Epicenter - Learn about Crypto, Blockchain, Ethereum, Bitcoin and Distributed Technologies - Muneeb Ali: Stacks – Building Decentralized Apps and Smart Contracts on Bitcoin

Episode Date: February 16, 2021

Stacks is an open-source network of decentralized apps and smart contracts. It is a Layer-1 blockchain that connects to Bitcoin. Stacks uses a Proof of Transfer mining mechanism with bidding and staking capabilities. Clarity is the programming language used, a Turing-incomplete language which aims to give developers a safe way to build complex smart contracts where the code itself clearly shows what the program will do when run. Muneeb Ali is the CEO and Co-founder of Hiro Systems, one of the companies creating Stacks. He joined us to chat about why he felt there was a need for Stacks, why he chose to build on top of Bitcoin, and the types of apps that are currently being run on Stacks 2.0.Topics covered in this episode:Muneeb's background and how he got into cryptoWhy Stacks was createdHow this works being built on BitcoinMining and bidding on blocksSmart contracting on Stacks - Clarity programming languageStacks vs “Bitcoin on Ethereum”Stacks vs RootstackThe apps that currently run on Stacks 2.0Episode links: Stacks websiteStacks whitepaperStackingProof of TransferHiroStacks on TwitterMuneeb on TwitterSponsors: 1inch: Discover the best rates and most efficient swapping routes across leading DEXes. Optimize on gas cost and execute DeFi trades faster with 1inch V2 - https://epicenter.rocks/1inchThis episode is hosted by Sunny Aggarwal & Friederike Ernst. Show notes and listening options: epicenter.tv/379

This is Epicenter, episode 379 with guest, Munib Ali. Hi, I'm Sebastankujiyo, and you're listening to Epicenter. The podcast will be into recruit our founders, builders, and thought leaders. On this show, we dive deep to learn how things work at a technical level, and we fly high to understand visionary concepts and long-term trends. If you like Epicenter, the best way to support us is to leave a review on Apple podcast. If you're on a Mac or iOS device, the easiest way to do that is to go to Epicenter. Dot rocks slash Apple.

And if you're new to the podcast, be sure to subscribe on Apple Podcast, Spotify. or wherever you listen. Today our guest is Mooneeb Ali. We had Mooneb on over five years ago on episode 101. And back then he was working on a project called One Name, which was a sort of decentralized identity system where one could register usernames and domain names. That project has evolved in a pretty drastic way to become Stax.

And Stax is a smart contract platform that is built on top of Bitcoin. It is powered by Bitcoin. In the world of smart contract platforms, there are just a handful of projects that leverage Bitcoin for security and stack positions itself, not as a side chain, but as a layer one chain that shares the Bitcoin compute power for security. They just launched Stacks 2.0, which introduces full smart contract capabilities with their native programming language clarity. And as you'll learn in this conversation, clarity has some interesting design differences to solidity and the way that we think about code execution on Ethereum. So I really like to see platform projects building on Bitcoin. There's an enormous amount of

value and capital in Bitcoin that just isn't productive. It's sitting there and it's hodling. Projects like Stacks makes it possible for Bitcoin to finance ecosystems in a way that feels much more native and intuitive than wrap tokens on Ethereum. And in the case of Bitcoin, that's like a trillion dollars of capital that could potentially finance an ecosystem of Defi and Web3 apps. So I'm hopeful to see projects like stacks gain traction and become successful because it can just grow the defy ecosystem in a pretty exponential way. If you need to make a swap, look no further than one inch. It's my go-to decks aggregator, and I know that when I use one-inch, I'm getting the best price across all deckss and AMMs. They recently did anirdrop and the one-inch

foundation has distributed more one-inch tokens. You can learn more about that on their blog. or if you want to use one inch, you can go to epicenter.rocks slash one inch. And with that, here's our conversation with Munib Ali. We're here today with Medib Ali. He is one of the co-founders of the Blockstack or Stacks project and the CEO of a company called Hero. Meneb was on the show quite a long time ago, over five years ago. We're talking about the same project actually, but, you know, on many iterations ago. And back then, it was called One Name.

And so, you know, it's really great to have, you know, have a lot of the old veterans back on the show to give updates on, you know, the state of what they're working on. So, Meneb, it was great to have you back on again. Thanks. Thanks for having me. So, yeah, can you give us a little bit, you know, maybe give us a little bit of a recap of your background because, you know, I'm not sure how many of the same listeners we have from five years ago. So for those who haven't watched that previous episode, how did you sort of get involved into crypto and this whole blog? blockchain thing. Yes.

So my background is actually in computer science. I did a PhD in distributed systems at Princeton University. And this is where the project kind of like started as well at the computer science department at Princeton. And we were actually focusing on building kind of like, you know, more secure, more decentralized internet infrastructure, like internet protocols, not nothing to do with crypto in the beginning. Right. And personally, I've just been doing research work in distributed systems like for 10, 15 years before starting the project. And for people who are not that familiar with distributed systems, think of that as the computer scientists who work on internet protocols or operating systems or these large-scale network systems like data centers and so on.

So I was pretty much working on cloud computing effectively during the early part of my grad school, and then got pulled into kind of like this really ambitious idea of not just a clean slate next generation internet design, but not doing it in academia. So we have actually seen a bunch of efforts in academia where, you know, at Stanford, for example, there was a product called clean slate internet design. They would build prototypes. they would write research papers about it, but it would never really get commercialized

or adopted in the real world. And I always felt that the real challenges are in that step. Like, how do you actually take some of these ideas and get millions or billions of people out in the world to use it? Otherwise, I think it just feels like you're just sitting in a bubble and you're just in a research lab trying out different ideas. So I think that was kind of like the motivation that let's try to take a staff,

at this very ambitious thing of trying to upgrade the internet, but do it in the industry. So raise venture capital. And interestingly, in the early days, we are very, very fortunate to have people like, you know, Navalra Wei Kant or Y Combinator and Union Square Ventures, who effectively funded us perfectly knowing that we are hacking on open source technology, and we have no business models in mind for like 10 plus years. Right, like this was the pitch, right? Like we're building like this technology in a completely open source way. And if it's successful, like it would likely have an ecosystem around it.

