How we think about storing crypto is broken

Transcript

great so I'm here to talk to you guys today this afternoon about why I think crypto is broken and what exactly we need to do to fix it see if I can get the remote so one of the biggest issues that I believe we have in the broader crypto space is we don't have the users we've had a lot of interest and innovations in crypto and I think that the press has a few opinions as to why they think that is slow expensive transactions etc etc but I don't actually believe that's the case I believe that the primary reason that we don't have users is that crypto is dangerous to use for most folks so

imagine that you want to set up your wallet for the first time on your smartphone or using meta masks what are you faced with but tons and tons of warnings about how you're gonna lose your keys if you don't write down your seed phrase super technical super scary for everyday people to get set up with the result is that when people encounter crypto for the first time typically it's on exchanges and we all know how dangerous that is all these are exchanges where people lost millions and billions of dollars worth of crypto in fact there's just a bloomberg article this morning estimating that the number of the amount of crypto is around nine point eight billion u.s. that's been lost due to private keys and exchanges

being lost so the question is how does this actually shape who is using crypto and how it's used well the biggest thing is people don't use crypto for useful things they don't want to spend it apart from the Bitcoin pizza there's not that many examples of people actually using crypto to buy real things in the real world moreover a lot of our focus has been on TPS the last few years so it's all about transactions per second how fast can we make the network how can we make ourselves like Visa tomorrow but I would contend that the problem is actually quite a bit different than that so today I believe that we're very closely approaching the chasm and the

chasm is the point at which we're going to jump from early adopters that's all of us in this room and people who are building and defy and encrypt oh to the early majority and my really big concern is that when we make that leap are most of the early majority going to be living on exchanges not actually holding them their keys themselves not actually having any of the self sovereign properties that we're all building so we posited the question about a year ago what if crypt it was more like cash now it may seem like an anachronism to propose that cash is old-fashioned there's a lot of limits to it you can't send it digitally but there are a ton of benefits to cash as well cash is understood by billions of people of all ages traffic cash transactions are final

so when I spend money and I give it to you I can't go back behind the scenes and pull it back they're effectively zero costs although there are there are costs to managing cash itself and along with that they're largely anonymous so we don't have to worry about things like snarks immediately lastly if you destroy cash it's gone and everybody realizes that unlike a private key on a hard drive or a piece of paper if I go and burn a 100 euro note out in the street everybody knows that I've just destroyed the value the other part is cash is really big even for just the US there's 1.7 trillion dollars in circulation

across the world despite the fact that cash transactions are slowly dropping over time over 50 percent of transactions are still done using cash so to understand if we want to make crypto like cash we have to understand exactly what cash is it's distinct from money it is a part of the money supply along with what we hold in our accounts but cash is something unique cash is based on trust it's basically a transferable debt there have been other experiments previously and how you make crypto like cash the simplest one is just printing a private key on a piece of paper it's really easy and it's really cheap I can take that private key

and I can hand it to somebody else they can then go pull the crypto off of it but of course I can't circulate that the problem with printing a private key is that the next person who has it doesn't know if the crypto is stored on it or if the person is going to steal the crypto in the future the next idea is to use hardware wallets hardware wallets are great except they're really expensive and more importantly they have tons of software dependencies built into them so they all have layers and layers of firmware which you have to trust when you hand off a hardware wallet to somebody else even if you're using it like cash and the latest iteration of this is something called a Java card now a Java card is interesting because as the name implies Java cards all run a miniature

java operating system and they date back about 20 years ago a lot of the credit cards that you have in your wallets are actually running a Java card operating system or one based off of it the problem with Java cards is there just like hardware wallets and that they have a bunch of software dependencies that you have to trust so if I hand you a Java card and I say hey trust this it's loaded with crypto you also have to trust that I haven't broken the firmware of that card and I'm not gonna hack it at a later point in time so really what this boils down to is that trust in crypto cash is trust in key storage we have to have really really good key storage and we think we've come up with a proposal to at least tackle this it's called kkong kkong is a physical crypto

cache note it looks like and it feels like cache so let's dive into the properties of it Kong has a printed face value just like cache it has a time log on it which is mediated by a smart contract so Kong itself is not a hardware wallet it can't actually spend directly the crypto associated with it instead it proxies through a smart contract which acts as the smart contract wallet so what the Kong secure chip does is it's I it presents signatures and the smart contract sees whether or not those are valid and the time lock means that unlike a prepaid gift card where I can immediately separate the digital value from the

physical we can lock the two up for some period of time so it actually circulates of kit as cash the original cog notes that we've issued they expire in October of 2022 that means that nobody can spend the token off of them until that point in time they can just use them like cash so they can spend them to buy things they can stuff it in a safe or under their mattress or they can burn it but that's the only choices that they have the last piece of this that I'll get into a little bit more and how we achieve these security properties is called a secure element so this all may sound somewhat familiar because it is this is kind of similar to how cash used to operate in a lot of places a hundred years ago so these are old forms of the US dollar

