S17 E43: Mike Belshe on BitGo, Quantum & Clarity

0:46Vlad CosteaHey, hello there, and welcome to the Bitcoin Takeover Podcast. This is Season 17, Episode 43. This episode is sponsored by Cake Wallet, SideShift, Braiins Hashpower, Orange Rock, and LayerTwo Labs. And we are live with Mike Belshe, CEO of BitGo, who's returning on the show. It's been 18 months or something since you were here. And so much has happened in the meantime. So it's really good to have you again.

1:12Mike BelsheHey Vlad, great to see you. Always love to see you. So I'm sorry it's been so long, but glad to be here.

1:18Vlad CosteaYou know, I saw you at the White House about a month ago when you were there with Brian Armstrong and the Winklevoss twins. And at the time I was trying very hard to get you on the show. And when I saw that, it was like, oh man, this guy's hanging out with the cool kids. He's not going to join my podcast again. It's like I'm throwing this small party and then you're at the biggest and most extravagant party. And I'm like, probably not going to happen.

1:44Mike BelsheWell, it's an honor to be there individually, but for the company, for the industry, I guess a few people get to kind of represent there. So anyway, it was fun and interesting to see kind of that side of things. But also, look, there's been a lot of hard work by the industry over the last several years. And obviously right now we're supposed to have clarity vote today, I guess, for whether it's going to go to cloture, for cloture to go into the next vote, I guess. But that is hard work.

2:19That is not just accidental. So there's a lot of people that have been behind it. And yeah, look, it's exciting times. We'll see what happens today.

2:27Vlad CosteaSo for those of us who are not in the US, why is the Clarity Act such a big deal?

2:34Mike BelsheIt's interesting. If you had asked me even just a year ago about how much do we need the Clarity Act, which is market structure for digital assets from the legislative point of view, I would have said, okay, it's good, probably should have it, help prevent FTX, things like that. But I didn't think it was really as required. However, so BitGo, we just went through the IPO process in January. And so you do the roadshow before the IPO, but before that you're talking with investors, basically raising money through Wall Street.

3:07These are investors that have been on the sidelines for crypto, digital assets, Bitcoin for a long time. Literally, they're asking questions like, what is a private key? I mean, very basic stuff. You find yourself paging in like this old information of things you used to share back in like 2015 and 2016 about how blockchains work and whatnot. And it shouldn't be that surprising. I mean, those of us that are in the industry, you forget, like we talk about all these topics and we go from the 100 level course to the 200 level course to the 400 level course.

3:38But there's a whole slew of people around the world, especially those that come from more traditional places that really don't understand how this works. They've been hearing about it, saying, interesting, maybe someday. All right. So all that education has to happen. And during that roadshow process, they're like, well, you must be really excited, that clarity, that must really help your business. And I was like, why do you think that? And you start to realize what's going on without a legislated path forward. They still have this feeling of like crypto Bitcoin is this wild west, has no chance of being adopted or used in a major way until it's got some sort of a sanctioned path from the legislation.

4:23So anyway, I believe this is true. And if you go inside of large banks, large regulated institutions, et cetera, like there's exactly three groups of people and how they feel about Bitcoin. Those are like, they love it and they're ready to go. Those that hate it, they're like, ah, you shouldn't kill it, whatever. And of course there's a bunch of undecideds, but all those undecideds, as long as there is not a legislative path forward that holds market structure for digital assets, how it's traded, how do you do derivatives, all that kind of stuff, they will continue to be persuaded by some of those naysayers that hold up, ah, it's not regulated, this is too wild, we can't touch this, it's too risky.

4:59What happens when there's a new administration comes in? So look, Europe, I think is ahead of the US right now in terms of having just a clear path of how do you build digital asset companies, products, financial services. We see it in South Korea, they're moving ahead. Of course, many countries all over the planet also have legislative paths forward. It's time for the US to join the crew. And instead we've got infighting over, we think Trump made too much money on some Trump coin. So anyway, we'll see what happens with Clarity Act.

5:34Of course, we have market structure for how we trade these things. I mean, we don't want to have another Sam Bankman-Fried. And if you don't pass some sort of market structure, you are at risk to have that again.

5:46Vlad CosteaDoes this affect the way in which Bitcoin is used for payments in the US?

