Hi everyone, this is Charles Hoskinson broadcasting live from warm, sunny Colorado. Today is August 30th, 2023, and I'm making a video to go through some of the questions from the Husk Said list. I haven't had time to do an AMA in a little while, and a backlog has built up as a result.
Just to refresh your memory, there's a wonderful website run by someone in the community called husksaid.com. It takes all of the AMAs that I've done, and I suspect they'll even take this video. They aggregate the information, and you can search for specific topics. For example, if you want to know what I said about Ouroboros, you can find all the different YouTube links of things I've said about it. They also take nice AI transcripts. I asked the people who created the website to create a section where you can ask questions, and they are now aggregating those. There's a very long list of questions starting to accumulate, so let's just start here. I’ll look through them quickly, and since I have a little bit of time, I figured I would start answering them.
Is 45 billion the max supply? I had tipped to Trump? No, that was derived before Trump ran for president, way back in the day. It's just a coincidence that he’s the 45th president.
When a proof-of-stake or proof-of-work side chain is bootstrapped with a set of nodes on the base layer, what role, if any, would stake delegators play in consensus? How we're imagining bootstrap to work with this—let me go ahead and open up a screen to show you guys. We’re going to talk a lot about the partner chain and side chain idea at the Cardano Summit because the team is working hard to pull together different threads and pieces.
The short line is that you want a situation where ultimately the user is in control, and the builder is in control of how they share consensus with Cardano. You have the mainnet, which we generally call SL if you refer to the Y Cardano document, and then you’re going to have the partner chain or side chain, which we also call CL, for control layer. The basic idea is that the control function is Minotaur, a paper we wrote about two years ago. This idea of multi-resource consensus says that this is proof-of-stake, and you can run in a BFT mode by default. That can come from the delegates, but you could also do things like proof-of-work.
One of the things we’re exploring with Midnight for a long-term release is a proof-of-stake BFT mode with proof-of-work, where the proof-of-work is useful and specifically related to the zero-knowledge infrastructure to accelerate it. You actually get a proof acceleration layer inside the system. The percentages of control between these two different sides—Minotaur allows you to have multi-resource consensus, so it’s not just two; you can actually have multiple consensus algorithms chained together. The balance between these different things is completely up to the person launching the partner chain.
What’s really elegant about this model is that, unlike other sidechain systems where you’re kind of captured and subservient to the main chain, you have a scenario where two different blockchains can come together in a partnership. For example, Cardano and Algorand. Algorand is currently trying to figure out how to build a proper incentives layer to get higher participation with their consensus algorithm. They have a great BFT protocol created by Silvio, which is a phenomenal piece of academic work, but they left the incentive side to be figured out.
While you’re doing that, you could say, “Okay, we’ll have the Cardano network be a backbone under Minotaur, and Algorand can take another side of it.” Maybe it’s a 50/50 split, where one side is compensated and the other side is uncompensated. We’ll spend a year or two figuring out the idealized staking model. In the meantime, rewards get paid to ADA holders, and once we figure out the whole model, we can just cut off the Cardano side, and it can go its own direction.
You could also include Solana in this model. The advantage is that they get their high-speed proof-of-history on one side, and they’d use Cardano to keep the network live, so they never actually have to reset the network. They’d still have a high-performance chain while using Cardano resources, which they have to pay for by sending the inflation and block rewards over to the stake pool operators and the delegators on the ADA side.
If you had a proof-of-work style system that’s a partner chain, you would have a reward function in your tokenomics. That reward function will decide how much goes to proof-of-work and how much goes to proof-of-stake. If there’s no proof-of-stake at all, then 100% of the rewards go here. In that case, this actually becomes its own layer one, running its own network. Because that basic infrastructure is there, you could conceivably roll it back through some form of governance mechanism that’s parameterizable by the users in the framework.
These are the kinds of things we’re thinking about and building in the partner chain framework for Cardano. Midnight is really the first case study, but this is an example of where World Mobile can go. You can have stablecoin side chains. I know that SingularityNET really wants to do Tota, and we recently wrote a paper on how to use proof-of-work to train a large language model. There’s a lot of potential here, and I think it’s going to build a very vibrant ecosystem.
What does that mean for the ADA holder? When I said Ethereum has a parasitic load problem with its layer twos and side chains, these things are taking transactions out of the main system and not benefiting it. Here, if they’re using Cardano infrastructure, they pay inflation and the coinbase rewards. So, if you’re an ADA holder, you’d condition getting ADA; you’d get all the tokens under this arrangement in your staking rewards because your stake pool operators and that partnership are maintaining the infrastructure.
It’s a lot more work for the stake pool operator because they have to run an algorithm node or a Solana node, so it’s non-trivial. There’s a lot to sort out here, and there are many conversations happening. We’re going to tell this whole story at the Cardano Summit and have lots of cool stuff about where it’s going and the framework that we think will be awesome for the community. The power of decentralized governance is key.