But, you know, it's not like, you know, we have any sense of any business models. We're not not not. It's almost like a R&D project at that time. And they were, they were very supportive in even funding that from a venture perspective. How long ago did you guys? start the project? So this is summer of 2013. For people who measure time in Bitcoin, if Bitcoin was at $90 at that time. So we last had you on five years ago, which is 50 years in normal time. Tell us about what happened since 2015. Yes. So I think to continue the story

a little bit, we quickly discovered blockchains. And, And we discovered blockchains as a very elegant solution for some of the problems that we were trying to solve. So imagine if you're trying to build like a more secure, more decentralized DNS system, domain name system, right? Or you're trying to figure out more secure ways of distributing your cryptographic keys on the internet. Like, the blockchains have very natural elegant solutions for these things. And we discovered blockchains and we're super. excited about it, kind of like started digging deeper into how blockchains can be used for solving some of these problems. And interestingly, at that time, obviously, you know,

there was Bitcoin, but then there were some Bitcoin forks as well. This is pre-Etherian, right? Ethereum did not exist at that time. I read the white paper. It was not public, public the white paper at that time, but someone sent it to me privately. And there was a fork of Bitcoin called Namecoin, which is still around. It's like a semi-dead project at this point, but it's still around. But it used to be kind of like the number two or number three alt-coin on coin market cap back then. And interestingly, the choices in front of us were should we directly use Bitcoin?

Should we build a separate notebook, which is what Ethereum did? or should we build on something that is kind of like sharing the compute power of Bitcoin, which is like what Namecoin was doing with MergeMai. And there are different design tradeoffs for these approaches. And the way we kind of thought about it is, interestingly, it's actually easier to build a separate network. So we're seeing a bunch of blockchains being built out. And, you know, it's amazing. I'm not saying building a blockchain is easy.

Building a blockchain takes a lot of effort and careful of designing and there's consensus problems to solve. But connecting two different blockchains together in consensus is actually a harder problem, especially when one of them cannot be changed at all. So that's Bitcoin. Right? So Bitcoin has this very, it's a great property that Bitcoin has. The Bitcoin is normally to change. So if you want to work with Bitcoin, you need to figure out how to work with it without changing it. And that actually makes things much, much harder.

So at that time, we were exploring effectively merge mining type techniques that can we start building these domain name systems or PKI systems, but still somehow share. the compute power of Bitcoin. And we can go into some of those challenges, but at a high level, there were many technical challenges with that approach. And the second iteration, the first iteration,

I would say it didn't last very long, right? We quickly discovered the problems with approaches like name coin or merge mining. And the first real iteration, which was Stax 1.0, was actually directly built on top of Bitcoin.

We actually knew that this is, in the grand scheme, of thing, not a very scalable design because every Stacks transaction is a Bitcoin transaction. So if a million domain names get registered on stacks, a million Bitcoin transactions happen. And when the Bitcoin core developers weren't like truly kind of excited about it. At one point, we were the largest consumer of opt return transactions. Op return is like you embed this additional data. in Bitcoin transactions.

So you're basically using Bitcoin for other purposes. Like you're not really doing Bitcoin to Bitcoin transfers. You're actually just using the additional embedded data. And we built like Stax 1.0. It's like I think of it as a virtual blockchain. Like it's a blockchain built on top of Bitcoin. And that has limitations. Like it did not have a full smart contract language.

It obviously had scalability limitations because it's tied with Bitcoin. But back then, like back when this was launched in like 2015 or so, it wasn't clear how Bitcoin is going to scale, right? Like this is pre the fork wars, right? Even though we thought that Bitcoin would always be used as a settlement layer, that was a camp you were in. It was kind of like still a question in the industry that can Bitcoin actually take more data right at the base layer? And now it's abundantly clear that no, Bitcoin is not going to take additional data. So I would say around 2017 or so, we started working on the stacks 2.0 design. So it's actually been like three years in the making.

And interestingly, I think this is where this becomes very fascinating. It goes back to the original design decision of 2013. Do you build a separate network? Do you build a Bitcoin or do you somehow share the compute power of Bitcoin? And I think that we've cracked the problem this time around with Stax 2.0, where it shares the compute power of Bitcoin, but it is not dependent. Like the scalability is completely disconnected from Bitcoin. It can scale independently of Bitcoin, and it can have full smart contract languages. So when you say that it's built on Bitcoin, what you mean is that it's kind of, it's a separate change.

a chain that kind of checkpoints into Bitcoin for finality purposes? Yeah, there's a little bit more than that, right? So think of that as it's not a side chain, like from a traditional definition of a side chain, but one way to think of that is it's like a layer one blockchain that directly connects with Bitcoin, right? So what I mean by that is there's actually consensus between two blockchain. right. Sure our Stacks blockchain has consensus internally as well, but it has consensus between Bitcoin. And how that happens is that in terms of, you're right about the checkpoint thing,

but the way it works is that you can do thousands of transactions on the Stax blockchain, and they automatically settle on Bitcoin every Bitcoin block. So for people who are familiar with Lightning, think of that as Lightning checks. channels that are settling on Bitcoin, every Bitcoin block automatically. And the way the consensus works is that the miners of the stacks blockchain, they actually view the state of both Bitcoin and stacks. They're competing to become a miner using the Bitcoin blockchain. But if they win, so there's a verifiable random function on Bitcoin,

they're sending transactions on Bitcoin to, participate in leader election. So they're actually bidding, bidding in Bitcoin to win a block. And the miner that wins a block writes a stacks block, right, because they have visibility into both, right? And there's consensus in between both.

So the interesting thing that happens is that, let me share some stats here. I think in the first 2,000 blocks of our blockchain, So it launched around January 14th. What we've seen is miners have sent something like, you know, close to 25,000 transactions just for these Bitcoin commits, right? They're competing. They're trying to commit Bitcoin.

And they have spent something like 43 Bitcoin, which is like around $1.5 million, in that competition, right? And it, and it, what's happening is that the value of a Stax block, they're bidding on that value in. Bitcoin directly on the Bitcoin chain, right? Instead of having hardware and A6 and burning electricity to compete for mining a block, right? And that's super interesting because what that does is it says that you don't need a separate proof of work chain. You can reuse, you can actually ride on the Bitcoin's proof of work. We're using Bitcoin the asset as a proof of

of computation, right? And we shared the security of Bitcoin because the way our consensus works, if someone wants to come in and change the history of stacks, they would actually go and attack Bitcoin and reorg Bitcoin, which, you know, good luck with that. That is by far the most secure system out there. So that's kind of like this works. And there's more to it. And we're happy to dive a little bit more into the algorithm.