5:53Mike BelsheI don't think it has any impact on payments, no. So this is really more about trading of tokenized assets. Bitcoin, I think actually is mostly through. So if you want to just talk about the Bitcoin side of things, I think from the SEC, the CFTC, Bitcoin is considered to be sufficiently decentralized and therefore it doesn't fall. Nobody has any question that it's a security, it's not a security. Maybe it's a commodity. I guess how you tax it becomes a little bit of a thing, but I don't know if Clarity would solve that either.

6:29So for the Bitcoin side, it's in a pretty safe place and then payments are different, right? So payments fall under and you got money transmission a little bit, which I think is still a bit annoying and outdated, but not nearly as crucial as the trading infrastructure for how we take these products into the markets.

6:50Vlad CosteaYeah, that's interesting because I've seen a lot of people on Twitter and of course, this is not proof that anything is well researched or whatever, but they were talking about the Clarity Act in the context of Bitcoin adoption and why it's so important for this to pass in order to be able to make payments at merchants and whatever.

7:09Mike BelsheI don't think it's going to affect payments, but it does affect the ability for large institutions to be able to touch the asset. So at this point, I mean, you can see like Morgan Stanley, Charles Schwab, Citibank, JP Morgan, BNY Mellon, all of these guys are now in on Bitcoin. They were not a couple of years ago certainly in on it, but I think they feel pretty comfortable on that. However, inside of all these institutions, a big part of the crew that works on the regulated institution that they have, and if we have a different set of politicians with a different interpretation of the exact same law, because there is no specific clarity law, then it's uncertain what happens or what those entities could do.

7:58So remember, when we had Operation Chokepoint just 24 months ago, the FDIC, which is the Federal Deposit Insurance Corporation, but it's insurance for banks. The guy that was running that was a guy named Martin Gruenberg, and all he had to do in order to prevent Bitcoin adoption was say, I'm going to scrutinize your business if you touch crypto. And he meant crypto or Bitcoin or anything. And then all the banks are like, I don't want that. And they were so scared of their regulator that they just backed away from all of this stuff.

8:32So without clarity, we are vulnerable to that happening again. And look, when it comes to payments, of course, it's got some overlap with banks. Obviously, Bitcoin allows for a decentralized type of payment. But I think when you're talking about payments at scale, you're going to have to interface to the banking system. So clarity would definitely help in that it would take away this possibility that a new administration comes in and just overrides all Bitcoin rules.

9:04Vlad CosteaNow, it makes a lot more sense because I hear a lot of people saying, well, we only have two more years and then it will be Operation Chokepoint all over again. And it seems like we had this time of relaxed, you know, SEC and whatever. And people were able to build whatever they wanted in the space, except for privacy. For some reason, that one is not as, I don't know, accepted, tolerated. We still have people in jail for running privacy services. But other than that, it seems like it's a pretty good time for crypto right now.

9:40But people are worried that in two years it's going to reverse. There's going to be AOC who comes into office and does whatever.

9:50Mike BelsheWell, we'll see. I'm not as worried about that as others are. I do think we've actually moved the ball down the field just in terms of people understanding digital assets, Bitcoin, etc. I think a lot has happened and more is going to happen. All right. What they're worried about, though, remember, under Biden, like all of the agencies, you had Martin Gruenberg at the FDIC. I mentioned him. Of course, we had Gary Gensler at the SEC.

10:19Who was at the OCC, what was his name? It was Michael Hsu, right? Michael Hsu, he wanted to be, I think he had ambitions to be a bigger role at Treasury or whatnot. But all these guys were put in place. They were all negative on crypto, and they kept everything down. Then as soon as Trump comes in, what does he do? He marches in with a bunch of executive orders. Now, that's not specific to Trump. The next president can come in with a set of executive orders as well.

10:46And when there's not legislation, you know the executive orders can kind of seemingly override everything. But remember, we actually have now a whole bunch of entities that are already engaging this activity. Morgan Stanley, Citibank, like Charles Schwab, Fidelity, BlackRock. They're already in digital space. And forget about what you think about politics. Like, you know, Larry Fink, who's been running BlackRock for I'm not even sure how long. He says everything is going to be tokenized. This is not a meaningless statement, and it's not a meaningless statement for regulators.

11:22All right, so between now and the end of the Trump administration, and I don't think that they're crypto friendly. I think all they are is their level, even keeled regulators. Remember, a regulator's job is not to determine I like what you're doing or I don't like what you're doing. Their job is to determine whether what you're doing is meeting the regulations that have been set forth by the law. That's it, right? And they're supposed to be very level headed, and they're not supposed to be charged at all.

