Getting Bridges back on Track

[00:00:00]Hi, this is Charles Oskinson broadcasting live from warm, sunny Colorado. Always warm, always sunny.

[00:00:07]Sometimes Colorado. Although I wish one of these days I can go up to Wyoming and just be there forever because I love Wyoming.

[00:00:17]Okay, so yesterday some piece of [ __ ] vile lowife scumbag went and stole a bunch of money on the one chain Cardano Binance Smart Chain bridge and uh it's opened up a broader conversation of why does this keep happening and I figured you know about time to make a whiteboard video and talk about how do we fix the root cause of the problem. You know, if you look at the last 36 months, the number one loser of funds is typically bridge hacks and they're catastrophic and damaging. And a lot of people really don't understand.

[00:00:55]Well, what does that mean? Does it mean there's a flaw in the cryptocurrency?

[00:00:58]Does it mean that people don't know what they're doing? I mean, what's what's the story here? Why why do these bridge hacks happen? So, I wanted to make a whiteboard video and talk a bit about what is a bridge and what you can do to remedy it and actually talk about why midnight is the solution to all of this.

[00:01:17]It's kind of ironic, you know, uh, midnight was built to solve these types of problems and then this problem happens in the Cardano ecosystem related to a u a midnight asset. Uh, so let me go ahead and get some my infrastructure working here. Okay, there we go. Yes. Sorry, I have like five coding agents running full-time, so this computer's a little slow. All right, so first and foremost, uh I mentioned this to Agalos, and I said, you know, I'm working on this problem, and he said, you know, we just wrote a paper on it. I was like, god damn it, you guys in the research team, you're just like so elite. So this paper's public. Uh the address is right here. So iaacr.org uh 2026 2022.

[00:02:02]uh you see all the authors and this is the first time in history in our industry that someone has actually done a rigorous formal analysis and universal composition uh of a actual model checking of bit and blockchain bridges.

[00:02:15]So when you want a precise mathematical definition of what is a bridge, what keeps it secure, what's the bridge model uh and try and develop a better understanding different types of bridges that can be used uh for example uh native consensus certificate or ZK bridges or advanced threshold signatures. This is the canonical 43page densely filled with math uh definition of it. So when we talk about Cardano land using formal methods we always do our job and this is any bridge this will give you a mathematical description of the bridge. You can even see that we have small step semantics and these types of things there. So this is uh kind of a reference point and it's got all the different citations as well of everybody's ever built anything anywhere uh for bridge and a lot of appendices about various different properties and they actually prove out uh the universal composition uh setup. Okay. Now the other thing is this is very specific to Cardano and specific to midnight. So one of the things I was interested in so this is a document I authored is can you create uh a trustless ZK bridge between Cardano and Midnight?

[00:03:28]And the first question is what does a bridge do? And most people think oh I move assets.

[00:03:36]Yeah, that's one use case that you can have for a bridge, but it turns out there's 42 things you could talk about in the Cardano ecosystem about bridges.

[00:03:46]So, uh you have the chain and epic context, you have the transaction and block facts, you have the UTXO membership, uh you have the non-membership, so spent transactions, you have the output address and datams, you have the mint and burn. uh you have asset supply at each epic script outcomes stake delegation rewards deposits treasury MIS and you can even make statements about the governance actions. So there are 42 things that you can make statements about because when you talk about a bridge, you say, I have one chain or I'm a member of one chain and I want to interact with a foreign chain somewhere else and I want that chain to do something for me and I'm going to rely on it and somebody needs to inject the state of that system in and every chain is different. They're not all the same. So when you scroll down further in this document, you can actually see here uh midnight there are 52 things you can talk about with midnight. You can talk about the chain indexer anchors transaction intent validity contracting calls private contract predicates z swap shielded tokens unshielded knight dust and aggregation infrastructure statements and these are very different than how Cardano operates. Okay, so that's kind of a baseline is understanding there's a rich mathematical underpinninging. And also when you talk about bridge infrastructure, you're talking about more than just the movement of ADA or Knight or Ether or Bitcoin to a different network. Okay, so let's talk a little bit about that. So you have and let me verify my pen is working. Oh, good. I'm not using that many CPU cores.