One inch is a decentralized exchange aggregator that sources liquidity from the top dexes and AMMs to save you money and time on swaps. One inch finds the best possible trading paths across over 20 supported liquidity protocols and splits them up across multiple market depths. I started using one inch last summer and since then it's become my go-to aggregator. I use it every time I need to make a swap. They recently launched V2, which has a brand new API. It greatly improves their routing algorithm.

And my favorite part about the V2 is the new UI. It's super clean and easy to use. These improvements ensure that you get the best rates on your swaps with the lowest possible response time. So the next time you need to make a swap, forget about getting the best rate or optimizing your gas fees. Make it easy on yourself. Just use one inch.

And you can let them know that we sent you by go to epicenter.rocks slash one inch. That's one inch. We'd like to thank one inch for their support of the podcast. How exactly do these bids work? So let's say I want to be a block producer and create stacks, blocks. How do I become the leader for this round? So the way this works is that you're effectively, you can look at historical data to see, you know, at what at what level of bidding are people actually winning?

blocks, right? And then there's also a smoothing function in the sense that we don't want miners to come in very quickly and then stop mining very quickly. We want there to be some sort of a smoothing function because the bids are actually going to depend on the STX-BTC price pair on exchanges. Some of the miners, I can tell you, that they model mining as a almost like decks where they're looking at the BTC STX price pair and they want to mine if they can get a cheaper rate than the Binance STX BTC price or some other exchanges BTC price. So they view this as an R of opportunity. So miners will only mine if they can get cheaper STX from mining, then whatever is the ongoing

real-time trading rate outside. And plus, obviously, there are more details there because there's some delay in them actually getting the STX and their other complexity involved. But at a high level, you can model that as a as a decks running on Bitcoin, right? And in the process of trying to buy up the STX that's being released, the consensus is a byproduct of that. But so how does the bid actually work? like do I send Bitcoin to a particular address? And so, and if so, if I don't get, if I'm not the winner of the auction, how do I get my bid back? So first, I'll explain the concepts with just a burn address, right?

So imagine that all the bids are going to a burn address and Bitcoin is being burnt. And what's happening is that there is a verifiable random function. and it effectively looks at if you burn more Bitcoin, your probability of winning goes up, but it is still a VRF, right? So you're not guaranteed. Your probability goes up, but you're not guaranteed to win.

So then, and it's over a window as well, right? So what happens is in any given block, the verifiable random function basically spits out who the winner is, right and that winner gets to write the the stacks block and collect the coin-based fees the newly minted tokens from stacks and the clarity smart contract fees and the transaction fees from stacks so that is the value of the block for for the miner but then this is a lot of bitcoin being destroyed right so we worked on a improvement where we were like instead like for example just in the first 2,000 blocks like 1.5 million

worth of Bitcoin would have been destroyed. So we worked on a improvement where the miners can actually agree on a set of Bitcoin addresses, which are the reward addresses. And these are people who hold the Stacks cryptocurrency on the Stax blockchain, and they also want to participate in consensus. So they are participating at a lower level, right? they're effectively locking up capital on the dominant fork, right? And when they send out this transaction to lock STX, they're broadcasting useful network

information, right? So these people become like active participants and consensus, but basically saying, here is the state of the blockchain that I see, and I want to lock my capital on what I think is the correct fork of the network, right? and they give out the Bitcoin address as part of that. It's very interesting because the mechanism itself is written as a clarity contract. So people can actually see it. It's a call to a clarity contract where you are locking STX.

And every cycle is a cycle is roughly two weeks, is 2,000 blocks. So the minimum time to lock is 2,000 blocks. And now the Bitcoin, so based on kind of like the last, almost like the registration period, we get a set of addresses, and the miners are now sending transactions to those addresses. And interestingly, the load on Bitcoin is in the order of the number of miners. So if there are, let's say, 20 miners who are competing, Bitcoin is going to see 20 transactions per block, right, independent of if Stax is doing 1,000 transactions or 5,000 transactions. the load on Bitcoin is in the order of the miners, right?

Which is a very, very interesting property to have. And in some ways, the hash power of the Stax chain is represented in Bitcoin. And the Stacks holders are getting that Bitcoin, right? So we were actually seeing, when people started getting Bitcoin directly from miners, like people were like, this is a magical experience, right? Like there are some tweets out there where people have posted videos of small Bitcoin transactions just showing up. Or, you know, there were some people who were like, I woke up to like 10 emails from my Coinbase account that you have received Bitcoin, you have received Bitcoin. And it's a very interesting experience because because all you're doing is giving a Bitcoin address, it's compatible with every Bitcoin wallet and exchange in the world already.

And that's how you're kind of like earning Bitcoin from consensus. I just want to understand. So the stacks holders, how, so they're basically putting locking coins and saying that, hey, I think that this fork is the valid one. That part doesn't happen on the Bitcoin chain. That's happening on the stack. So, but in that case, if there's a fork of the stacks chain, wouldn't the same people be able to say, well, I don't know which one? I mean, this comes back to like the traditional proof of stake, like nothing at stake problem where it's like, well, I might as well just lock on both sides of the fork. Meanwhile, if there's a way to lock on Bitcoin, then that would be.

Yes. So this is, this is not a critical part of our consensus, right? The miners are only responsible for consensus. What this is, think of that as a additional information where that helps is. It actually helps honest miners, right? People who are honest miners, they can actually tell where the majority of the users of the blockchain are. But from a pure consensus point of view, it actually does not impact consensus in that sense, other than giving additional useful information to the honest miners.

I would also like to backtrack here a little bit. So you said if there's 20 miners, there will be 20 transactions on the Bitcoin chain. So how many concurrent miners do you expect there to be? Because basically on Bitcoin and Ethereum and other large chain, there's much many more than 20 concurrent miners, right? Yeah. So we have actually seen, you know, quite a large number of miners as well. Like on the test net, we got like 800 miners and so on.

On the production network, we've seen 80 plus miners. I think 46 have successfully mined or something like that. But interestingly, just like how mining typically works even Bitcoin, our expectation is that a handful of really sophisticated miners will control some percent of mining rewards. And then there might be mining pools as well, right? And then there are a long tail of small miners. So I think in that sense, like if you look at that distribution for Bitcoin, let's say you ignore the very, very small folks, you just look at who the big. miners are or who the big pools are, like you basically get a handful of miners in a steady state, right, which is something that might happen on stacks of wealth.