where you could actually take the note to your central bank and you get silver metal out in this case one in five dollars worth of metal the only difference is that instead of getting metal out of the the Cong note you're getting crypto and you don't have to go to any party to do that it's totally on non-custodial when you hold the note you can claim the crypto so the secure element chip is a really unique kind of chip which gives us these properties unfortunately the term secure element is kind of a broad term we're really focused on a few specific properties that must be in place the first one is that the chip must self generate its key pair so we the creators of the con cash

even though we've manufactured it cannot go and ascertain what that key pair is after the fact the other part is it obviously never reveals that private key to anyone so unlike a hardware wallet I can't extract the private key after the fact and perhaps what we argue is the most important part is they are non programmable now what does that really mean it means that I can't just write an app or some set of firm form firmware and load it onto the chip but rather the chip is programmed in hardware from the factory so as a result I can't change the logic on it I can't modify the firmware it only does a very limited number of functions in a very specific way there's only one key problem with the few chips on the market today that

exist like this they use the wrong curve so in Bitcoin and aetherium we use k1 however all these chips which were designed for SSL and against government standards use P 256 but we can solve that problem the way that Kong works is we have the secure element chip on the Kong note and that in turn talks to a smart contract the smart contract contains the tokens which are associated with the note and those could be anything from Kong to aetherium to whatever token you desire and it also allows us to translate that P 256 key so the Kong smart contract has the

capability because we have a full-fledged smart contracting platform at aetherium to interpret the elliptic curve correctly compare the public key and compare the signatures and then all of a sudden were able to release the token on demand we see this actually is opening up an entirely broad new class of physical digital devices which were calling silos or silicon lock contracts you could use this for crypto identity cards crypto art or asset tracking anything where you want really really really good security guarantees about where the keys are stored and how they're stored and you want to associate say an NF T or an identity token with someone but this leads to the next point

which is if we think that hardware and our device is important and key stores important we need to think at an even lower level on how to secure these things and one of our big pushes is saying that we need open silicon to achieve this so here's one of my favorite examples of a chip that is different than it should be so this is at a microscopic level inside Samsung chip I believe from 2011 and here the person who's designed all the logic gates on the chip actually inserted this art onto the die of the chip so this is looking at multiple micron view very very small no one would ever know this it was actually a security firm that

found this in their that designer could have just as easily put in backdoors where they leaked the private key material we have another project that we're working on which is called arcs and the idea behind arcs is how can we quantify the trustworthiness of different ships using crypto economic means the project is pretty early but the idea is that chips are really important for us to store private keys and so we should worry about how they're made what designs they're based on and where they've gone along the supply chain so this is really not just the starting point for crypto but the entire distributed web when you start to think about what Google and Facebook are

really doing they're arbiters of our keys and they're arbiters of all of our encrypted material this brings me to the thought that we're just on the verge in terms of time of requiring these things at mass scale so in 2009 we saw a massive collapse of the financial system which really created fertile ground for things like Bitcoin to emerge where people lost trust in the financial system last year we saw the exact same thing happen with big tech like Google and Facebook where all of a sudden people are questioning putting their data on to these services because they don't know what happens to it or it's manipulated however the problem is that unless we give people a really easy user experience to hold their own key

material will go right straight back to the problem that we have today where they'll delegate all of that to exchanges and centralized parties so to recap what we've created is something that's usable self sovereign gives people control over their own keys to achieve this we're using a very special kind of chip called a secure element and we believe ultimately we need open silicon versions of these chips that we can audit and have crypto economic incentives backing them if you want to get started and take a look at this we released our white paper which talks both about Kong cash but also more in detail about how we're actually

escrowing the funds against a given Kong note you can take a look at the escrow contract on our github and then lastly we have a bit of a community in key base if you want to ask about using hardware for your own projects if you're interested in the COG note definitely come up and see me if you'd like to see it in person and take a look at it thanks a lot and I'd be happy to take some questions hello thank you very much for the talk that's really nice I'm curious like what happens when the expiration date of the

the bill it is Richard like after 2022 it's as you told how can the the do the person that is in possession of the the bill actually convert it and like to actually transfer to some wallet he or she controls yeah so one of the things that I didn't demonstrate with the note itself is every note actually has built-in NFC capability so you can use your Android or iOS app directly with a note and tap on it and that allows you throughout the lifespan of the note a to verify its authenticity because you can generate signatures based on the private key and be ultimately at that claim date you can using the current block hash

generate a signature that allows you to claim the current funds so the reason we use the current block hash is because if somebody had pre signed some data historically and that was sufficient then they could steal the funds off of it so we want to know that you've held it very very recently other questions anyone don't be shy I'm just kind of curious if you're gonna do open fab you know who you're gonna

have fab your chips because there's not really much choice in the world and nobody really trusts any of the existing players anymore yeah so the fabrication of chips is a fascinating and terrifying world that we believe needs to be totally opened up so the the biggest problem there is to begin with at the toolchain side so we finally are starting to get open tool chains so we can audit designing chips but then on the fab side the standard sell libraries that all the fabs users still closed and we're hoping that some of those parties start to open that up so that you can choose at a different process note an open source thing but yeah open fabs is is really the next big ultimate problem which is even if we have open designs

for chips we want full auditing and I think that even with ambitious projects like this and and other projects that are working on much more sophisticated processors it's probably another decade out before we get those but fingers crossed awesome thank you so much everyone thank you [Applause] great great