[00:05:30]That's good. Okay, so you have Bob.

[00:05:33]There's always a Bob. You can never get rid of Bob.

[00:05:37]And Bob lives on Cardano.

[00:05:40]And then you have Alice. And Alice lives on, let's call it BNB, Binance Smart Chain.

[00:05:49]Okay. Oops. We'll put uh BNB here.

[00:05:54]And let me go ahead and erase this.

[00:05:58]All right.

[00:06:00]And this is Alice. Okay. She lives on BNB, so she's a BSC girl. There you go.

[00:06:07]All right. So, over here we have Cite, Cardono Knight, and that's issued by the smart contract.

[00:06:17]Now, already there's some people running around Twitter saying, "Well, why can't we just roll back the CIE supply and do this whole thing?" Cine is completely decentralized. It's embedded in Cardano.

[00:06:29]And part of the Cardono native asset standard is that all CNTTs follow the same decentralization principles that ADA has. So this is fully decentralized the CI you can't do anything. It's decentralized. Okay? It's already been distributed. It's in the hands of people. It's a real cryptocurrency.

[00:06:47]So, Binance said, "Hey, look, we like this glacier drop thing, but we want all of the night instead of taking custody of it on Cardano, we want it on Binance Smart Chain." So there is a bridge that's been around for a while and it's called Juan Chain uh Chinese bridge and they've been in the ecosystem for a few years and they've received catalyst funds uh from 2024 and beyond uh to build bridges for Cardano many places.

[00:07:16]So one chain provided a bridge to be able to move Cardano native tokens from Cardano to BSC.

[00:07:26]Now, there's many different ways to do these bridges, but generally speaking, you can do a multi-ig.

[00:07:33]Uh you can do a ZK bridge.

[00:07:36]Uh you can do threshold signatures.

[00:07:40]You can do all kinds of things. But the basic idea is that you'll take these tokens CIET and you'll lock them.

[00:07:50]And when you lock those tokens, then you get a receipt.

[00:07:58]And that lock will be connected to an address, a redeemer address on the chain that you want to go to. In this case, BSE. Okay. So you take the redeemer address and the receipt and then you go over here and this bridge will inject into the Binance Smart Chain the transaction with the receipt and the redeemer address. Okay. So it'll inject that in. Now once it's over there, what'll happen is it will create a BSC wrapped.

[00:08:46]Let me turn that off. That's a little annoying because it keeps warping my writing. A BSC wrapped uh C knight.

[00:08:55]So you're going to get a BC knight.

[00:09:01]Now, this is only valuable because there's a one to one peg between the BC knight and the knight.

[00:09:12]So, how do you go in the other direction? Well, you go in the other direction. Alice will destroy the BC knight.

[00:09:24]So, there's some sort of provable mechanism to take PC Knight out of circulation. And then she gets a receipt and the receipt is bound to a Cardano address.

[00:09:39]So, what happens is the bridge will say, "Yep, that receipt's right." And then it will unlock the corresponding amount.

[00:09:52]And then suddenly she now has Cite at uh address dot dot dot dot dot dot.

[00:10:01]Okay. So the bridge basically is in a position where it can understand both chains. So typically you have a full node and you have a full node and the bridge watches And the bridge is like this this guardian and it's there and there has to be a well-defined logic. So it's asking questions like when Bob says, "Oh, that's my scenite." Well, is it real?

[00:10:35]Well, the bridge needs to be able to replay the entire chain, search through it, and figure out yes, that's real knight. When Alice asserts that that's real BC knight that's going to convert to C knight. Well, the bridge has to be able to say yes, that's real. That looks good. Okay. Now, you have custodial and non-custodial.

[00:11:01]This is one of the distinguishments.

[00:11:05]So custodial means that you actually send it to a large pool somewhere and uh that large pool uh is under the control of a trusted third party.

[00:11:16]Non-custodial that typically means that you are the person who can unlock it on both sides.