11:50The idea that they're charged is really the bug. So what we have right now is a set of administrators that are willing to take a look at what you're doing. And then hopefully their job is not to say, Oh, gotcha. Now I'm gonna fine you or throw you in jail instead of, should be, oh, wait a minute. This is not up to snuff. Let's look at the intent. All right. The intent is good, so let's just help them fix it, as opposed to like, we're gonna come after you with enforcement.

12:18All right, so we have that today. I think they're level headed towards crypto. It's allowing us to all move forward. By the time we get to the end of this administration, we're gonna have a whole bunch of rulemaking in place. It will be in from the SEC. It will be from the CFTC. It will be from the OCC. So the number of major financial institutions touching Bitcoin is going to be huge. And the number that are touching other types of digital assets is also going to be very, very large.

12:47And the markets are probably gonna be pretty significant. So good news for Americans. Like once those things are in place, it's difficult for any future president. Just like yank the banking system out. Yank the trading system out. Yank access to your money from all these Americans that are going to have digital assets. So I think we're probably gonna be okay with or without clarity in terms of not having an administration rug pull us. However, The market structure is important. If you believe that the U.S.

13:21Has the largest capital markets in the world. And when I say believe, I think it's measurably the best capital markets in the world. It's not even close here. We talk about China versus the U.S. from a GDP competition. That's one thing. But from a capital markets, there's no doubt that the U.S. capital markets are the strongest. One of the things you build with market structure is robustness. It's just like building a complicated server infrastructure. It's a how do you put together these module these systems so that you know when a pricing feed goes down, the market doesn't blow up that when this other component doesn't go down like the whole thing doesn't go down.

13:59And when you look at the U.S. markets now they only run 9 to 5 Monday through Friday. But that

14:06Vlad CosteaWas gonna be the next topic, actually, because I've seen the impact that we've had on the markets and the fact that now they talk about tokenizing stocks to do 24 7 trading

14:16Mike BelsheCrypto inspired. Look, perps, I think are a fantastic example. You can have purpose without 27 24 7 markets. And these all come from the crypto industry, and they're absolutely a better product for those that are looking to have hedging capability across any type of asset. So, yeah, look, there's tremendous innovation. Anyway, just to finish off that thought on market structure. When you're building market structure protects against risks. If the capital markets of tomorrow are going to be digital and everybody in traditional finance, not the Bitcoiners.

14:53All the traditional finance guys are saying this is true. Then, of course, it has to be protected from any single party blowing up and failing. And that's what market structure is really about. So like, I think there's very little risk that when this presidency ends that there'll be just a massive rug pull like what we had prior to this administration. However, yeah, as the US capital markets shift more and more digital, we need these protections in place. And due to political infighting, the lawmakers right now are struggling to do what's frankly common sense.

15:34Vlad CosteaYeah, it really sounds good. I hope that the best outcome is what we're going to get honestly, because it's been such a roller coaster in terms of regulations concerning Bitcoin and other digital assets. And sometimes it's good to set a precedent for the rest of the world here in Europe. It's only getting worse with MiCA and whatever. It's really absurd. I've seen shops that used to accept, for example, Bitcoin payments. And then now they require them to get a license, which costs them like 10,000 euro or something and lawyer fees.

16:11And most businesses are like it's not worth it.

16:16Mike BelsheAnd it's unfortunate. Look, I think on the Bitcoin payment side. I don't know. My own personal view is that stablecoins are going to fill the payment gap that is digital. And I think Bitcoin is going to end up serving kind of a different purpose at this point. It's the store of value. It's the non fiat network. In terms of payments, I think it's going to be a while before people want to pay using that. Just like we don't pay with gold. You can pay with gold today if you want it, or you pay with a tokenized gold.

16:47People are going to prefer to pay with a tokenized. Dollar tokenized euro, etc. And then when it comes to digital assets, the regulators are all kind of gone bananas about what regulation is required just to move money between two parties. If you and I want to send some money between each other, we ought to be able to send that money. And if there's an ATM at a store, I ought to be able to use that. But frankly, this is part of a bigger problem, which is the war on cash broadly.