It's effectively modeled after Bitcoin. And the block rate is larger for stacks than for Bitcoin, right? So basically, you have many more blocks. Each Bitcoin blocks corresponds to many stacks blocks. So how do I bid on a specific block or do I bid on the right to mine the next 15 minutes worth of blocks? So how does this work? Yeah. So we have this concept of microblocks.

So to keep things simple, stacks blocks have a one-to-one correspondence with Bitcoin. But we have further divided a stack's block into microblocks. So microblocks, interestingly, also get confront. These confirmations have nothing much to do with Bitcoin. Basically think of that as your user, you broadcast a transaction on stacks. It will very quickly get picked up by a stacks miner and get included in a microblock. And then, you know, and microblocks are a little bit different. Only the winner of the last stacks block can produce microblocks.

And so a microblock is like, you know, the last winner is just going like, boom, boom, boom, and they keep producing blocks, let's say, 6, 7, 8, depending on how long Bitcoin is taking to produce a block. Let's say a microblock is every call it 30 seconds. But once I'm in the microblock, am I in? Or is it kind of like a snapshot of the MAM pool? Yes. So that, interestingly, is how we can display that information to the user

as basically other miners can, include the microblock in their Bitcoin commits as well. So what's happening on Bitcoin is that people are very quickly doing replaced by fee, the miners are doing replaced by fee, to always have the updated information and the updated microblock in there. But you can actually just look at the Bitcoin mempool and know that the probability of your transaction getting confirmed is actually close to 100% because all of the of the stacks miners have you in their microblock. And then there are incentives there where a percent of the rewards are based on including microblocks in your transaction, like for the other miners, for the miners who don't even

win the main block, they are incentivized to include as many microblocks in there as possible. So it's a little bit of a game theory there, but it's pretty interesting. I would say in 99% of the cases, as soon as you get a single confirmation on a microblock, your transaction is getting in because every single miner is incentivized to include it on whoever wins the Bitcoin block. They would have you in their microblocks.

Let's say, okay, so let's say half the miners end up not seeing one of the microblocks on time. And so half the miners with their bids, they included up to microblock number 100. while half of them included microblock 101. Which one now, so now a new leader is selected. They continue to build off of the one that they saw, off of the bid that was in their thing,

or do they have to like take, you know, whatever the majority of the miners picked, or how does that work? So this is where we separate kind of like confirmations from finality. Finality is only driven by Bitcoin. So whoever was the miner who wrote the Bitcoin, block, basically you will see one finality effectively, right? And then Bitcoin confirmations are kind of like finality numbers.

Once they reach high enough, there are guarantees that nothing can roll back on stacks. So if you have someone determined who is allowed to mine your next block, how do you deal with them the threat of them censoring transactions? So interestingly, when thinking about censoring transactions, we have this really cool feature where people can actually send stacks transactions on the Bitcoin network as well. So transfer transactions and locking up, we call it stacking. So if you want to lock up your stacks to participate, they're supported in Bitcoin and more forms can be supported as well.

So if you're afraid that, hey, someone is not going to let me transfer my stacks or not going to let me participate in stacking, you can actually just broadcast transactions on the Bitcoin chain and they will get picked up in consensus. And I think that's actually pretty beautiful because it's Bitcoin as a backup communication channel if you are afraid of censorship. What about from the other direction? Bitcoin miners basically censoring all stacks bids. so that way, like as the Bitcoin miner, I can guarantee that I win this stacks bid as well. Yes, we did a bunch of work there. Like, I think there's very interesting a game theory.

So think of it this way that, first of all, the transactions are Bitcoin transfer transactions, right? So it's unlikely that these types of transactions would get banned on Bitcoin altogether, right? Because we're just doing transfers on the Bitcoin chain. And then if, miner, let's say, wants to play around with a stack chain and say, hey, I'm going to censor other people and just have my own bid in there. I think A, that's where the rolling window comes in.

That miner would actually have to try and control things for several blocks, right, to have a real shot at winning the block. And secondly, then it starts getting into collusion with other miners, right? Because if other Bitcoin miners are not not doing the same thing, what would happen is that you will see one block in which a miner was actually not letting in other stacks miners, but because of the window, that even that block doesn't really impact how the VRF works. That minor would have to control a long number of consecutive blocks, which is a very hard thing to do on Bitcoin, right? And interestingly, another version of, of this is that what if, you know, Bitcoin miners think that this is very profitable, why don't we mine the, the stack chain, right?

And not let anyone else mine. So the worst case scenario here, absolute worst case scenario, if you work out kind of like the game theory, is every Bitcoin miner is also a stacks miner. You turn into a merge mine chain, essentially. it turns into a condition where 100% of the Bitcoin miners are the Stax miners and they're not letting anyone else mine. But which is fine because if you consider Bitcoin mining to be decentralized, it's basically like an additional requirement to mine the Stax chain at that point. To mine the stack chain, you also need to be a miner in Bitcoin. But that is the worst case scenario, which I'm very, very comfortable with. But that is basically a similar security as like merge mining, right?

It essentially turns into merge mining. In some sense, yeah. So how does the Stax chain deal with Rheorgs on the Bitcoin level? So I think in general, Bitcoin dictates the finality, right? So if Bitcoin forks, Stacks forks with it. And whatever fork wins on Bitcoin is the source of truth for it for Stacks. And that's how it's designed, right? Like, it's basically, it's designed to use Bitcoin as a settlement layer and as a source of truth.

Cool. Yeah, I think that's a very clear-cut answer. And I appreciate that. So the Stax blockchain can do more things natively than the Bitcoin blockchain. So, I mean, you have a smart contracting script. Can we talk about that for a while? The language is called clarity. So what are its list base as far as I know?

So what are the benefits of clarity over something like solidity? Yeah, so I think this goes back to, again, fundamental design decisions. Like, just like there's a fundamental design decision that do you want to build on Bitcoin and use Bitcoin's approval work or do you want to build a separate network, right? Like once you make that decision, it's a type one type decision. same with the design of the language, right? If you decide that you want to have a Turing complete language like solidity, you know, it's very hard to roll that back, right?