[00:11:25]So this is how the Bitcoin DeFi side works.

[00:11:28]You take your Bitcoin and you mirror it into Cardano, but you control both sides of that equation and there's a onetoone correspondence. It's also relevant because what if Alice locks some stuff on from Cardano to Binance, she sells half of it and then she decides to come back. She's going to be getting potentially different cite than the Cite that she locked. So all kinds of things have been added into bridge design like liquidity pools to make these faster uh custody accounts all kinds of things.

[00:12:05]And this is where the attack vectors typically live. There's some usually some sort of multi-IG signature uh that exist and there's some sort of logic rules that exist about the unlocks. So you really have four vectors of attack that exist here. Okay. So you have the Cardano Plutus contract that was written onchain.

[00:12:35]You have the Binance Smart Chain uh Solidity contract that's onchain.

[00:12:44]And then you have the off-chain BSC infra.

[00:12:53]And then you have the off-chain Cardano infra.

[00:13:01]So any one of these can be hacked. They all have to work together to make this a seamless process. Why? Well, if bad things go and the indexers are wrong or the view of the blockchain's wrong, you may be able to fool the smart contracts into creating fake cite that doesn't exist or you may be able to redeem fake BC knight to uh the uh uh Cardano side.

[00:13:31]Okay. Oftent times also when you build a bridge, you also have upgradability.

[00:13:40]If you find that there's flaws or you need to add new features or new capabilities, you'll go from smart contract v1 to smart contract v2. Well, what does that mean? Well, somebody has to have admin keys, and those admin keys give you the ability to change the version. Well, if I compromise those, I could update smart contract v2 to say, "Give it all to Jim." Evil Jim. Evil. Evil Jim. He's got a little cane.

[00:14:13]Yeah. He's Jim has got that top hat.

[00:14:16]He's got a little monle. He's a classic guy. Little Jim, you know, give it to evil evil Jim. Uh, so he can update the contract basically through admin keys to just give it all to him. So bad bad things. So typically when you have hacks these are the public side and this is the private side. So the public side means that everyone in the world can see these things because they live on a blockchain. And on the private side, a restricted subset of people, usually the operators and other things can see that. But this can be compiled code and it's not necessarily exposed to the internet directly. It's behind some form of interface. So usually when there's private compromises, it's insider threat, an employee, a contractor, um a trusted third party or something like that cuz it's a or I'd say semi insider is probably air quotes. And here this is typically a black hat uh who found on vulnerability and exploited it.

[00:15:33]So where the hack occurred tells you a lot about what the nature of the hack happens to be and all of these are equally threatening.

[00:15:41]Now, if you read that formalization paper that we wrote, we cover exhaustively all the different styles of running these types of things. And when you talk about bridges, right now, this is just an asset locking, but what if you want to have a hybrid application?

[00:16:00]So you have Cardano here and you have a smart contract on Cardano and that smart contract says, "Hey, look, I'm only going to execute when Binance Smart Chain uh reaches block height XYZ.

[00:16:23]Okay, so execute when How do you know? How does a Cardano smart contract know that that has occurred in Binance?

[00:16:39]This type of transaction is a special purpose transaction. That smart contract living on Cardano, it doesn't know a lot about Binance. It's it's told a lot of stuff and facts and principles, but ultimately it has to have some sort of root of trust and uh somebody has to tell it usually as an oracle that yes, this thing occurred. So when you go multisig, it's a multisig signed message. Uh but it can be anybody and it can be anything and we'll get to that in a second. So you have a smart contract, you have a certain block height and trigger. Okay. Well, something's going to tell it that. And either that's a trusted third party, uh that's a federated trusted third party. So that's the multi-IG case. Uh there's some sort of uh bespoke threshold signature scheme or the heaviest is ZK.

[00:17:41]And this is the world that Midnight lives in. Okay, Midnight said, "Look, there's 42 properties about Cardano that people would want to talk about." 42 over here. And there's 52 properties that somebody would want to talk about with Midnight and Binance Smart Chain probably 50 plus. Okay. So what if we constructed a zero knowledge proof and we have a verifier on both sides a ZKV.

