zkSessions: DeFi & Privacy - Snapps: DeFi's (private) key to mass adoption

Transcript

all right um cool so um hi everyone i'm from mina um because we have limited time i'll spare everyone in the intro and only do it like about mina only do it after the actual content i want to make it less less marketing um but um zooming out right what is what is the define landscape like and this is very high level um you know approximation with many inaccuracies but i think it gets the point across so once we you know look at any likely by structure at the bottom here is like blockchains starting with ethereum and many others didn't want to get political so you know lots of competing chains other than ethereum there's all these liquidity protocols um we just had um you know avion and then

um there's end user applications and sometimes these are vertically integrated right the the protocol ends up being the end user application itself but sometimes they're separate and then we have a bunch of users um so if we look at the current landscape of d5 this all works um by putting many dollars into the system um usually in the form of collateral right um and the one issue this this works right we've i mean obviously everyone's aware of how big defense got but the one issue that disgraces that all the users have to be rich like they have to all have a lot of collateral available to post and to get access to the system um so what can we do if we don't want to you know only limit this to

rich users um because that's not the way the normal world works um and normal world the the non-blockchain world has figured out a way right um it's by sharing a lot of personally identifiable information whether it's in the form of your um you know social security number here in the us or it's equivalence elsewhere or a lot of other otherwise personally identified information which which works um not everyone has to be rich but once applied to this defile landscape as it is it would create a lot of issues where like this all trickles down and you know because blockchains are public forever there's lots of pii at the blockchain level and obviously this won't work and that's why

we're stuck in this over collateralized world right now so what can we do we have to put some locks some privacy locks somewhere along this chain to make the system work and so for example we could put it on the between the end user apps and liquidity protocols right in fact one can argue that this is already being done for example with the likes of coinbases and binances of the world you know to the extent they start providing information or access to d5 protocols that's what they're going to be doing but obviously that's not the end world we want to live in we could put it between the liquidity protocols and the blockchains themselves but then the issue is that all this chain of events and applications and services providers up here

end up being you know pretty much like the existing world like we have a lot of huge companies and apps and businesses that have a lot of pii and it's not in fact too different from the non-blockchain world and how things work so what if we could put this privacy layer between the user themselves and the applications and everything to look below it right and as i think vitalik was the first one to say this but data's become a liability right it's no longer an asset it is a liability um so i'm sure many applications or protocols don't even want to have access to this much pii anymore um so we are actually figuring out a way to do this at mina we call them snaps after snarkified

applications um and what is a snap it's actually pretty similar to adapt but it acts two very crucial pieces of features number one is that it can prove facts about a user's data without having to share the data itself and this it does this through the magic of zero knowledge um and then number two is um what we're or i'm calling universal trustless oracles um which we're building this nifty feature by being able to prove any https payload inside of a zero knowledge proof by allowing that allows anyone any single user individual using their own computer without having to rely to relay any any trusted centralized oracles to prove any any fact about any data that's carried

over from an https payload from the existing web um so you know this this can unleash and unravel a bunch of interesting use cases i shared some examples here this is just like very just skimming the top right now um but yeah so we believe snaps can actually enable existing defy applications with these features that allow the privacy to stay or the data to stay with the users and for that data to not be a liability to further down the supply chain um so that's actually the main content maybe i have one minute left just to skim real fast through who we are and what we do and yeah please um you know if anyone has questions um contact me or hop on to them you know discord over

here pretty busy um we're about to launch also our main net uh which which takes me to what mina is mina is actually a new layer one blockchain um it's the world's latest blockchain um and why is it the world's largest blockchain um mina actually uses xero knowledge proofs in its very core that the exis the entire chain is consistent you know built on zero spruce and we use recursive knowledge proofs to fix the size of the entire blockchain um so it's somewhere between 13 and 22 kilobytes we're doing our latest calculations for me that's launched um and um you know there's lots more details but i'll spare everyone in the details and as i mentioned on top of the blockchain

link and fix size small tiny blockchain one can also run snaps on top of this so you hope to bring these privacy features to all the defy apps around town yeah i think that's my slides um i'll pause there and see anna how you want to do this yeah so i would say like if it's okay for folks i think we will be able to get the hop-in working again in a shortly i'm actually speaking with the folks there um i didn't expect them to come back so quick so what i would suggest is in the chat on hopin do you want to leave any questions that you might have and if anyone has any questions there's one question

great that was awesome um can you please expand on the limits of what you can prove about users as data great question um so broadly this is about um how complex the snark construction can get right um and although i'm not the right person to speak about that because i'm in charge of the business stuff over here um you cannot for example like prove super sophisticated computations um like i'll just give two pieces of example one is that you can very easily prove stuff like is the user's credit score within a specific range or is there a specific number of upvotes on a reddit post you know just

random stuff but it it will be much harder to like get into some funky math and you know do all sorts of complex computations so hopefully that helps explain a bit okay there's one more question any more details on the snap design a publication or a more mature idea someone posted a link i will post a more specific link that goes into the snap section um but beyond that we're we're soon going to be releasing a bunch of uh demos about this actually uh we're working on a very cool one with another d5 developer um that that should be out soon um but also we're like laser focused on our main net

launch as well right now which is coming in a few weeks which is a whole another big endeavor um and right after that um the entire team is going to be focused on releasing tooling and libraries for snaps but anyone have ideas questions want to collaborate on stuff please reach out to us