17:18So cash is kind of a thorn in the side of regulators because they really want to have control over all money movements. They want to see everything that you buy, everything that you spend and have that power for whenever they need it. And cash is an anonymous form of payment. And so you probably know this. Like in France, you can't pay more than a thousand euros in cash for something. You can't buy something with more than a thousand euros. You know that?

17:42Vlad CosteaSame all over the European Union. I think they tried to harmonize. That's what they say. Harmonize the legislation all around the 28 EU countries.

17:53Mike BelsheTake the most restrictive rule and apply it everywhere. Look, this is this is I think the real issue is they want to take away your ability to have private monetary transactions and they don't like cash. So I just saw that there was here in the US, some city banned like Bitcoin ATMs and then people were like, this is anti-Bitcoin. And my response is like, no, this is anti cash. You still buy Bitcoin. They didn't prevent you from doing that. They're just preventing you from paying for it with cash at an ATM.

18:27They don't want the ability for you to just go and take your cash and convert it into Bitcoin. They're perfectly fine with you going to an exchange and doing that same thing.

18:40Vlad CosteaYeah. Just a side note, I think you forgot to retweet the podcast and maybe that part of your audience would love to join in and maybe write some comments in the chat if they want. Also, I forgot to tell you that it could be a good idea to write down the number from one to 100. And there's going to be a contest in the chat. The person who guesses that is going to get $100 from the sponsor Cake Wallet. $100 not in cash, but in digital cash of their choice.

19:09It can be Bitcoin, Monero or Zcash.

19:15Mike BelsheAll right, you are retweeted.

19:18Vlad CosteaAnd don't forget to write down the number if you have a pen and paper. Don't reveal that to me. I'm not supposed to know.

19:28Number one.

19:32Sure. I'm curious how many people will say one in the chat right now. Anyway, you mentioned perpetual trading. I also need to mention that I have a sponsor called Orange Rock, which is handling that. And you can do trading 24/7 without KYC, without any limits to volume. And it's basically an interface for Hyperliquid, which is another innovation of this space, except that this one works on a very nice user interface like an everything app. And it's built by the Bitcoin.com people. You can find the app.

20:05It's called Orange Rock on iOS today. And I also need to ask you about BitGo's business right now, because it's been such a wild 18 months. And last time you were on the show, you were talking about self-custody and how BitGo started as a company that was helping customers with self-custody. But right now, I don't know if it's ironic or it's just the natural evolution of things, but you're like the largest Bitcoin bank.

20:34Mike BelsheLook, actually, I think it's the right type of bank for the future. So yes, we support custody, but we also support self-custody. I think we're the only ones that do both. I'm actually surprised that it's something people don't see. So with BitGo, by the way, we just opened it up for all individual access. If you want to try it, please do. You can use it for self-custody. It doesn't require full KYC for that. That's self-custody. That's a software component. If you eventually decide, you know what, I want to put some of this in custody, you fill out the extra steps to go through the custodial part that does require us to know who you are.

21:15But then you can have both custody wallets and self-custody wallets side by side. All of the wallets, whether they're self-custody or custody, are multisig or MPC with a two out of three model for security that would protect against things like some of the Coldcard stuff. Happy to go into details on the security of how that all works. But the idea is that you should have control over your money. And, you know, some amount of self-sovereignty may be something that you value and want.

21:43And then some amount of hiring a bank is something that you want. So, look, I mean, historically, if you just go back 150, 200 years, there was a time when people actually didn't use banks for all of their savings. It was obviously a very different time. It was not electronic time. But do you think that banks kind of right out of the gate were like universally trusted and there were no rug pulls? There were rug pulls. You know, there were banks that would establish and say, hey, we'll store your money and keep it safe.

22:12And then one day you wake up, you go down to the town square and the bank is like empty and the guys ran away with all the money. Absolutely happened. And it took a while to build that into a trustworthy system. You know, banks, you know, then went through an era where they established these big buildings with marble pillars and like all kinds of permanence. And the reason was to show that the bank was investing locally and they weren't going to run away with your money.

22:40So anyway, the same thing is happening right now, I think, with digital assets. They used to be a pretty small piece of people's net worth. When they get to be worth more, like every individual and there's no judgment, has their own thinking about how comfortable they are with what they hold on their own versus what they want to put into some sort of institution. And it's a trade off that you have to make. Like, you know, on one hand, when you hold it on your own, you get the self-sovereignty part of it.