[00:18:18]So instead of Bob going through a bridge, what Bob will do is Bob will just generate a proof and he'll connect that proof PI to a redeem address on BSC.

[00:18:36]Then anyone bridge operator whatever can submit that to Binance SmartChain. Bob himself can. Okay. So then you send it over. Now when it goes it goes to a verifier contract. So the ZK verifier BSC and it gets that proof and it ingests it. It looks at it and says oh yeah uh that proof is unforgeable and it meets my understanding of how Cardono works.

[00:19:04]So what I'm going to do is mint that BC knight and I'm going to send it to that redeemer address just like that.

[00:19:17]Now you notice something. There's no bridge operator there. There's no third party there. Anybody can transmit that message. It's a trustless system.

[00:19:27]And then Alice can do the exact same. Uh she can generate a proof and send it on over. So one of the reasons we built Midnight is we wanted to have a generic programmable system that's fast and capable. So you could build these types of things and then you could create embedding contracts that are loaded with lots of facts and they have a verification system and once those are there basically you can set it and forget about it. Then the root of trust will just be on the upgradability.

[00:20:01]Okay, that's basically it. So that instead of the operator sitting there watching and facilitating the actual messaging and you send it there's escros and custody, you can do perfect mirroring where you can move assets over or one to one uh and then you can basically create verification contracts and then they would just keep them up to date by injecting current state because hard forks happen, semantics change, consensus changes like when Ethereum went from proof of work to proof of stake or us going from Prow to Laos, those have to be encoded in and you have to update your systems then okay so as we look at this that's the ideal situation and you can combine ZK with the old systems and what you can do is develop basically a multimodality bridge and this is kind of the holy grail so you can have zk plus bespoke threshold signature schemes. Okay, so you can create some bespoke certificate and you can create a ZK system and only when both agree do you allow the transaction to move through. Typically you look at these things in terms of livveness and safety.

[00:21:25]So safety is can I steal your money and livveness is can I use the service?

[00:21:30]So ideally this is your number one priority and people should never be able to steal your money and you would like to have a lot of uptime in your bridge but sometimes things don't work. Well ZK is a clientside concern. You generate the proof and usually a threshold signature scheme or something is a server side concern.

[00:21:53]It's done on your behalf.

[00:21:56]So if you combine these together in the right way, that multimodality means that if there's a flaw in the ZK system, you lose livveness. Nobody can send money, but you don't lose safety. If there's a hack on the threshold signature system, nobody can forge a ZK system. So you lose livveness, but you don't lose safety.

[00:22:18]And then what you would typically do is use a trusted execution environment and you'd have some sort of multi-party uh computation protocol for the admin update. And so you can or you can use a Samir secret sharing scheme. You can basically layer these things together.

[00:22:34]And what it does is it gives you the ability to do admin keys that are basically unhackable. No one can ever compromise an M m of n infrastructure.

[00:22:44]It just never happens inside the system if it's done correctly. Okay, so this is the operational side of bridge excellence. And if you have a powerful ZK system like Midnight, it's very possible to build one of these things in a way where for all intents and purposes, hacks just really don't happen anymore. Uh unless there's like pervasive insider threat, but this is why that is mitigable and there's all kinds of ISO standards and other things to make that really difficult to do.

[00:23:14]Okay. Now what happens when an attack occurs?

[00:23:20]Okay, so this is all just like normal operation but attack. What happens when an attack occurs?

[00:23:29]What do you do? Well, first you need a warning system.

[00:23:35]You need to alert the other bridge operators, the DeFi applications, especially lenders because what the attacker will do and they did both these things in Cardano to some degree of success is they will do two things. one, flood Dexus, and number two, lend to another asset, usually a stable coin like USDT or USDC, and then they will mix it to launder.

