or some protocols and algorithms in those Protocols are classified like that so let's dive into it for example in amm fee from one user is being distributed among multiple liquidity providers and it doesn't matter how much liquidity providers you have is it like 1,000 of wallets or one million of wallets so this swap fee will be distributed proportionally to their liquidity with 01 complexity for example for money market we see that those who borrow it borrowers they are paying gradually interest and this interest is being redistributed among lenders liquidity providers so it looks like two different algorithms I involved here one is uh collecting assets from multiple participants and another is Distributing uh value among multiple participants but actually money markets implementation uh is uh a bit tricky and they use the same algorithm so it's the same algorithm of distribution but for borrowers they are just Distributing same asset they depth it's like Distributing of depth token but it works so let's get back to this classification so you can see same asset instant dispersing it's basically amm and many other amm here are just the most popular ones for example same asset gradual dispensing you can see it's a compound maker Dow synthetics die it's a d savings rate a so they are dispensing values even if it's depth they are dispensing deps to borrowers and they they're dispensing uh values positive values to those who provide liquidity so if we would discuss same asset instant collection it's basically drug pool and uh options and probably uh Insurance protocols so whenever Oracle decides that all the insurance sellers or options sellers should be charged they will be charged with one transaction so this happened uh instant it's collection proportional from every participant and it's in the same method which is the nominal so one more thing here is same asset gradual collection uh this is I have slide about this it's here it's uh about uh money streams or subscriptions so the basic idea here is that people can have subscriptions which is actually linear streams among wallets and uh what's the most important part of the protocol that each account here has some speed of incoming and outgoing streams this allows to compute your next balance without enumerating all your streams like you could run Netflix on blockchain right now without like a waiting for thousand uh TPS hundreds of thousands TPS so you you can have subscriptions of millions of users where where you don't need to enumerate all those streams so collecting assets from millions of streams is the same cost as collecting from 10 streams or thousand streams or one stream so let's get back to classification again so yeah you can see that different asset is not very popular uh Direction so I I see here only different asset gradual dispensing that's basically farming and uh project synthetics they were like the first project who started uh doing this even without Smart contract initially yeah and then I with my colleague Michael uh implemented farming smart contract for them uh I'm not sure that other algorithms are not existing I'm not aware of all defi algorithms of all defi protocols probably some of them could be here uh so what I'm pointing to that all those protocols they have algorithms which allows them to involve potentially infinite amount of participants with uh constant gas costs this means that those protocols they are already ready for blockchain adoption so you see that uh Dy is like a Future Finance but it's not yet there always because we have lack of computation uh computational uh uh like in system our gas is limited so we just can't enumerate millions of participants and do random algorithms like which are possible in traditional Finance but all those protocols they are kind of ready already they're ready for adoption uh and uh that's why they are getting uh a lot of traction in dii so whenever one operation involves uh huge amount of users like thousands millions if you would implement it with one complexity with one of those algorithms you are safe it would work it would work uh and it is already working for most of the projects but if you are trying to do something else which with higher complexity uh like you have much much less chances that this would work so what I also propose to uh discuss protocols and architectures in terms of those algorithms because this would give a more clear explanation how exactly uh your algorithm Works how your protocol Works how it engag multiple of users and uh yeah what's the economy behind the protocol uh uh that's probably it let me check if I have uh more slides oh yeah okay I forgot about this so yeah if we would look at instant distributions they have one uh common issue or they are all event based and this means that they are all triggered by transaction this also means that they are potentially vulnerable to uh front running attacks and sandwich attacks so what happens with uh for example opium protocol funds are locked in the pool until Oracle have a decision should they pay or not that's fine Lock Works and uh it protects from uh front running attacks sandwich attacks but for example for amm it's not true they're not protected and that's what happening you see there are for for example there are three different pools involving usdc usdt and d and someone have 3 millions of those assets so they are injecting their liquidity right before swaps why what they are achieving they are achieving it's like virtually having six millions of liquidity by really having only 3 Millions so you see it's like Unis version two or something like that if you will have uh liquidity 2 millions in each pool before huge swap you will have like virtually six six millions but if you will have more assets involved this coefficient is even more crazy like if you would have 10 different assets you will have much more different pools and you will have much better coefficient of liquidity like 10 different assets 10 Millions will probably turn into hundreds of millions of virtual liquidity and this is uh already exploited by uh AKA jit liquidity Bots they provide liquidity just in time whenever they see huge swap which they