23:08That's great. Nobody can rip it apart from you. You don't have to worry about rug pulls. You don't have to worry about security hacks. You don't worry about any of that. At least not from the third party. Instead, you put that on yourself. Right. And are you going to hire guards? Are you going to put bars on your windows? What are you going to do? And then with custody, of course, you've got the opposite side of that. So Look, BitGo provides both. We will continue to provide both. We provide a very, very large spectrum of coins.

23:33I think it's the largest coin support in the world. And yeah, that's what we do. So by the way, it's completely free for individual users. You can use the self-custody or the custody. Let us know what you think. If you've got feedback, always open to that. Always trying to make the product better.

23:50Vlad CosteaIt's kind of amazing that you're not charging at least the amount that is the price of a Coldcard. Just to be like, Well, you can buy the Coldcard or you can use our service.

24:02Mike BelsheWell, look, we're fortunate enough. We've been in the space since 2013. So we built a number of services on top. It's true. Like if you're just using the self-custody, like, you know, we're not gonna make any money. But hopefully some number of people are like, Hey, you know what? I need some of this custody component. They move to that. We've also got the best trading rates. We've got staking and other things. So we make money off of services that way. But we're using those to be able to allow people to use the basic service without charging.

24:30When it comes to Coldcard ledger as well. You know, both of these are cold storage solutions for individuals. I'm a huge believer in if you've got large amounts of asset, it needs to be cold. Remember your computer is vulnerable 24 hours a day, seven days a week. If it's connected to the to the Internet, we do make mistakes as humans. We download malware. We click on phishing links. We do all kinds of things accidentally. Even the best security professionals will do those things. So having a cold is really important.

25:05Now, with Coldcard, we learned about another thing that's also really important, which is eliminating single points of failure. So sad part. I think that, you know, BitGo was founded on a sorry all the way back in 2013. I'm not trying to say I told you so, but as the asset size gets bigger, it's important to understand this. What is security? Security is eliminating any place where you might have a single point of failure that you could lose your money. That's what security is, has nothing to do with crypto.

25:34That's a principle that you would see in any financial system around the planet, putting in controls so that no single party, no single entity can take the money. Unfortunately, that single key seed phrase that you have is a single point of failure. And the reason you put it cold is because it's so critical that you don't want to have it on your computer, making it so that your computer is a single point of failure for your money. Unfortunately, even if it's cold, you still have a single point of failure on that one guy that generated that one random number, which is that one key which protects your money.

26:09And that's where Coldcard, you know, let its clients down. It's pretty tragic. Of course, a lot of people very, very terribly affected by that. What BitGo does with this multisig MPC thing is there's always three keys and you can implement those three keys using completely separate implementations. So with our Bitcoin multisig, for instance, we've got one key, which is in our HSM. That's the one that we own. You get two keys. That's how it's custody to you rather than custody to us. And then you can generate those keys however you want.

26:43We use our web-based browser. It'll generate one on the browser and generate one on a backup system. All three of those are completely independently implemented. What that means is you've got three different random number generators that are used for those keys, three different implementations on them. So we've eliminated that single point of failure on having a single implementation that is completely controlling and protecting your money. Anyway, that's what we do. And when you do the self-custody side, cold storage is a little bit harder, right?

27:15Like, how do you take those keys? How do you take them offline? Can be done. What Coldcard was doing was just giving you a single key, taking that offline. And then hopefully if that was implemented right, that'd be great. That's what Ledger does, too. And then, of course, there's tricks that, you know, bad guys use to try to find that private key. Remember, Bybit was just hacked, I think, a year and a half ago. They were using Ledgers. They were even using a multisig Ledger, but they were using blind signing and hackers were able to trick their user interface and then convince them to sign anyway.

27:53So. All right. Anyway, using eliminating single point of failure is a really important part of security.

28:01Vlad CosteaYeah, I mean, that should have been so obvious in hindsight. I think what we learned from the Coldcard issue is that these people were super paranoid. They were not connecting their device to their computer ever. They were using this what they call the air gap. But I'm not 100% convinced that it was truly air gapped. And they were using like nine volt batteries to power up their Coldcards and were like, yeah, look at me, I'm so cypherpunk, whatever. The issue was entropy generation, which is about the randomness of your private key.

