The Path to Decentralized Custody

Transcript

testing hello hello hi Bonjour mom job with you my name is Zen I'm one of the co-founders for Taurus and I'm glad to be here in the Paris talking to you guys today this talk is about let me just close the notification so you guys don't have to see that this talk is about the path to oh [ __ ] okay disaster averted or maybe not okay this talk is about a path to decentralized custody and again

I'm Zen from Taurus Bonjour everybody um so who here's a DAP developer who here develops tabs primarily okay I see a couple of hands alright and if for those of you guys and also you guys who have interacted with a decentralized application in recent months I think you guys might have seen that the onboarding experience for applications has become somewhat like this it's a little bit messy there's a little over set back there tons of wallet providers providing wallets which pledged for a better UI UX but this is not really the experience that we've all been gunning before today for this with this talk I hope to come I hope to

convey how Taurus is approaching the key management problem its path to decentralize key generation as well as maybe convince some of you in the crowd to use us over some of these competitors so for those who are new to Taurus the best way to experience Taurus and the wallet itself is to actually just go to app toward us and check it out for yourselves you guys can you guys would be able to sign in Google and the app basically allows you to log in via olaf's so this is includes like Google Facebook reddit discord or any one of these logins to get a public/private key pair which you can then use to interact with the thier own

blockchain or any blockchain for that matter and really the front end is just sugarcoating over what's happening on the back end how we do key match when it's a little bit different from some from from conventional ways who here knows of Shamir's secret sharing by any chance okay I see some hands yeah several of you guys for those who don't it's basically a polynomial interpolation so for example if you have a line right and you have one point on that line you can't define the line right you need two points so you need a and B this is true for higher powers of X so for x squared you need three points for X cubed you need four points like a

curve you need three points basically engineers secret sharing the y-intercept is always the public/private key pair and here we have a to our three Shamir's secret sharing because of any of the two points of a B and C you can always reconstruct the function and get back with the y intercept which is your public private key pair what we do on the backend futurus is we basically do this n times so each each node on the network creates a random polynomial and you basically sum this all up to get a final polynomial a master public/private key pair with master shares on the network and the result is each of these

nodes we take we basically take this we basically take this we basically take there we go okay so yeah so we're facing about technical difficulties there so we basically take these points that we've just defined on the line and we distribute them across a set of notes on on the network so these are run by by Nance a third name service ethos can tenements kale mattock silica and us as well when you log in via apt order us by Google or something like that

you take your off token so you take you talk to the Authenticator which is Google in this case you get your auth token via the iframe and relay it to the nodes on the backend each individually verifying that you are who you say you are - against Google itself and this is the case this is generalizable so this is the case for like Facebook reddit or it could be even a signature keypad and a reason and you can build on top of this so that's a default seed key but what you can build is for example a password manager which we've also done

for you can check out the implementation and alpha told on us this basically allows a user to add an additional salt via a password you hash the salt with the key or you encrypt the key with the salt so that the key distributed and stored and network itself is an encrypted key and when you login you retrieve that key do the decryption only in the front-end and use that key we've also built out gasless transactions via smart controls which people are free to use honestly we never really wanted to do this we don't consider ourselves smart contract developers I mean smart contract wallet developers but we weeded out necessity to supply some of the gaps

that have already integrated us these gasless transactions and our philosophy over here has always been to extend extensibility to all smart contract wallets so basically what we're doing is we're ensuring our front and it's integrable so our control worlds be they from argent which is ours as a fork of or theorem gnosis safe whatever might be so that was the fast form of integration which is the Taurus iframe integration you can do so via script tag and all of that but there's another form of integration in which people can actually talk to the notes directly on the backend you you could you can imagine our iframe currently as just being one of these applications which basically

talked you which basically have a verification or execution script which is on the BFG layer of the network and this script is run every time you want to retrieve a shimmy as you could share for note so the tours iframe basically has these verifies the google verify the Facebook verified the discord verify the Twitter verify and so goes on but and any application can do the same so for example we we have some applications of pipeline good Dalek I Bob which are doing this form of integration so you basically create the experience is basically you create a Google client ID you just create a Google OAuth flow and you relay that to notes on the backend