You're going to live with it for a very, very long time. And we have always been in the camp, just like we've been in the camp, that the broader thesis here is Bitcoin has network effects. Any experiments that we're seeing in other blockchains will, eventually get created on top of Bitcoin, right? So just like the early internet, there were a bunch of separate networks, but in the end, you know, TCPIP and the internet ended up taking all the, all the traffic. We think Bitcoin is like TCPIP and all these other things will end up getting built on top, right? Like maybe not in that extreme of a form that, you know, everything else

disappears and there's only Bitcoin and things on top of Bitcoin, but I'm talking about the bulk of the market, right? Like, that's our thesis. So on the, on the, on the, on the, on the, smart contract language side, we are in the camp that the only thing that matters for smart contract in the long run, right, in the long run is the security properties of the language because the unique thing here is that the smart contract run on a blockchain, you know, they cannot be modified once they're deployed and you get people are keeping hundreds of millions or billions of dollars of crypto assets on these computer programs, right? So being able to reason about the security models,

and give formal verifications for the smart contracts. And basically just optimize for security. Don't optimize for anything else other than security. Like that is really the thesis. So yes, it's a list-based language, but that's more of a syntax thing. The design principle is we don't want a Turing complete language, but we want to have a language that's general enough

that most of the programs can be written. Right. That's where we started. Then we took pretty contrarian decisions like, for example, we don't even have a compiler. We don't want to have a compiler. We explicitly wanted to avoid having a compiler because even a compiler can actually introduce unwanted bugs into your code, right? Because what you're publishing in blockchain is like code is law, but what people are publishing is actually compiled code. So they don't even know if the thing that the thing is law is doing the thing that they wanted to do, right?

Very few people can read by code. Yeah. And so we actually publish source code, which is interpreted instead of compiled. You can even compile it locally for performance reasons or whatever, but it's interpreted, which means that in the blockchain, like if you go

to our Explorer, you can actually see and read the code and the comments in there as well, which I think is a much better UX experience as well, other than the benefits of knowing exactly what it is. But then the interesting part and most of the work behind clarity was to ensuring that we don't introduce pure incompleteness by mistake, right? Because you can do that in a design of a language. So you have to be very precise about everything.

Like how are you even mapping out, you know, some storage internally? And effectively, what that translates into is just like in Ethereum or Solidity, there is a gap. estimate. It's an estimate because you don't know what the program is going to do. And what is exactly going to execute and when it executes then you know what the fee was. In clarity, there's no estimate. There's precise fee because you know even before executing the program exactly precisely what the program is going to do. And that, I think that is the night and day difference in the two designs where in the industry right now, what's happening is people realize that his smart contracts are important,

let's just get a lot of manual audits on the code. Or they would try to have some sort of a formal verification by writing additional code, additional syntax, and then doing partial verification on that code base. What we are saying is, yes, obviously, you know, manual audits are great, but you get formal verification out of the box just based on how the programming language is designed. And I think that's a major,

major win for developers who care about the security of the smart contracts. So this is like some middle ground, though, between like the language can be non-ture and complete, but you're saying that there's like no need for gas system. But don't you have, can't you still have cases where there's like race conditions and stuff that like, you know, basically, you know, you make a transaction expecting it to do one thing, but maybe someone else did a transaction that changes how your transaction. acts. Do you not have that kind of situation happening on here? Or do you avoid that as well using like, I don't know, UTXOs or something like that? No, it's a account-based system. But I think

we'll have to dig deeper into like what that example is. But think of this way that it's about execution paths, right? So you can calculate like the exact execution paths for what the program can potentially take. Right. So you know, let's say, let's say, let's say, that you are attaching the precise fee for the longest possible execution path. Okay. You can still have different potential execution path, but at least we can bound the potential paths.

But given the, given a state, given an input state, you know precisely what will happen with the program. Right, yeah. The problem is we don't always know what the input state is because other people can make transactions before you. I mean, that's exactly the same problem as on Ethereum, right?

I mean, because it's a state full chain, you don't always know what the state is that your transaction is going to act upon. Right, but I think over there, the Turing completeness actually adds a lot of ambiguity, right? Because in theory, like, it's a Turing complete language, so it's very hard to actually do formal verification on it. It's impossible to upper bound the gas costs. While here at least you can buy, you know, you can take all the different code paths and you can at least say here is the like upper bound by like this like, you know, provably this is the upper bound which is I guess not possible in a true incomplete language.

Right. Can I go back one step? Seeing that clarity is not a pure incomplete language, what are the sort of computations that clarity cannot do? when compared to something like solidity. Yeah, I can give a very simple example, right? For example, it doesn't have loops. But interestingly, you can write most programs without loops. Like, you can just write the same type of logic without using loops.

And then we get a lot of questions around, hey, can you actually call other contracts? And the answer is yes. You can call other contracts, but there's no re-entrancy, right? Re-entrency does not exist. So it's a little bit like, you can even, like, if you really, want quote-unquote loops, you can actually write a separate contract and then just like call it a bunch of times. So it's a like the developers need to like get into the mindset of like how to write this code. But once you pick that up, like, we actually ran an experiment, right?

Like for four months before launching Saks 2.0, we are doing a bunch of hackathons with the goal of like let's try to get developers to build like some of the more popular. smart contracts on Ethereum, mostly from the lens of, like, do we need to change clarity in some way or add certain features or someone? And interestingly, the developers were actually pretty easily able to build literally every major celebrity contract, right? And yes, it was informative, like we wanted to add certain changes. And in fact, we ended up doing something pretty interesting because a change to clarity language is actually a hard fork to the stack of blockchain. So we don't expect clarity to change that much.

But we were able to figure out this very interesting way where you could implement a clarity feature or enhancement in clarity itself, which is not a hard fork, it's just that that function will be slow. And if that thing is getting usage, then in a future hard fork, you could have a native implementation. So one of the learning from this hackathon process was that,

okay, maybe there's something that clarity doesn't have right now and people feel the need for it. Well, maybe they can just write it in clarity itself. And down the road, you can optimize it by making it faster. Do you foresee that clarity will become like the language that people write in? Or do you think that there's going to be higher languages that then compile down to clarity? Because I notice in your GitHub you have a language called clarity script, which is actually like a... Like a JavaScript, yeah. JavaScript-esque language that compiles to it.

I looked at clarity, and it seems to be like LISP language, which is like, you know, has all these nice properties. But at the same time, it's like definitely much harder for like beginner developers to approach. And so what's the strategy there? My preference would be that people just learn it, right? Because it takes a little bit, a little bit of time, but I can see why there can be a learning curve as well. Like the way I'm thinking about this is to write. smart contracts, you have to go through a learning curve.