28:36And then some situations that was very easy to crack if they did not have a passphrase, which is another password on top of your BIP39 seed phrase, then you were screwed if you did not have good enough entropy. Likewise, you were screwed and a lot of Bitcoin was lost. Some guys were attributing this issue to the fact that maybe that nowadays we have powerful AI and that can be another topic because it's such a hot conversation in the US right now. But I also think that this is negligence and it's just the fact that now we have the tools to look into open source software that's poorly written.

29:14And people did not care enough to look before or they were just sloppy when they audited. But that's not necessarily an AI-discovered issue or something that the AI created. It's like a human mistake that we conveniently ignored for a long time. And now it surfaced because it was easy for an AI to find it.

29:39Mike BelsheYeah, look, you got a bunch of things in there. First off, let's talk about random number generators. So I'm a computer scientist by background and by training. And, you know, prior to crypto, random number generation has been a hard thing in computer science for a very long time. For people that are not in software technology business, this sounds like a difficult concept to understand. Like, what do you mean? How hard can it be to generate a frickin random number? And it turns out it's pretty hard. By the way, I saw a really cool thing.

30:10Apparently, you can go to any AI and you say, hey, select me a random number between one and 30. And they all say 17. I tried it. You can go to Grok, you can go to ChatGPT, et cetera. They'll say 17. Go ahead and try it right now, Vlad. See if it comes back with 17. One and 30. Anyway, that's a different issue. That's not a random number generation issue. But random numbers are a little bit difficult. And so we've got secure random number generators and we've had them for a long time.

30:45The securing of your digital asset, your crypto, your private key is effectively a big random number. That's all it is. There's no magic to it. It's a big random number. So the most important thing is getting that randomness right. And yeah, even good computer scientists have failed on this over and over again. So like AI, I guess switching over to that topic, gives us the ability to audit so much more code, so much faster, build code, et cetera. It can find all these things. And sure enough, you know, it's probably part of what was used.

31:18But again, I can't see it. 17?

31:22Vlad CosteaYeah. Grok said 17. But when I tweet, I didn't tweet this. I Googled this. I guess I triggered Gemini, which is Google's AI. And this one said 20. So let me try this with other models. I have a bunch of them. I have a local Qwen. Like I want to try in the local model, see what's going to happen. I mean,

31:42Mike BelsheI think this is a really fascinating example of how language models seem to be understanding. They seem to be doing things that humans would do. They're so amazing in terms of their breadth of data that they can pull from. But when you ask it to give you a random number, it doesn't actually know what you mean. Instead, it says, oh, 17 seems to be something that humans like to see as a random number. And they're all giving you that back. So anyway, there's probably AI experts that could give you a better answer as to how that comes about.

32:21But if you think that your AI is smart, it's a great example of where your AI, at least today, is not understanding the actual question, even though it's giving you an answer that looks like an actual answer. So anyway, yes, AI, great news. It's going to find all this stuff and it's going to knock out these bugs. And on the go forward, software that's written with AI will be hardened from the start. I know to some degree, maybe what we should do is go back and rewrite everything from scratch using AI.

32:57It's going to be better coders than we are. Done right. You can have stronger test suites. You can have more robust fallbacks. And of course, you can write it without the bugs. Of course, you can't be just vibe coding blindly. I think at least today, we still need computer professionals to make sure that we kind of guide it in the right direction. But it's getting better every day. It's amazing.

33:21Vlad CosteaIt is. And it's not just the quality, it's also the speed. Because of course, you can pay a human to work for one month on this project. And maybe it will be better if he's some sort of world class expert. But this kind of stuff takes less than five minutes for Astra, I think, is amazing. I've tried it. It's just mind blowing. There's also Fable by Anthropic. And also Claude, which is good enough. And also Qwen, like it's an open source model. I can run it on my laptop and it can do some amazing things like the stuff that I use it for.

34:01I'm really impressed.

34:03Mike BelsheAll of these models are orders of magnitude better than any of the models from just three years ago. So like the power at your fingertips is incredible. The price is about to come way down on this stuff. And as it does, it's going to be even more accessible. So yeah, look, I think we're still figuring out a few things like, you know, how do we get to the looping and self improvement part? That's a little bit harder. At BitGo, we've been investing for, I don't know, two and a half, three years now, pretty heavily into this.

34:37We track exactly how the AI is working, exactly how it's being used. Every time you check in a change list, it goes through five different agents that are checking different things from security to style to database load, all kinds of stuff. And then, of course, we still require humans to review it. The new bottleneck we have is the humans are too slow on the review. The AI can put out tremendous amounts of code. But maybe one of the most interesting things for reducing bugs is the ability for the AI to look across a broad code base and make sure that it's working across all the different modules.