Let’s switch back to the questions. Is there a misunderstanding of the notion of a security? Staking rewards do not mean it’s a security because the only way you get the reward is by doing work. If you have a dividend on a bond, whether you’re fishing in the lake or sitting there and helping people honor their commitment to pay you, that’s completely immaterial. The fact that they have an obligation to pay you is irrelevant. No one in a properly designed proof-of-stake system gets a staking reward unless they’ve done some work for that reward. It’s compensation for work. When you delegate to a stake pool, they’re working with you in a partnership, but that stake pool still has to make blocks. If they don’t make blocks, they don’t get paid. The same is true with the partner chains; if they don’t make blocks, there are no block rewards. It’s very important to understand that.
Tell us about bison meat. What does it taste like? How do you cook it? What spices or sauces, if any? If you make your own sauce, will you bottle and brand it? I am going to make a hot sauce over in Wheatland to create some jobs there. We’re going to call it Wheatland Widowmaker, using the hottest peppers in the world. We can put all that synthetic biology and genetic engineering to use to do that. I just like creating jobs in my backyard.
If you’re really keen on how to cook bison meat and how it’s different from beef, there’s a lovely video from Red Meat Lover. It’s a great YouTube channel that has a taste-off comparing American bison steak versus USDA prime beef steak. There are a lot of resources available on this topic.
As far as being a stake pool operator, is it possible for one stake pool to handle mainnet and also side chains simultaneously? One SPO is a logical entity. Yes, because they’ll containerize in practice, and each container will run the agents to do that. For additional detail, I see a lot of SPOs having more than one stake pool, but my question focuses on one stake pool.
So, say pool two can handle transactions; they’re going to have to run different infrastructures because they have different consensus algorithms and different logic. They’re actually running a full node on both sides.
Could you update us on why Jed was depegged? The value per coin was sometimes over a dollar ten a month or so ago. It’s just market mechanics and the nature of these things. It’s important to understand that all stablecoins, whether asset-backed or algorithmic, will not stay exactly at a dollar if they’re pegged to the dollar. What happens is that somewhere in the market’s perception, belief, or mechanics of trading, they can either be undervalued or overvalued.
For example, let’s say you have a stablecoin that’s fully asset-backed, and you have redemption rights. If you notice a market anomaly where that stablecoin is worth 90 cents, if you believe those redemption rights are solid, you’re going to buy that coin until it reaches a dollar. Then you’ll redeem all those coins for actual dollars and get 10 cents per coin, which will push the price up. If it’s a dollar ten, you’ll sell to anyone willing to buy it because you know your redemption right is only a dollar, creating a 10-cent overvaluation.
This push and pull of price is a common mechanism. Jed is a living protocol; it still has everything you’d expect in a living protocol, meaning it’s changing as you go. There’s enormous economic data, and there are already plans with Cody to upgrade Jed to version 1.7 to fully implement the extended Jed protocol. There’s also something called the Jed Alliance, and there’s a lot of activity in the Jed world.
It’s a fast-growing project that got mired a little because it was often in the corner, but it’s now being brought to the forefront because algorithmic stablecoins are the only way to create asset-backed stablecoins with volatile assets. If it’s asset-backed, it will eventually be regulated in the traditional financial world as a bank, which is rightfully so because they have bank risk in that structure. But if it’s in the DeFi world and strictly Web 3, it is crypto, but you have to work on the protocols.
This has happened in every algorithmic stablecoin. The question is whether it can resolve itself. Let’s check real-time where Jed sits. We’re at about a dollar four, so it’s returning to the peg. If you take a look at the all-time chart, it sits around a dollar, a little bit higher. For the most part, it’s staying within a reasonable band of about plus or minus five to ten percent of the anticipated value. The market cap is about 3.4 million, with a volume of about 124,000, so it’s not a very big project, which is rightfully so because the algorithm still has to mature.
If you’re really keen, go to the Jed Alliance. They are discussing various ways to upgrade and grow Jed and Cody’s Jed, the Sigma USD. They’ll come in a C1 implementation with various contributors working on it.
What tech will allow us to have fast transactions? The transaction is done within a few seconds instead of having to wait 20 seconds like the current block. That’s called fast finality. There are two routes to achieve this: one is Hydra, and the other is through a finality gadget. Currently, our thinking is to solve finality gadgets in three different levels. One is VRF improvements, which should get it from where it sits today to about an hour. That’s like an hour with no guarantee of no rollback, especially for side chains, but in practice, it’s just a few minutes.
The next step is to implement an actual finality gadget in the system, which is a piggyback protocol that runs on Ouroboros. That’s a post-genesis work stream called Ouroboros Paris, which will take you from about an hour to around ten minutes. When input endorsers come, we can upgrade it one more time to get to about the minute range. To achieve sub-minute latency, you create smart contracts that provide instant finality, and someone takes that finality risk for about a minute for a higher transaction fee.