want to get their fees they inject liquidity and then they withdraw it back so potentially it's possible to avoid this to protect uh those liquidity provider providers who are not cheating like that but you know they are improving uh exchange rate for users so that's good but they're making uh liquidity providers less efficient those who do not use this strategy they are less efficient than those who use this strategy so yeah this could be leveraged if we will have delayed deposits or delayed withdrawals or better deposits yeah so technically it's possible to use something but the problem is here and it requires something to be resolved probably we can ignore it and we will have two sets of liquidity providers those who provide it constantly and those who provide it uh like a jit so yeah that's is not an issue for gradual distributions for example here uh when we have money market it's not when you return your depth this uh percentages are being distributed among those who provide liquidity right now it works differently your depth is growing every second and every second your all depths of all borrowers are being redistributed among lenders so that's how it works your depth is growing every second and collateral of everyone is growing every second so there is no any room for front running or sandwiching here because it's gradual it's not event based or instant distribution so let me check if I have more slides no so this means that we have almost 10 minutes for questions thank you thank you so much um so take a seat we got 15 minutes of questions and I'm really going to need your guys' help so um if we could switch to the slido view um that's that would be really really great and please get your phones out I think most of you that have been here know what you need to do at this point um and yeah we can just talk about some questions um my biggest question right now is um we're going to do the Q&A not this one uh the um what's next you all of that stuff super cool classification completely understand but what where do you think that maybe the industry is going to go next and maybe also talk about what you guys are also going to do next uh I had a recently discussion with my colleague Michael and uh I I think that's reasonable to think that the very first inventions in defi which were like they're pretty obvious right now it's am money markets um like other protocols which are were also listed here uh they were invented because it was like the very first things to be invented they were not actually existing in this form in traditional Finance in this kind of algorithms they were not existing definitely because there was no any requirement to optimize this High to have no computational complexity at all to have constant uh complexity and I think that we will see more cool protocols involving these kinds of algorithms maybe we will see something new but I'm not sure I try to kind of classify and analyze it from different sides I'm not sure that we will see anything different but we can see new applications of uh these things they are also called accounting algorithms and I see that there is like a huge room for Innovations in this area so yeah probably we will see something new for options like approach I mean we saw a lot of implementations of options which were kind of uh highly inspired from traditional finance options but we could see something different which provides something really similar close to original options but it's different thing and it's more optimized for execution on blockchain and have own advantages okay cool there's a lot there there's a lot there and what are you what are you what kind of what's one what's next for one in oh uh we are currently exploring different opportunities we are uh having research and development phase for multiple different products we're working in direction of uh cross chain actions for users we are working on intents that's very interesting about intents you know everyone is talking about intents but the thing is that I don't see if anyone trying to summarize like what are the features what should be done by uh this intents what should not be done what are restrictions and that's something what everyone should decide for themselves it's like optionality yeah mostly people understand any signature based actions that it's actually intense but yeah probably there should be more things like there should be more things which how you can Define restrictions or trads of your intents you could have time based trads or anything else based trades you should have invalidation invalidation per per selector or per action or Global per user or something like that I mean there are many things and no one is clearly defining it because everyone who came up to intense protocol is building own thing the only common part is that it's uh signature based so yeah but signatures are pretty hard you see yesterday there was announcement in Twitter that next ethereum hard Fork is going to have ethereum Improvement proposal 3074 that's something which would allow smart contracts to behave on behalf of your wallet with just your signature a form of account abstraction in a way or yeah you know this thing means that previous approaches for account abstraction they are not not going to make it so this earc that's a statement I thinkc 4337 it's like capitulation I mean there is no any need to use it anymore if this thing will be included in next hardw this is a like game changer for wallets because you will be able to execute arbitrary logic from your wallet smart contract will execute it for you but yeah the same thing Audits and uh you can make Atomic things from your wallet that's uh pretty interesting cool thanks for sharing your thoughts there uh we're going to transition to the audience question so please continue up voting so we're just going to go top to bottom uh so the first one is uh what do you see as the main Revenue driver of 1 in in the future um I'm we got a lot to get through so we're looking for short short answers yeah I I'm mostly focused on other things uh like on potential growth and uh I can share my vision of the market uh I see that defi is a missing piece of Finance world and