[00:24:12]Okay. So, if the dexes get alerted quickly, liquidity can be pulled. And if the lenders can be alerted, uh, lending pairs can be pulled. So, they can't get the exits. So, they get the assets, but then they're unable to do anything with them. If the assets are on the Binance side, the operator can break the one:1 peg. If the assets are on the Cardano side, the operator can't. You'll notice there's an asymmetry in these bridges. C knight is not the same as BC knight. BC knight is a synthetic asset that's one to one peg to Cite. The only reason it has any value is Cite backs it. So if somebody steals BC knight, the bridge operator can just say that's no longer redeemable and the C knight's okay. They have a different problem then, but it stops that. If somebody steals the C knight, it's equivalent to stealing ADA or Bitcoin or Ethereum. And so then they can actually liquidate these things and they can go down that road. But these are the two common attack vectors that exist. So an early detection system and that is a blockchainwide pub subsystem. This is something I've wanted to have for a long time. There were powers that B that stopped it from occurring and I'm very bitter about that. But it is a very high priority for Card Cardano for 2027. I think it's the eighth time I've tried to do it. It will get done after Laos. We're going to get a blockchain wide pub subsystem. Why?

[00:25:36]Because all the dexes, all the DeFi, all the wallets, all the exchanges, uh, they can subscribe to channels and those channels can be like the bridge operators channel or the Cardano Foundation's channel or intersects channel and you can create a flood message.

[00:25:59]You flood the entire network with a danger danger Will Robinson a warning message a warning warning warning a compromise has occurred and then they can know because they've been notified or even built automated systems to pull liquidity and stop things. So basically the minute something is detected uh within seconds to minutes the circuit breakers switch and it stops it. We typically call this uh circuit breakers.

[00:26:30]They build them in many trading and financial systems. Basically, stop trading, stop the operations of the systems. You can build that into a decentralized oper uh mechanic and have either a natural way or a uh a manual intervention way to turn the system back on. So this super vital to your real-time warning system inside the system. Now the other nice thing is uh bridges they work in many directions.

[00:26:59]USDCX really helps us here. If you have just pure USDC you can go straight to tornado cache or these types of things launder it mix it and run away with illotten gains. But if you have USDCX you have to go to circle and then go into the main network. So that bridge can be cut off. So that's one layer of attack that's been mitigated.

[00:27:24]Now um there's other things you could do in the detection system and depending upon the systems there, but then you have the cleanup and this is another reason we built midnight and actually um the white hat response.

[00:27:41]I forgot to mention that. This is really cool.

[00:27:44]Now you have a wallet.

[00:27:49]And this is new stuff, by the way. We're so far beyond the Ethereum people and other people. They're like living in the past, man. We're living in the future.

[00:27:57]We're They're living in the past, man.

[00:28:00]I'm Dennis Hopper and Apocalypse Now, man. They didn't pay me. They just gave me crack. I'm ready to go, man. All right. There we go. Many of you will not get the apoc apocalypse now reference because you haven't seen the movie, but you've probably seen it referenced like Family Guy or American Dad and say, "What's this reference to?" Watch it.

[00:28:18]Francis Ford Copela. It's a masterpiece.

[00:28:20]All right. So, you have a wallet and it's non-custodial. Okay. So, non-custodial. Buyer beware.

[00:28:31]Uh, you know, it's just whatever the [ __ ] happens. You lose your money, you lose your money. Bridges operate usually under the same logic. What if you could embed in the wallet's creation a white hat license?

[00:28:49]>> [sighs] >> So basically there are a series of white hat best practices and you authorize the wallet maker to delegate authority or do it themselves to be a white hat in the event that an active attack is ongoing subject to some sort of redemption process for you get your funds back. So you have a little box here and at the time of creation of wallet you can check it or you can not check it and if you don't check it your funds are excluded from a white hat. If you do check it then they'll be swept into it. It's pretty cool concept isn't it? They have never really invented this in the cryptocurrency space.

[00:29:35]Why?

[00:29:37]Well there's two big problems here.

[00:29:42]identity and nonreputability very closely related.

[00:29:54]Okay, so identity is how do I know who owns the wallet?