It’s usually how you do it; you don’t have to push the protocols too much. It’s hard to get sub-minute latency on finality because there has to be coordination across the entire world. Your minimum is 100 to 200 milliseconds, so you’re already at a 20-millisecond budget just for coordination overhead with most distributed systems. To get true finality, it’s difficult, so minutes are usually what you aim for.
Does that impact cash registers or these types of things? No, because you can have those smart contracts that take that risk and get you there.
How is the Edinburgh Decentralization Index going? Do you really think it can become the standard measure of decentralization? We’re pushing very aggressively for them to get that measurement of Bitcoin done because that’s the bellwether. We’re also talking to European regulators, UK regulators like the FCA, U.S. regulators, and Middle Eastern regulators, trying to get them to say this should become a component of a larger story on how they think about whether a cryptocurrency is sufficiently decentralized or not. It’s coming along quite well. After Bitcoin, the next measurement I’d love to see is Cardano and Ethereum, so we actually have a comparative matrix and can see where everybody falls. I’m very excited about that.
My 96-year-old mother gets about five to ten scam calls a day. Everyone’s emails are full of spam. You talked once about how Cardano could solve identity on the internet. How would that work? Could Cardano solve the problem that my mom faces with spam callers? Absolutely. If you have a DID-based identity, you can have people sign. For your 96-year-old mother, you can create a whitelist of people that can be managed by you, someone else, or an entity you pay a subscription fee for. Only things signed by that whitelist can actually call her, and everything else gets auto-rejected. That’s how you protect people who can’t protect themselves—children, very elderly people, and those who may not fully understand whether the person they’re talking to is real or legitimate.
You move from a blacklist to a whitelist and curate that list as a power of attorney or trusted authority. This is very easy to implement with DIDs.
There’s a wonderful website I want to mention called Do Not Pay. This is super cool. I was at AI4, and the guy who started this is here. It uses large language models, and you sign up to use AI to help you fight corporations, cancel subscriptions, and automatically cancel free trials. It’s all relatively cheap. I don’t own anything or have any relationship with this company; I just thought it was one of the coolest things.
In the United States, if robocalls contact you, you can sue them for a thousand dollars. They have a service where they can give you a fake credit card. If you get called by a robocall, you give them the fake credit card number, which then gets sent to Do Not Pay with the address and contact information of the robocaller. They then sue that person. There’s a guy in New Jersey doing this for Do Not Pay, and he’s making twenty thousand dollars a month with it. I just thought this was one of the coolest services.
It’s an example of how you can use AI to do these types of things. Services like this are coming. Whitelisting is another example of a service, and some combination of AI and whitelisting, along with putting people in trusted authority, would solve that for her.
What do the folks in Dubai think about the theme of the hackathon? It does not seem to me that there’s interest in technologies of transparency. Actually, they are quite interested in transparency. When you have too much money, you get very interested in transparency because you have trillions of dollars to invest. If you don’t have transparency, you get scammed, and they’re having that same problem. The Dubai government has actually been more forward-thinking about transparency and accountability than the U.S. government is at the moment.
Is RISC-V on IOG’s radar? Yes, RISC-V, for those who don’t know, is an open-source instruction set complementing RISC and the regular RISC. They’re trying to build open hardware. Sci-Fi is one of the companies involved, and there’s the RISC-V Alliance. They’re asking questions about open-source hardware. We are considering this for the optimal high-security stake pool, where you have RISC-V, compile a certified kernel like seL4, and then have a certified operating system like Mirage OS. You basically use that as a complete stack end-to-end on the software side to run a stake pool operator.
We’ve also been looking at how much it would cost for a person to be their own micro ISP. The United States is starting to open things up. Getting IX level access in Wyoming is only $5,000 a month, which is pretty interesting. Maybe there are some subsidies where people can get access and control the network layer as well.
Thoughts on Nick Bostrom’s Life 3.0? Not enough time for that.
Why is the MBO called Intersect? Is it a meme? No, it’s just the intersection of all the different interests of Cardano—the NFT, DeFi, and our interests and your interests all come together. They intersect, we talk about it, and we come up with a conclusion.
Any updates on the Wyoming stablecoin progress? We are making great progress there. The state is making great progress. They actually have public open meetings. The next one is on September 7th. If you want to see, I’ll probably be up there in Wyoming. I don’t know if they’re holding it in Cheyenne or Laramie, but by law, they have to do it. It’s just a question of how to get it done.
Well, anyway, I have another meeting to jump into, but I’m just trying to work my way through the Husk Said list. I hope this was helpful for you guys. If you have a question, ask. If this could be chopped up and the Husk Said people can get the answers underneath the questions, that’s a good interface. Maybe that’s the next version of it. I’ll keep chipping away at it. Thanks, everybody.
1:07:35
26:05
26:19