I'm sorry, but it seems there is no text provided for me to edit. Please provide the transcript you'd like me to clean up.
Hi everyone, this is Charles Hoskinson broadcasting live from warm, sunny Colorado. Always warm, always sunny, sometimes Colorado. This is the third time today I've got Logan on the mic, and we're just going to get through it. This is the Midnight Special for November 16, 2023. Guys ready? Let's go!
Okay, first step is we're going to go ahead and share the screen. There we go.
So, a lot of community questions about Midnight. A lot of people are interested in Midnight, and there's a lot of stuff floating online. I did a whiteboard publicly with my 30-minute keynote, but we didn't really have a lot more time for that keynote, unfortunately, or else I would have gone into a lot of detail. So, we just kind of left it for this, and I said, you know, I owe you guys a whiteboard. I've been terribly ill for the past two weeks, but we've pushed through, and that's okay because we've got a scholar draw and Obsidian.
So, what we're going to do is say, "Hey, Midnight notes, how about that?"
In the beginning, there was just the Cardano asset and settlement layer. What I’ll do is shut this notepad down so we have a bigger screen here. So, in the beginning, we just had the asset and settlement layer of Cardano. Basically, what this layer was all about was that you have Cardano assets and settlement. There we go, stroke width, all that stuff. We'll zoom out just a little bit.
So, you have things like Cardano native assets, you have ADA, you have the extended UTXO model, you have Plutus, you have Ouroboros, and other artifacts that exist there. It’s a really technological marvel when you look at what we’ve constructed from Byron to Vol, on-chain governance, the Cardano network stack, liquid non-custodial staking. There are thousands of innovations that live here, and there’s a design philosophy that lives here that’s very powerful.
The goal is to gradually upgrade this system so that you really make it the best place to issue assets and the best place for fast settlement. There are programs like Ouroboros, there are PR programs like Babel fees, and huge improvements have been made to Plutus and the family of languages that compile into it, like AEN, for example, which is a community project. There are refinements and iterations to the extended UTXO model, and obviously, a lot of discussion around improving the native asset standards, including new standards like security tokens, etc. All of that is wrapped up with best-in-class on-chain governance, and this is the concept of CIP-1694 in MVG.
Now, that’s all fine and dandy, but if you’re a DApp developer and you want to write a great DApp, let’s just expand one of these things right here. We’ll put a little background inside that thing, maybe a little happy blue. There we go, and we’ll fill it like a shade.
So, let’s say you’re a DApp developer and you’re creating a nice little DApp. Your DApp is going to likely consume a large collection of different things. You’re going to have some off-chain stuff; you’re going to have perhaps, you know, an oracle, maybe like Chainlink. Obviously, you’re going to do some stuff on Cardano. Maybe you are using Cardano native assets, or maybe you’re using Marlowe to do your settlement and all your financial logic.
So, all of your ownership stuff is kind of managed there, and maybe you connect your off-chain and on-chain stuff with Plutus. Now, what happens when all of a sudden you say, "Hey, I want to add a regulation layer to that DApp?" Right now, that would probably be another off-chain provider, maybe that’s like Acuant or Verisign or something like that, doing your KYC and AML. There are a whole bunch of options there.
Well, the whole concept of the Cardano service layer, in a nutshell, is to say, "Well, hang on a second here. We want our infrastructure over time to gradually decentralize." So, wouldn’t it be cool if we allowed people to build service layers inside the system?
These service layers are basically decentralized infrastructure. They have some characteristics about them: they are, for all intents and purposes, their own blockchain. They have their own tokenomics, their own transaction logic, their own network logic, and their own consensus logic. There’s going to be some form of connecting tissue that goes between these networks and allows them to talk to each other. Usually, we call that some form of a bridge.
Now, the problem with this idea is that it effectively leads to a very difficult-to-manage and fragmented ecosystem. Back in 2016, when I wrote "Why Cardano," I introduced this concept of a computation layer. I’ve kind of renamed it a little bit, and I’m starting to call these things the Cardano service layer because I think of it like microservices or service-oriented architecture.
At the end of the day, this is what we care about: the DApp. What I want to do is take as much of this off-chain centralized stuff and put Cardano at the center of the DApp. If your DApp has an asset, make it a Cardano native asset. Write all your financial logic in a domain-specific language like Marlowe. Do some connecting tissue in Plutus, and all your off-chain stuff.
What I’d like to see happen is an ecosystem of service chains form. Those chains basically provide different views or contexts with different capabilities, and every single one of them has different logic behind it. Excuse me, every single one of those service layers has different logic behind it, meaning they have different tokenomics, different transaction logic, and different network consensus logic.
But what we can do is bundle this in just the right way where they actually have a common ancestor, a common set of rules, so that basically they’re in partnership with Cardano. We’ll get into specifics with Midnight.