You have to understand what the smart contracts are, right? I know that there are approaches in an industry where they're just trying to use like Rost as a programming language or even see as a smart contract programming language. I just disagree. I think people actually in the process of learning why clarity is different, like they would actually learn the reasons for how smart contracts are different from normal programs. But realistically, so ClarityScript is not the only one. I think there were some people in the R-Veeve community

who were actually trying to write similar like Jala script to Clarity, trans-compilers and so on. So I think that's already happening, right? People have their preference for what language they want to use, and there might be like these compilation step, right? Like you go from that language to clarity. But I think in terms of what I would like to see is I would like to see more education material of when someone is just learning how to program,

like telling them that why doing things this way is different from doing things in the other solidarity type world. Now going back a little bit to architecture of the chain itself, one of the main sort of taglines that's used here is that it's built on Bitcoin, stacks is built on Bitcoin. And so after going through this whole discussion about the consensus protocol, I definitely see in why this is sort of inherits its consensus security from Bitcoin in more in a way that something like a separate chat network like Ethereum does not. But from an application perspective, does a stack smart contract have any different relationship to Bitcoin the asset than an Ethereum

smart contract might? Because with Ethereum, we have, we have like wrapped Bitcoin or like, T-BTC on Ethereum, and we've had BTC relay for, you know, number of years at this point. And so from that application level perspective, how is this more Bitcoin native than something like Bitcoin on Ethereum? So that's the second biggest thing of clarity, right, that I was going to get into. And that is that clarity has full visibility into Bitcoin state, right? So clarity has native Bitcoin SBB proof. and it has, just like I mentioned earlier, that we can use Bitcoin as a backup channel to even do

transactions, like Stacks transactions that show up in consensus on the SAC chain through Bitcoin. Similarly, whatever is happening of Bitcoin is available to programmers in clarity. So let me walk you through a very simple program where, you know, let's say I want to do donations to Coin Center. and I say that, you know, here's like 20,000 stacks. Like up to 20,000 stacks, I want to donate to Coin Center. And based on how many Bitcoin donations go to this Bitcoin address, I will match an additional amount.

And I can write that as a Clarity Smart contract. People are sending transactions to Bitcoin address for CoinCenter. Right. They don't even need to know about clarity or stacks. The stacks contract automatically knows the state of that Bitcoin address. And if it triggers that, hey, now X amount of Bitcoin has been donated, it can send a transaction on stacks for the additional donation to CoinCenter. So you're writing logic that is triggered by changes on Bitcoin.

That's, I think, super interesting. And then you can look at the reverse of that as well. So this functionality that I describe, like it's live today, right? Like you have full visibility into Bitcoin, you have Bitcoin SPV proves natively available in clarity. But the reverse path is actually even more powerful where a clarity contract can trigger a change in the Bitcoin state, right? That work is more in the R&D stages.

And we are looking at various Bitcoin scripts. where, you know, let's say an ideal example would be you lock up your Bitcoin and name some sort of a time lock. And there's logic that executes on the clarity side that when that program, clarity program basically finishes executing, it can actually trigger where the unlocked Bitcoin on the Bitcoin chain is going to go. Once you have that, now you have right capability to Bitcoin as well. That is super powerful. And so what are some of the like techniques being used? used, you know, even in the R&D stage to, like, do that. Yeah, I think, I think that's super early.

We were, we were focused on building, building out this tax group. I know, made that at this point. But, you know, we look at, you know, hashtag logs. We'll look at tap root. We'll look at, you know, there are a bunch of, like, interesting other newer projects as well. I know, I know Jeremy Rubin has been working on, like, some, some sort of more advanced scripting language on the Bitcoin side. So we are exploring all of that, basically see what can be done to, in a way, kind of like connect clarity with state changes on Bitcoin.

And then going to the example of kind of like Ethereum and wrapped assets there, it's a little bit like if you are moving a wrapped asset on a disconnected chain, like it has nothing to do with Bitcoin at that point. Any asset that you're creating on stacks is effectively secured by Bitcoin. Let's take the example of an NFT. There's an app called Boom. People are creating NFTs there. Those NFTs are actually literally secured by Bitcoin. You can see that here is the Stax transaction. Here's how it got packaged up.

And here are the hashes on the Bitcoin chain for it. So I think in my view, an NFT issued on Bitcoin is more valuable because it's more durable. And similarly, other types of digital assets issued on the Stax chain. chain, they're directly getting secured by Bitcoin as well. But so that means, though, at the moment, there's no way of transferring Bitcoin onto the Stax chain. Because I think one of the main, like, things that people are, want right now is the ability to use Bitcoin in DFI, use it as collateral in different DFI things.

And so right now, Stax doesn't enable that yet until the work on, like, having the other direction is completed, right? Yes. So I think what's happening there is there's a spectrum of solutions, right? One is the custodial approach. Just like you have wrapped Bitcoin on Ethereum, Tokensoft and Anchorage have announced that they plan to have a wrapped Bitcoin asset on stacks. And again, the only real benefit you're getting is that asset itself is secured by Bitcoin. So if you don't trust the consensus or security of other chains, at least you have the peace of mind that this. this thing is being being secured by Bitcoin. The next level from that is keep network type of approaches, right, where there's some sort of a more decentralized way of moving an asset into the chain. But the extreme end of the spectrum is the ideal world where the user experience is, you just lock up Bitcoin on Bitcoin and use it on stacks.

And then when you're done using it on stacks, they just unlock on Bitcoin. Right. Like that's the, that is the dream world, right? Like where that's the thing that everyone wants, where it's like a more seamless kind of like, I just locked Bitcoin on Bitcoin, used it in the in the transaction, took them back or they go to somebody else if, you know, the ownership of Bitcoin has actually changed hands on during the time it was in the transaction. So the other offering that builds a similar offering on top of Bitcoin is, root stack, right? So how does that compare to stacks? Yeah, I think, first of all, I love to see innovation around Bitcoin, the people who are trying to do different things on Bitcoin. That's always great and great to see. The two big differences are the consensus mechanism over there is merge mining, right? And I mentioned our experience with merge mining on Namecoin, where because the merge mine chain is,

unless you somehow get 100% of the Bitcoin miners to also mine the merge chain, the merge mining chain is always like a subset of the main chain. And then another kind of like thing that happens is, let's say you have a large Bitcoin miner that has 35% or something of Bitcoin. If that miner starts mining on the merge mine chain, that minor might actually have 60% of the merge chain.

Like, we went through those problems. I found some issues on NameQuine where there was literally a single minor, like, having 75% of the hash power. And when I, when I disclosed those issues, like, somehow, you know, the community members were like, we, yeah, we have known about this for a while and that's fine. And, like, how is this fine? Like, this is at this point, a single miner is effectively running the network. Right. So I think there are challenges. And then there's practical challenges where off the merge mine chain.