[00:29:59]How do I know how to give funds back to you? Non-custodial wallets traditionally do not have KYC. There's no KYC. So, how do I know it belongs to Alice or Bob or Jim or Jane? And you are who you say you are? That's a big problem. And non-reputability is how do I know you legally gave me the right to do that? I need some sort of signature from you.

[00:30:22]How do you sign something associated with the identity uh to resolve that?

[00:30:32]How about this midnight passport?

[00:30:39]Once you have a first class identity citizen with selective disclosure, what you can do with your non-custodial wallet is you can link the wallet in a trustless way and in a private way to an identity.

[00:31:00]So from the blockchain's perspective on the outside they see a bunch of garbage.

[00:31:06]Nobody knows what that is. But then an event happens and that commitment basically proves can be used as a proof that Bob owns the wallet.

[00:31:32]Super cool. And when you have a ZK system with selective disclosure, this is something you can do. And to take it a step further, that blob can also have contained within it a that signed uh thing that you've agreed to the White Hat license or any other logic for that matter. How about that? So when a bad thing happens, when you're doing the cleanup and the white hat's taken some money or they're trying to reimburse customers or an insurance policyy's triggering in, you have to unblind these things and look into it. Well, with the zero knowledge system, you never actually have to decrypt anything.

[00:32:11]Somebody can just show up, generate a proof, and show, yeah, I was the person.

[00:32:16]and they can generate it from the passport which lives in trusted execution environments on your phone disconnected from your primary wallet which lives above it. So the wallet or the bridge can be compromised, but this core credential much safer. Kind of like trusted hardware like a ledger or treasure. It's something that's much harder to break and it's individual instead of global. So even if one of them gets compromised, not all of them get compromised where usually a wallet or like SecondFly or a bridge that's a global compromise for everybody out there.

[00:32:51]So this is kind of the framework and what's really interesting cool is you'll see a lot of thirdparty solutions and they say oh I can do wallet recovery okay great or hey I can do identity okay great or yeah we got a pub you okay great or we have a slightly more efficient bridge or we can do a zk bridge or this type of thing okay great but there's no universal system that combines all of these things into one toolbox that's easy to use and programmable and the language is Typescript the most common programming language in the world. Uh compact is a derivative of that. Okay, TypeScript number one programming language and it was just so easy to use. So you can write a ZK infrastructure, you can link it to trusted execution and MPC, you can use the abstraction privacy primitives and then you can build an entire solution like this.

[00:33:49]So really the magic of midnight is it solves the bridge problem but in doing so it also solves the wallet problem and in doing so it solves the white hat problem and in doing so solves the identity problem that underpins most of this industry.

[00:34:05]I was doing an interview the other day and I said, 'You know, our biggest issue is we need to be adults, not kids.

[00:34:15]And adults have responsibilities, accountability, empathy, and adults understand that you can't just laugh at people when bad things happen to them.

[00:34:27]You live in a rules-based society. uh we live in a society where we want reciprocity and we want to be kind to each other. And so when you look at a financial operating system, the legacy system has laws.

[00:34:42]The legacy system has consumer protections.

[00:34:46]The legacy system has courts. The legacy system, so we call that dispute resolution. The legacy system has all kinds of things like freezing of assets.

[00:35:02]Uh it has law enforcement.

[00:35:05]Okay?

[00:35:07]It has checks and balances. All kinds of stuff. If crypto is going to replace the legacy system, we have to ask what makes crypto special. This you're in control.

[00:35:21]Not a third party. You are in control.

[00:35:24]But what makes the legacy system special? Ultimately, when you put all these things together, safety, somebody stole my identity and withdrew $5,000 from my bank. So, you go to your bank. Well, okay. We'll uh we'll uh credit your account. Uh somebody stole my credit card. We'll cancel those transactions and we'll issue you a new credit card, etc., etc. every day, thousands of times per day. That's a conversation people are having to some success or no success, but they at least get to have the conversation. When you live in this land, you never get to have the conversation.

[00:36:01]So, if we want to get adoption of billions of people, we have to convince them that we can match or exceed the safety of the legacy system.