So, common rules, okay? Common ancestor. Those rules basically say, "Okay, is it a layer two or is it a side chain?" What if we say it’s both, meaning that it can have its own security mechanism? Some part of the network can secure itself internally, and that will be connected in some way to the logic of the blockchain.
I’ll give you guys two examples on that. At the same time, it borrows security from the Cardano chain via quorum sampling of the SPOs. The SPOs can go and run a supernode, and they can actually keep part of that chain secure.
Now, if these are service layers, let’s actually have a common set of rules in terms of the tokenomics. On the consensus side, you want some commonality. We call that commonality minor, but basically the idea is that when you’re setting up your partner chain, you start from a default where the Cardano network has the option to secure that chain.
So, it almost behaves kind of like a layer two, but then you can also add on additional layers of security that are relevant to the logic of your system to connect your system to other systems, maybe like Ethereum, maybe like Bitcoin, or maybe it’s an internal thing like you’re having your own proof of work or something inside the system because it’s useful for your context.
You can kind of layer these things together, and you can split up your tokenomics accordingly. Then what you do is follow a dual token model. The dual token model basically says that you issue your deflationary asset on Cardano as a Cardano native asset. You issue a token like that, and this kind of represents the control, the governance, the ownership of the system. Ideally, a deflationary monetary policy makes sense there.
Then you have another token that lives in the service layer, the service layer token, and that is effectively the fuel token of the system. That can follow a different monetary policy, especially to give you predictable pricing. Ideally, what you’d like to happen is that the token that lives on Cardano has some mechanism to generate, with some degree of regularity, the fuel.
That’s how you connect those two. What Cardano really does provide is a mechanism to sort out a distribution of the service ownership and resources, so you get a kind of a common distribution from Cardano regardless of the particular service that you’re looking at.
This could be like World Mobile Token, for example, if they wanted to run World Mobile Chain as a partner chain. This could be like an AI chain or something like that if they want to go down that road. You could have dozens of these things.
You think, okay, an AI chain looks like ChatGPT. Every time we query it, it is a transaction. You pay a predictable fee for that transaction. So, there’s a token that would live here, and that’s a fuel token. You want bandwidth? There’s a token that lives here, and then you’d have a Cardano native asset. That native asset acts as a regulating function to basically create that fuel with some degree of regularity. That’s how you connect those two things together.
If they’re one token and they’re both connected together, the deflationary nature of the system is at odds with the variable monetary policy that you need to have with the fuel system. So, it’s good to actually separate them in some way, and this is kind of the thing that we’ve been exploring on the Midnight side.
So, you have something that lives on the asset settlement layer, you have something that lives in the service layer, and then basically one generates the other. Security comes from here, and then you parameterize minor to basically get security from either just that one source or multiple sources.
When transactions happen and blocks are made, it generates that ownership token, and that gets paid to the ADA holders and the stake pool operators through the delegation process.
Now, if you’re a stake pool holder, if you’re a stake pool operator (SPO), in practice, you’d be running a supernode. There are some examples of this; like Firefly is kind of a supernode maintained by Hyperledger. That supernode basically gives you the ability to plug in through containerization multiple chain structures, multiple chains.
The idea of having a common framework, a common ancestor, is that there’s a uniform process for the stake pool operator to have their Cardano container. Then they’ll have their Cardano service layer one, Cardano service layer two, service layer n, and each of those service layers effectively are things that they can maintain. They just drag and drop and pick them.
We’re kind of looking at this supernode concept in parallel, and one of the things that we’ve been exploring is this idea of log space mining. We wrote a paper on this a few years back on the proof of workspace, and now we’re trying to explore it from the Minotaur perspective of how to apply a concept of log space mining so that you, as an SPO, do not have to maintain a full node, but rather you have a lighter-than-full node for the side chains that you’re maintaining, the partner chains that you’re maintaining.
You kind of just have the tip of it maintain the state and the transitions of the system, and then these are all containers effectively. As a delegator, the idealized thing is metadata of the stake pool certificate. You’d be able to see effectively which pools are supporting partner chains and which chains are supporting.
We try to make that a lightweight experience for those SPOs with some sort of supernode architecture. Again, Firefly is a good reference for how that would be done for each one that they’re supporting.
What would happen is, as they make blocks and transactions are done, rewards are paid, and those rewards basically flow through. That’s how you create an incentive layer for ADA holders, and you also create an incentive layer for SPOs to maintain these services.
Minotaur gives you the ability to kind of decide how your security is going to come together, and then you can build other bridges here. The dual token model basically separates the ownership, governance, and control of the system and fuel generation of the system from the actual fuel of the system. The fuel will live on the service layer, and the native asset will be the ownership of the network here on Cardano.