Like, it's about incentives. Like, what is the incentive? for a very large fraction of the Bitcoin miners to continuously mine this chain, especially when I think the way my understanding is RSC also doesn't have coin-based rewards, right? So they're really saying that the transaction volume on our chain is going to be so high that every single Bitcoin miner will want to mine our chain, right? That would be a great world if we can reach that, but I think getting there has some sort of,

initial hurdles and incentive problems. And secondly, I think the big difference is that they basically have solidity and EVM, right? And I would not recommend, you know, young entrepreneurs and engineers who are basically trying to learn program smart contracts to go down that route. Ideally, you know, and many other projects are doing that. We're not the only ones, right? If you look at Tesos, if you look at Algorand, Algorand actually also collaborated with us on clarity. like they're in the same camp that these need to be non-tiering complete, much more focus on security type of languages, not EVM-based.

I think what happened is because Ethereum was early and it got the initial kind of like adoption, now there is a little bit like, hey, let's just be compatible with EVM because that's like an easier sell to developers. But if someone made the bad decision like in the early stages of the industry, doesn't mean that the industry needs to live with that. for forever like I'm talking 10 years down the road like it'll I think it'll be a at least in my mind it'll be a pretty pretty sad outcome if people are still fighting kind of like re-entrancy and all these types of like problems in smart contracts instead of just using more more secure languages there are so many options available clarity is not the only one right Zilika has it has an interesting approach Tesos has an interesting approach algorithm as I said they have a lower-level language and now they're working with

clarity as a higher level language as well. Cool. So let's talk about some of the apps that run on top of Stacks 2.0. Can you give us an idea of how many there are and kind of what the popular ones are? Yeah. So I think, first of all, there's a divide between Stacks 1.0 and 2.0. 1.0 did not have smart contracts. So 1.0 is a chain that was directly on top of Bitcoin.

and over there, effectively, the use of the blockchain was effectively to register usernames, domain names, and pointers to storage, right? And then we had a decentralized authentication system. We had the rest of the stack so that people can build things like, you know, decentralized blogging, decentralized photo sharing, like even messaging applications and so on. And we saw a bunch of them. A lot of the attraction goes back to our app mining.

program, which is a very interesting experiment. Like, we learned a lot from it. It effectively was a more automated way to incentivize developer behavior. Let's say that there are certain metrics based on which you're handing out rewards to developers, right, to kickstart kind of like an app ecosystem so that people are building more applications. Interestingly, what we saw was initially it was working really well, right? Like developers would basically come and build these applications.

Like, we can even score them on decentralization, right? Like, hey, how decentralized is your app? Or you can just put in, like, or you can score them on the UX of the app. We had this process of like multiple reviewers. So I say there are five independent reviewers who are giving a score on the UX. There is a score for how decentralized the app is. There's a score for, you know, how much users an app is getting and so on. And very quickly we started seeing that there was almost like some sort of unintended behavior.

where developers were more interested in just gaming the scores for the rewards, versus building, you know, really high-quality, decentralized applications for love of it, right? And looking back, it might actually seem more obvious that, of course, they were just, it's like an exam, right? Are you taking the class to actually learn, or are you doing it just to pass the test, right? And there are so many students who are just focused on, like, I just want to clear the exam and get a good grade. I don't care about the learning part.

For those reasons, we ended up pausing the app mining program. And it got us like more than 400 applications built on top of the network. Some of them were really high quality, really good apps. But then there was a long tail of people who were effectively, you know, some people would build like 10 applications. They can have 10 entries in app mining and kind of like get more rewards and whatnot. But they're not really passionate about any single one of those applications. So the app planning program has been paused for almost a year now, more than a year.

and we have definitely noticed a difference that there's more organic community like during the time app mining was running like we got a lot more attention there were a lot more people but they were basically there for the money right like a good good quantity of them but now the community feels much more organic

the developers are there because they believe in decentralization they believe in privacy they believe in building on bitcoin right like building on bitcoin is a very core kind of like cultural DNA of the community that we're building. And I'm super excited about it. So in terms of apps, you can just go to app.com that has a list of some of the top applications from Stax 1.0 that are also migrating to Stax 2.0.

They need to upgrade certain things in the app to migrate to 2.0. Pravica is a great example. Pravica is almost like a decentralized zoom plus a decentralized WhatsApp. So you can do messaging and video calls. and it's amazing, right? Like it's like you look at what's happening in the world and these are the kind of solutions you're looking for where there's no company in the middle.

There's a completely encrypted video chat session and you look up the keys for people using their usernames that are registered on the stacks chain, but then secured by Bitcoin. So with your private key, you can actually just see in which transaction your name was registered, how was it packaged on Bitcoin, and you know that no one is going to change the Bitcoin chain.

And then in the newer applications, because Clarity just went live, we're actually seeing there is a project called Swapper, which is like an automated market maker, like Uniswap, written in Clarity, which would be super interesting to see there's a project called DualX. It's like you can take an option to buy an asset and trade that. Right. So people are trying to now innovate more on the. the defy side of things.

And I think the key differentiator, and going back to what you mentioned earlier, that the blockchain has only been live, 2.0 has only been live for like two weeks now. When we can have like more native Bitcoin movement from the Bitcoin chain into the stack chain and back and so on, and I think it might happen over phases. Like, sure, you'll have a custodial approach.

You might have a semi-decentlyze and then a fully decentralized approach and so on. But that means you can feel. inject Bitcoin capital directly into Defy. Written in Clarity, which is a secure programming language, and all of this is happening directly on top of Bitcoin in the sense that, you know, all the state for these contracts is eventually getting settled back into Bitcoin. And I think that is a world that I'm super excited about. If you look at the Bitcoin community,

there's a lot of resistance to any sort of token that's not Bitcoin. right? And a lot of the apps that run on top of stacks 2.0 have native tokenomics. So how do you see that playing out? Do you think tokens, I mean, basically in many Bitcoin circles, they are decried as, you know, this money-grabbing scam. So how do you kind of see these two come together? Do you see these two come together? Yeah. So I think, let me tease that apart a little bit. And actually, even before getting into that, like, there's one important kind of like thing that we have completely missed from the discussions right now. And that is the ability to earn Bitcoin rewards from consensus.