how Finance world was built without this layer of blockchain and defi is kind of a mistake but there was no any other way to have it there was Zero chances that we would first have blockchains and then we invent all this uh economical and financial Primitives yeah it's obsolete at this point yeah yeah it's uh Legacy actually yeah and uh what I tried to say with my presentation is that defi is not something which is coming in like 5 10 years it's already here and most of the protocols which were in this table they are ready for high scale like they are ready for millions or billions of users even if blockchains layer ones and layer twos are not ready for this they are already ready because they involving millions of participants in single transaction they are all like their balance grows or drops but all these things happens on whenever someone can make transaction all this happens without transactions in gradual approach I mean uh this thing like money market it's already working and right now thousands or millions of people can borrow from thousands or millions of users and it's already working it can work uh we don't need like billion of guas to spend this thing we need only guess when people do some actions but when they are not doing anything this algorithm allows them to get this money streaming depth streaming to money uh and this happens via mathematics and soon a lot of privacy based off of the last 24 hours of what's been going on here so coming soon with privacy um next question is do you stand with Unis swap unicorn Emoji uh uh it's new information for me that I should put Emoji somewhere okay we should put Emoji of unicorn slac yeah or something like that this would be even more promising yeah I think that's interesting thing what's happening but probably there's no any other way to get high level of uh confirmation of this defi being accepted worldwide without this this thing what is happening so probably this is one is happening for people that don't know what's happening to to Unis I mean that's probably legal stuff or yeah yeah I mean that's probably one of the steps which is required for the future yeah the industry is growing up there's going to be a lot of regulatory tomorrow morning here um any thoughts about virtual autom market makers VM that's interesting I'm not sure how much they're different from automated market makers uh is it about they have a narrative problem I I missed this thing yeah virtual I'm not sure what exactly it is but I think it's something like amm but with some restrictions like uh concentrated liquidity or other things like Unis swap V3 V4 like curve balancer I'm not sure but whoever whoever asked that question can go find Anon afterwards and have a discussion about that I think uh the last question and then we're going to move on to the next panel is when one inch on salana Wen without the H there are many cool blockchains out there where one in is not presented yet I would say we are constantly working on this we are looking at different chains and considering like opportunities for sure right now we are on evm chains because uh it's much more easier to scale among those chains and support them uh development for other Chains would automatically mean like months of work and also uh this access to nodes and stability and other things it's uh like a different level so no one said it was easy yeah but yeah you see also like tone blockchain iser in and uh it's uh who who here someone said this earlier in the outside who here is a fan of Tron not Tron that the answer is no uh ton telegram who his knows about what plans they have what they're coming out with no really okay well pay more attention I think not to I'm not I'm not Shilling them I'm just saying yeah I'm also not Shilling I'm just curious what I knew recently I mean uh that's the blockchain which have sharding already and they force uh blockchain defy developers to sh their smart contracts by just restricting uh storage amount so storage in tone blockchain for smart contract is very limited it's like few hundred slots so you can't even have a token smart contract because you can store balance of multiple users so the way they propose you to develop software is make Charing of your smart contract it's like chared architecture for smart contracts chared per user I had a presentation about this four years ago that's cool story actually I had presentation about sharted smart contract architecture uh if you would try to find this on YouTube you will find it immediately and uh two days ago I discovered that tone fible token standard called jet tons is based on uh my presentation they have this in their uh enchantment proposal they have this link that was unexpected and fun that thing which I came up to for chared architectures of smart contracts with a non-interactive checks uh for calls among those contracts sharded logic is it's being used actually in sharded blockchain where they force people to Shard their smart contracts the idea behind the sharding is that they rebalance sharts they move different smart contracts among sharts to rebalance computational complexity of the blockchain that's interesting approach I mean it's like one of the few blockchains with shards which is actually working right now I'm not sure that it will be possible to build build such a huge defy thing which appeared on evm already because you know evm is atomic and atomic is much more easier for development sharded blockchains they have don't have atomicity whenever you call any other smart contracts it's it's a synchronous call it's like a separate transaction and whenever it gets finished you can get a call back but if you send some tokens and then you change it your mind it's not possible anymore because those assets could be spent on this shart or something like that so I'm not sure what exactly is possible to build but it's uh something is possible to build people build tokens people build amm I will be curious to see what else they can build in this highly different environment we'll find out I guess maybe some AI agents thrown in for some fun um thank you so much please give a round of applause but you stay set down stay sat down [Music]