35:19So I think humans have had this problem for a long time. And it's why we have like, OK, we take this project, we chop it up into teams. This team owns the security component. This team owns the database storage. This guy owns the platform. And then when you want to do something that's sweeping, you need all these teams to coordinate. And hopefully you figure out each expert doing their job. AI can now just rip through that entire thing and find little details that those humans failed to communicate to each other about subtle parts of the change that impacted each of their subsystems.

35:55And that's where bugs come in. So anyway, I think that's probably the most amazing part is that they can hit across your entire code base in a way that humans can't match. And that's today.

36:08Vlad CosteaYeah, it's amazing. Also for video, like I was able to do very accurate Seinfeld representations just inserting random people into Seinfeld episodes. And this is amazing. Like three years ago, people were making fun of Will Smith trying to eat spaghetti. And that was so bad. And now it's almost like they trained their model on Seinfeld, which I guess is a copyright infringement. But it's so good.

36:36Mike BelsheThink of how many podcasts you'll be able to do. I mean, you say you're on season 17. I don't know how that's possible. Bitcoin was invented 17 years ago. I don't know. Did you have all the seasons? No, no.

36:47Vlad CosteaI started in 2019 and I had like 10 episode seasons for a while. And then at some point I got to season 15 on the same year when Bitcoin was 15. And I was like, I should just continue like this. It's more convenient. All right.

37:01Mike BelsheWell, we there's so much recorded of you. It cannot be that hard to create an AI Vlad. And then you could have like seasons 18, 19, 20 and 21 all run in parallel.

37:14Vlad CosteaI mean, it is possible to the extent that it's going to guess like who predictably joins the show every year and then assumes what their points are going to be. I guess we all have our own patterns and maybe it can do that, but it cannot... 17! Tell the future, it cannot. You know, us as humans have very weird fascinations, which sometimes are biological just because we fill in a certain way. We're going to read something and then we're going to become obsessed with that. And that's going to become our new rabbit hole.

37:48And we're going to talk about it. There is no way for the AI to anticipate that. But if we have like a predictable business or something, I guess it's not hard for someone to pretend to be Mike Belshe impersonate like your whole business pitch and say that, well, BitGo is great because we have two-of-three multisig. So that part, I guess, is predictable. But the part where you went to the White House and you actually have human stories to tell and you're doing stuff with your life.

38:20Like there is no way the AI can anticipate this.

38:24Mike BelsheI think it's going to be able to make it all up in just a couple of years. But you're right. I accused you of being the AI and you know, I can be the AI just as easily. And the scammers are going to love that, right? I get all the time, you know, some message saying, oh, I thought you told me on WhatsApp to buy something. And was it you? I was like, oh, I hope you didn't send money. That's without much AI. Happens every day.

38:55Vlad CosteaYeah, that's horrible. It's possibly the worst way to use it. I mean, the same tools, the same hammer in the hands of different people can have a different purpose. Fortunately, most people choose to build with it than harm. So correct. I don't think there's a fix for this. I don't think any amount of regulation or brakes on the AI models will fix anything.

39:19Mike BelsheLook, I mean, this is a place where, you know, back to kind of the whole politics and like, should you be required to have money transmission license or a MiCA license or payments license or whatever? Regulators are stymied by the fact that digital assets can move across borders instantly and transparently. They've never had to deal with this before. In the past, they always had the bank that they could go to and they could put rules on the bank and that's how they could control the money. And so, look, I am somewhat appreciative of this problem, right?

39:53Like, I don't necessarily agree with why they're trying to do it, but I do understand that, like, they've tried to have controls on how the money flows in and out of their citizenry. That's no longer possible, but they're struggling because they're trying to enforce the same rules. They want that control back. Unfortunately, there's not really going to be an answer. It could be, I guess maybe the one answer would be that all of the nations come together and pass laws across the entire globe. To set a standard set of rules.

40:29Fortunately, I don't think that will happen. I don't think that they will come to an agreement. But they could try to do that as a way to kind of lock these things down. But yeah, this is just a new era, right? Money can move easily now.

40:46Vlad CosteaYeah. Some people say that the internet was a mistake. I tend to disagree with that. But I do get where they're coming from. Like everything that we used to think that is orderly. And what normal, whatever that means, is being disrupted.