So the Stacksholders, when mining is happening, every Bitcoin block, literally Bitcoin is being transferred to the Stacks folder. And it's a, it's a magical experience, right? Like, there were people tweeting about it, they're showing their screenshots of basically getting Bitcoin in their wallet. They can give any wallet address. It could be an exchange or Coinbase or however, desktop wallet, cold storage. And there's an economic tie between stacks and Bitcoin. Like in the sense that if there are more smart contracts that are gaining traction on stacks, the Bitcoin yield is going up for these tax holders.

So it's a little bit like, imagine it's a, we've seen in the crypto industry, this tribalism, especially like, you know, between Bitcoin and Ethereum as well, where sometimes you'll see the base like, hey, Ethereum is going to be more valuable than Bitcoin. Bitcoin can just be recreated on top of Ethereum and so on and the other way around. But Bitcoin has actually never had the capabilities that Ethereum has. It's mostly a passive asset like gold that people are kind of like this buying and it's just sitting there. We are not only bringing the smart contract and app functionality to Bitcoin. We're doing it in a way that doesn't compete with Bitcoin.

It can grow in harmony with Bitcoin. Stax is never going to say that Stax is going to replace Bitcoin. It's not even possible from a protocol perspective because you need the underlying reserve currency and the reserve asset to actually settle to it, to actually benefit from it, and then Stax has Bitcoin rewards, which are tied to smart contracts. like in the sense that if smart contracts are taking off, these Bitcoin rewards are going up and people are kind of like, you know, more interested in the long-term success of both Bitcoin

and stacks. Right. So let's have that framework in mind and then dig deeper into the Bitcoin community who have, they have resistance to new types of assets. Like there are people without naming them who have flat out told me that they love every single thing about this project minus the stacks. cryptocurrency. They love everything, right? It's open source, it's decentralized, it's bringing, you know, apps and smart contracts and all this innovation and all of that. And, you know, they praise

even the work we have done on regulation, that this and that, but they have some sort of a mental barrier to a separate asset. And then I, then I, then I basically walk them through the journey. And like, why don't you fork our project and run it without, without the asset and basically see what happens. So what would happen is, that all this state for smart contracts that is being secured outside of Bitcoin, no one would have any incentive to store that state. So then you're talking about can we shove this state back into Bitcoin, which is not going to happen.

It's a scalability problem for Bitcoin. And then also, I think incentives matter a lot. Like we just launched and their miners were competing over mining blocks. Like we're not in a situation where we are begging people that, hey, can you please come and run nodes on this chain, out of the goodness of your heart, they're doing it because they are financially motivated to do it. And it's the same thing with Bitcoin.

How do you think Bitcoin took off? Like it has incentives. It pays the operators of the network to operate the Bitcoin network. And similarly, you need to pay the operators off of the Stax network to operate these smart contracts. And I think the other interesting angle with Bitcoiners is you would ask them that do you think Bitcoin is a settlement layer? and they'll be like, yes, Bitcoin is a settlement layer,

and Bitcoin should settle everything in the world. I'm like, okay, great. You know, Stacks settles on Bitcoin, and if there is a token that is issued on stacks, that token also settles on Bitcoin. So I think there's a little bit of an inner conflict where you cannot at the same time believe that everything in the world should be priced in Bitcoin.

Everything should use Bitcoin as a settlement layer, and at the same time deny the existence of those assets that should be priced in Bitcoin and should settle on Bitcoin, right? So once you start walking through this, you see the conflict. And also, like, I shouldn't say that this is the same reaction from all of the Bitcoin community. Like, I think even the Bitcoin community is a spectrum. There are some people who are extremely bullish and extremely helpful and extremely excited

about what we are doing. Then there are people who are like, hey, they kind of, accept it. They're like, you know, I don't like the token, but I, overall, what you guys are doing is, like, really beneficial and I kind of like it. And then there's the extreme form where they're like, the token, I just hate it enough that I'm going to hit the entire project because you have a token. So I think it's a spectrum, but over time, I think especially with more education, especially when people realize, like imagine how many people are going to discover Bitcoin because they're going to earn their first Bitcoin through stats or how this is actually making Nick Carter

was on a podcast with them. He made a really, really important argument that stacks miners are willing to pay really high Bitcoin transaction fees because the value of a tax block to them is way higher than a normal Bitcoin transaction, right? Which is actually a very sustainable model for Bitcoin, especially if all of these Bitcoiners, they would love to see Bitcoin. at 50,000, 100,000, a million, right? What is the thing is going to happen with Bitcoin fees when that happens, right? Bitcoin fees are going to skyrocket. And to sustain the fee market, the only sustainable path is that you're actually packing

so much valuable information in a Bitcoin transaction that it is okay for someone to pay $100,000, $1,000 transaction fee. Our miners will gladly pay a $1,000 transaction fee to just get their transaction. in. They actually do very aggressive, replaced by fee. They're always bumping up their fee and updating their transaction because they really, really want to get into a Bitcoin block. And I think long term, like, that is the kind of use that can actually justify a very high price of Bitcoin as well, because you're packing so many information.

I think this is a perfect note to end on. I think this kind of sums up your entire offering, really nicely. So can you tell our listeners, where should they go to find you and to find information on your project? Yes. So I think I'm fairly active on Twitter and it's just my first name at Munib, M-U-N-E-E-B. And then the project is on stacks.com. And there are actually like multiple entities in the project that you can then discover. I run a developer-focused company called Hero. There's a nonprofit called Stax Foundation. There is a community-focused entity freehold. There are mining companies like Damon and clients like secret key labs or new intern labs and so on.

So the project has really grown over the years. It's very decentralized. It's very much in the cap of Bitcoin. It shares a lot of the DNA with Bitcoin and it's effectively trying to bring more innovation to Bitcoin. Perfect. Thank you so much for coming on, Muneup. Awesome. Thanks for having me. Thank you for joining us on this week's episode.

We release new episodes every week. You can find and subscribe to the show on iTunes, Spotify, YouTube, SoundCloud, or wherever you listen to podcasts. And if you have a Google Home or Alexa device, you can tell it to listen to the latest episode of the Epicenter podcast. Go to epicenter.tv.tv slash subscribe for a full list of places where you can watch and listen. And while you're there, be sure to sign up for the newsletter. So you get new episodes in your inbox as they're released. If you want to interact with us, guests or other podcast listeners,

you can follow us on Twitter. And please leave us a review on iTunes. It helps people find the show, and we're always happy to read them. So thanks so much, and we look forward to being back next week.