[00:36:13]And by the way, we can do so with principles.

[00:36:17]We don't need to accept a financial panopticon where everything you say, do, and think and buy is tracked and traced in a untrusted third party at any time can just turn off your money. We don't want that. So this is squaring the circle and we call this web 2.5 the best of both worlds.

[00:36:37]And by the way, that is the majority area of growth of our industry the last two years. Say we want token price to go up, we want the network to grow. Well, who's growing? XRP.

[00:36:48]That's one area they live in. Who's growing BNB?

[00:36:52]Who's growing? USDC.

[00:36:55]Who's growing? Tether.

[00:36:58]Who's growing? Canton.

[00:37:01]It's the web 2.5 space. That's where they live.

[00:37:06]So, if you want to square that circle, what do you need? You need privacy.

[00:37:12]You need smart compliance. You need abstraction. And ultimately you you need to be able to link the two worlds together. Uh so that's why midnight was created. So you know the attack is the attack. We've already mostly recovered. It'll take a little while for the price to rebound uh because it was a pretty big chunk. But Midnight survived 500 million night being stolen uh which is about half the size of the Ethereum crowd sale uh the money raised to create Ethereum. So not an insubstantial amount of money and relative to liquidity, pretty big event.

[00:37:47]We survived that. The network is strong.

[00:37:49]Midnight itself was not hacked. Uh the Cardono smart contract with the Cite was not hacked. Third party bridge was. And why? Because this is a complicated thing. And no matter how hard you try in an age of fable and mythic and Kimmy K3 and these things, it's easy to break this stuff. So you got to move to a different security primitive. And we have to move to multimodality. You need some management layer and you need a ZK layer and they take each other and check and you fail on livveness not safety and we formalize that model. So hopefully next generation bridges can follow this but then when it occurs you need a warning system and then you need a legal space for white hats to operate in. So you need a white hat license and to make that actually work you need midnight passport.

[00:38:36]So we're ahead of the game. We've been thinking about this for a long time.

[00:38:40]I've been in the industry for a long time. I'm getting old. Uh and uh pretty much ancient, frankly. And I'd like to solve this because if we solve this, cryptocurrencies become real. They're not real if you say buy or beware. And no matter what you do, um you can't win some point. It's like those zombie survival games where you're in the house and the zombies are trying to get you and you survive round after round.

[00:39:10]There's no way to win that. It's just name of the game is how many rounds can you survive before the zombies get you.

[00:39:15]That's cryptocurrencies right now. I'd like to have a game you can win. And the game you can win is a game of insurance, a game of early detection, a game of checks and balances, a game of dispute resolution, a game of legal embeddings, a game of oversight, and ultimately a game of having consumer protections. And then people can decide where they want to live.

[00:39:39]What this gives everybody is choices.

[00:39:44]Notice I didn't say that we have to get rid of non-custodial wallets. In fact, they not only will still exist, but they'll be strengthened by this because the system uses as a primitive. It'll never be taken out because right now they want to take these things out for consumer protection. If you look at where Europe is going in Mika, if you look at a lot of the people on the political left in the United States, their argument is cryptocurrency should only exist within tightly regulated custodial environments. So you give all your money over to a trusted third party and they manage it on your behalf like an ETF. That's what they want. And the angle they're using is every day hacks occur. Every day uh North Korea steals people's money. Every day sanctions compliance has an issue, etc., etc. So if you can find a way to add consumer protections into a non-custodial setting, they can't ban the non-custodial route. It stays there. So this strengthens the non-custodial case and then best practices will emerge and because you can preserve privacy with consumer protections, the vast majority of people will use them and marketplaces will form that provide risk protection because insurance is very simple. You pay a subscription or a transaction fee goes into a pool and somebody on the other side provides the collateral. As long as the policy doesn't pay out, they get the premium. That's an RWA with yield and it's worth trillions of dollars. We were just in Bermuda and the BMA and these other guys, they're like, you know, how do we do wallet insurance and how do we do bridge insurance?