That’s kind of the rough idea of how this all comes together. When you’re a DApp provider, you say, "Oh, okay, I want to use Midnight now for all of my KYC, AML, and my data protection privacy stuff." Then there’s this question of, "Well, what currency do I pay in, ADA or Dust?" The reality is it actually doesn’t matter, and this is why Babel fees were developed.
The whole concept of Babel fees basically means that the user of the DApp pays in the currency that they have. If they have ADA, they pay in ADA; if they have Dust, they pay in Dust. There are mechanisms under the hood that effectively sort out how that settlement consumption is going to work.
At the end of the day, under the hood, it’s all ADA, but Babel fees sort out the settlement of these things. More of these native assets on the system create more demand overall for ADA as kind of the fuel of Cardano. This is a recognition of some facts that we have to contend with in a broad ecosystem of cryptocurrencies.
The first thing you really have to accept is that the world is multi-chain. No matter how much we want to be Cardano Maxis, the thing is that this was built to be a tank. Ouroboros is absolutely extraordinary, Plutus is great, extended UTXO is great. We’re not going anywhere, and with LAOS and other things, we’re going to make it really fast.
That input endorsers agenda is there, and there’s a great roadmap here to make Cardano best in class. There’s a whole bunch of great infrastructure here, but just because you have all that infrastructure, what happens when you need privacy, data protection, identity? What happens when you need an AI component? What happens when you need decentralized network components? What happens when you want to get rid of Infura and Alchemy and have a decentralized host?
Basically, something like a decentralized Alchemy where you’re running completely in a web3 context. That infrastructure is no longer there. Does it make sense to overload the Cardano chain with all these features? In many cases, they’re very contradictory to each other.
You’re going to have to admit, for example, a mutable world because you’re paying rent to utilize some space, and when you stop paying it, it may stop storing that. This is a blockchain; it never throws anything away.
The reason you need a service layer is you need the ability for services to come and go, and you also need the ability for services to have different design philosophies. For example, having all this really sophisticated SNARK technology is awesome for a lot of use cases, but if your goal here is a super high-performance ledger that’s public by default with high liquidity, it doesn’t make a lot of sense to conflate these two.
The point of a service layer is you don’t have to give that up. You have a coherent supernode notion management model where your stake pool operators can comfortably manage many service layers at the same time. That’s the point of log space mining, and each of those service layers is kind of like a service-oriented architecture where DApps can call them as needed, and they pay in the currency that they care about.
At the end of the day, Babel fees just kind of settle all of that, and the dual token model means that you can separate the ownership, governance, and control of the system from the underlying fuel so you can start getting predictable prices.
This has to be competitive with Amazon, this has to be competitive with OpenAI, this has to be competitive with T-Mobile, this has to be competitive with Verisign, and a lot of the services that provide privacy services. That’s not going to be competitive when every day the price changes by 600% based upon market conditions.
We thought a lot about how to put these types of things together, and we live in a multi-chain world. We also live in a world where interoperability is king. You have to have an interoperability strategy.
The reality is Ethereum’s not going to go away, and Bitcoin is not going to go away. You need some sort of mechanism or mechanisms to bridge and make other networks partner and care about your system. You need some sort of way of doing that, and that’s effectively what we came up with with this concept of a service layer model.
You’re saying, "Oh, okay, it’s all right for these guys to do their thing. They have great ecosystems, they have millions of people, and they have DApp developers thinking about stuff." The reality is that those DApp developers are potentially customers of Cardano.
The problem is that right now, the DApps that live in this ecosystem on the DApp that live here, they can’t talk to the things in Cardano. By having a service layer that has a lot of hooks for interoperability, what that effectively allows you to do is have a DApp. Instead of just being a Cardano native DApp, it could be written in Solidity here on Ethereum but then call a Midnight service or call our AI chain or call our decentralized network or whatever we’ve got here, especially with this decentralized hosting.
That’s something that is kind of a holy grail, and it’s a very fragile point overall in the Ethereum ecosystem. If you care about interoperability as one of the truths, you have to have some sort of strategy for how all these things are going to fit together.
A big part of the Cardano side chain agenda, that common ancestor, that common set of rules, is this idea of trustless bridging. It’s another component. The first generation is a little clunky, but eventually, what you want to do is use recursive proofs to give yourself a high degree of assurance for supported chains. When you see a transaction go back and forth, as long as you know some baseline things about these chains, which it’s pretty easy to inject into the system in a trustless way, that’s actually accurate.
The rollup community has been talking about this a lot, the recursive SNARK community has been talking about this a lot, and we as an ecosystem are starting to move in this direction of building bridging that is semi-trustless or trustless in general. So, interoperability relies on this: multi-chain interoperability.
The other thing is this: this is the biggest one and the most important—legacy friendly. You need to have the capacity and ability to bring many users on board from the systems that have billions of people, like the Microsofts of the world, the Facebooks of the world, and others.
That’s where we invented concepts like RVP and this idea of hybrid
1:07:35
26:05
26:19