you are then able to collect the shares and reconstruct the key in your own front end this implementation has several implications though there are positive implications and that it's immediately web and mobile friendly also adapts get literally full control of the user experience there's no longer and in pop up flows no longer any basically the consent flow is derive by the tap itself there are there are also issues with but there is a problem in that users don't get the same key so for example with this particular flow it's very similar

to the universal logins flow in that each tap gets each daps user it gets a specific key according to that down so for example Bob at toward on us which logs in via the Taurus iframe is different from we'll get a different key when he or she logs on to for example Kaiba over here but within it's specific and it's pertain to the same key and we really like this integration moving forward so we've been out for about a year and a half at this point and we launched out of beta last month there's no two ways about it the Taurus Network currently is a permission network its federated we basically the

notes operating you saw just now and the next step is obviously moving towards further decentralization initially when we started the project we really wanted to make it a fully decentralized project but we decided not to go to tokin route instead try to build decided to build kind of in towards incremental steps of providing a product which has actual value so the first version is what we have today and you can see that app that toward us or integrated into one hundred hundred of applications in app space today but what we want to do is build a second version of this there are some obstacles in the way the first one is obvious because we use Shamir's secret

sharing we have a different set of assumptions from what what somebody would build if they were trying to build let's say a theorem or Bitcoin this is because these jamia you could choose can't be revealed can't be committed to on can be committed to Unchained but can't be revealed and it's basically the underlying assumptions are threshold assumptions there are also other problems such as pinning it to rewards incentives for example in Bitcoin when for for example a Bitcoin the security of the network is directly pegged to its incentives because bitcoin is just a currency and its miners are rewarded in its currency but and that kind of makes it such that if the currency's value

goes up network value in the network is goes up and therefore rewards also go up this is different in the theorem the theorem has that concept as well with efj but they also have a generalized execution context which allows for other things like for example if you put your house on a theorem that value can't that value doesn't increase the security of the network at all that token just yeah it's it's not possible to quantify and that's the same with public/private key pens we cannot quantify the value that's held on public private key pairs well what we can do is realistically try to make it very very hard time we also feel that custody in general is here to

stay this is because custody is a great service that we have in the world there's a reason why we don't keep gold in our sock drawers instead we keep it in banks and stuff and even I mean let's say you could run your own SMTP server but most people choose to go with a front-end client protonmail Google or whatever that might be this is Shawn and crypto ecosystem as well in a way in a way for example most mainstream users I'm sure not many people in this room but more mainstream users keep a lot of their funds on centralized exchanges or basically custody solutions like coin based custody or something like that so our model of choosing to potentially

decentralize this is we are making it user focused user chosen so basically users anybody would be able to run a tourist node they would be able to charge and stake the amount they would like to charge or stake for keeping somebody's share in return users choose who they want to keep their shares and park their shares with if you are comfortable with a hundred notes that's great if you're comfortable of five notes that's great too and it might be a mixture of notes which are relevant to you for example JPMorgan might write a note maybe Google might run a note as well as maybe your mom might run a note and you might choose you might to choose to park your shares of all three of them for a and your mom won't charge because

that's free but yeah I think that's about it from me for now do talk to us if you're keen on the I both iframe integration or the direct tourist network integration we would love to help to get you set up with that if you adapt we can really help with the OAuth logins or any particular off that you want including your own and we're also looking for people to help us build this if you're keen on potentially exploring the space let me know I would love to open the floor to any questions that people might have yeah yep yes exactly it's very similar to that

the staking mechanism is very similar to that in that Inn and we don't really have a token right now but in the next version this this it's basically show miss you could chairs are just points or that you saw on that curve right and there are ways to reshare a show means you could share so it's called proactive secret sharing where you basically reshare the whole key from one thing to a new polynomial and you recreate the polynomial what you can basically do is on every login you require an attestation on that Association there and that that that proves a login and if they if that thing doesn't exist and iShares revealed a commitment of the

shares revealed on the network then you can slash and that's how it works yeah any other questions that I might be able to answer okay my time is up thank you very much guys [Applause]