[00:41:18]There's like really good money for those types of things. We got to figure this out. Like how how do we make that work?

[00:41:24]If you have a non-custodial wallet like this and you can bind it to an identity, you can bind it to an insurance policy at the same time and then you can make the decision. Do I want my wallet to be recoverable?

[00:41:36]Do I want an insurance policy? Do I want a white hat protection clause? Check the box. Pay the fee. If you don't want these things, that's fine. You're basically a legacy Bitcoin account at that point. And if something happens, buyer beware. It's the equivalent of cash in your wallet.

[00:41:53]You lose your wallet, you lose your money. No one's going to come bail you out. That's your cash. It's your decision.

[00:42:01]That's the difference between me, Cardano, and the legacy financial industry. The legacy financial industry says, "You don't get to make that decision. We're going to make it for you, and we're going to make money off the decisions we make."

[00:42:18]What I tell you is you get to make that decision. But I don't want you to make a decision between Mad Max and roving pirates come and steal my money and I have to live like a 12th century peasant burying whatever gains I have so my feudal lord doesn't come and steal it from me or uh turn over everything to an untrusted third party and hope to God they're somehow honest. I want you to have a third option, which is you're in the driver's seat, but then you can build checks and balances and controls and create marketplaces for these types of things. This is one of the reasons why we built Midnight. It's actually one of the principal reasons. So, when this bridge hack happened, all these people came, they said, "Oh, well, that means midnight's dead. And isn't this proof that midnight's gone away?" I said, "This is the single biggest advertisement for why midnight is needed.

[00:43:09]We're not going to stop the hacks.

[00:43:12]They're only going to get worse. There's been more vulnerabilities discovered in the Linux kernel. That's the OG of open source projects in the last two months than there have been in the last two years because of Fable and Mythic and these other things. The US government, even the NSA, their systems are remarkably porous and vulnerable. It took only 11 hours to break into the NSA's most crown jewel servers uh with these new AI systems. They're vicious in terms of their cyber capabilities. So, the hacks aren't going to stop. But crypto can't be broken. Okay? If I have a digital signature or zero knowledge proof, if it's written correctly, uh, implemented correctly, which you can do a lot to make sure of that, very low probability that somebody's going to figure out a way around that.

[00:43:56]Mathematics gets in the way of AI. Math beats AI. So if we layer our systems with multimodality, you know, you call that high assurance.

[00:44:07]And that's the realm Cardano lives in.

[00:44:09]And that's the realm we all live in.

[00:44:10]This is why these things are so infrequent in the Cardano ecosystem. And usually they come up when a third party from outside of our ecosystem builds something or an entity is just not competent with its engineers.

[00:44:24]It's very rare to see a formal methods built system crash and burn. So, I hope this lecture has been valuable to you.

[00:44:32]I'll go ahead and give you guys a link to the uh formalization that Agalos and his team wrote. It's a very lovely paper. It's a little hard to read if you're not a cryptographer, but uh you can still kind of throw it into Dr. Chat GPT and it can give you a nice little summary. Um and it can tell you what you can and can't do and gives you all these nice categories. And I I think they have done an amazing job of uh doing things and actually they even formalized a few things in quint. So the model checking was done in TLA which is super cool. You typically don't see that in an academic research paper both a UC formalization and a Quint uh formalization. So this is a very rigorous piece of work. Anyone who's building a bridge if you're all vibe coding these days you know who you are. uh this is the first thing you load in and you use it to generate your formal tests, your property tests and your uh model checking and these other things. If you don't do this, go away, go home. So, I'm very glad the uh cryptographers put all the effort into it and got it done. It's 43 pages of nicity uh and it was written in many different places. Estonia, France, uh Scotland, very interdisciplinary team uh and uh even Microsoft research participated in writing this. the Microsoft Azure AI guys uh Dematar Chichv uh was one of the co-authors. So uh just wonderful paper.

[00:45:56]Okay, so that's all I have to say. I hope this video is helpful for everybody. Um everything's going according to plan and uh you know crypto is dangerous and you know we created Midnight to fix it. Thanks everyone.