41:06Mike BelsheWell, I think the challenge of like, once you have the, so we have some guy who's posing to be me or you or whatever, and sell some product, and he scams people. Okay, we've always had scammers. We try to stop scammers. We should try to stop scammers. But how much do we expect that to be the burden of our governments versus that to be the burden of us as individuals? And I think we have to be really careful here. Like somebody falling for a stupid scam on a telegram or WhatsApp, I think has to be their own responsibility.

41:42And we should not look to the government to solve it. And in the past, the government has looked to solve these things by putting in money transmission rules. If you really want to solve it, the only way to do it is just continue to lock everybody down. And the problem with that is that it is going to take away self-sovereignty, it is going to tie you into a one true banking system. And actually, I think it compounds on itself. Every time you ask for a protection, well, then if you find a new bug where the scammers get forgotten, the scammers are clever.

42:14Scammers got AI too. Every time you find a new one, they're gonna have to lock that down and lock that down and lock that down. So it really does come down to kind of a freedom and civil liberties versus I want to have safety and protection.

42:29Vlad CosteaIn the background, I've asked Claude Fable 5.1 to give me a random number from one to 30. It said 17. The one true number. This is crazy. It also said that if you want an actually random one, run this script. So it gave me like a script that I can run in command line to get a random number. So at least it did something extra compared to Grok. But I paid $3 for this. I'm a fan of regulations now. I'm like, what? The hell? I'm running this.

43:02It's called NanoGPT, this thing. And I just pay for what I use. And it seems like these new frontier models are super expensive. I paid $3 for this. Crazy.

43:15Mike BelsheYeah, look, I don't know. Here we can get into the discussion about AI. So we've got pretty hefty bills at BitGo for the AI frontier models. And of course, internally, we're looking at like how they get used. And some of the things we're using AI for, yeah, you want to use the frontier models. Others, at this point, like the OpenRouter and the cheaper models work just as well. And we don't need to be spending that top dollar. And the difference is, the difference in price is astronomical, right?

43:46I mean, we're not talking about like 10% less. We're talking about 90% less, 95% less. And these AI bills have gone up so quickly as operating expenses for pretty much any software company worth their salt. Of course, you have to be looking at it. So I don't know. I think Anthropic and OpenAI haven't really quite realized this yet, but their business model is in trouble. And they're not going to sustain these prices.

44:16Vlad CosteaI agree with that. I mean, I do hear that a lot of their use is subsidized. And the fact that you get a cheap subscription, but you're actually burning a lot of tokens, which are worth a lot more money, is not a good business model. But they're just trying to maybe get customers right now and have them used to, I don't know, familiarize with the brand or addicted to using their service. But I do find that these local models, which are open source and can run on any computer that has more than 16 gigabytes of RAM, they're amazing.

44:53Mike BelsheWell, the giving away basic models for retail or individual users, heck, that's similar to what we're doing right at BitGo, where we'll take your custody, we'll take your self-custody, and that's more or less a free product. We'll make money elsewhere. But businesses, I mean, we're spending millions of dollars on AI models. So this is not small amounts. These are significant costs, right? These are like up there with how much you're spending on personnel. So if you're running a business and you're spending that kind of money, of course, you're looking for ways to optimize it.

45:31And by the way, we've met with Anthropic and OpenAI. They are, I think, so consumed by the race that they're in that they're not particularly concerned right now about, like, hey, whether you move over to OpenRouter or cheaper models. They're kind of like, yeah, you should. That's what they say. We were kind of surprised. But, OK, look, we think we just have to do that, by the way. And I think when you've had big infrastructure rollouts in the past in other industries, we've seen a simpler, similar thing.

46:04Right now, there's so much capital going into the infrastructure build out of AI and, of course, is needed. We need lots of data storage capability. We need lots of energy. All that stuff's going to go in. But it's really costly. People are so excited about it. They're spending a lot of they're investing a ton of money into it. Now, the question is, like, are you going to be able to get the return? Look, I think in the end, the value always accrues to the applications. It doesn't accrue to the protocol.

46:32It doesn't accrue to the network. Yes, there's money to be made on the infrastructure. There will be pieces there. But it's the applications that are built on top that will receive all the rewards. So there's going to be this huge capital inflow in until there's not enough capital to continue to invest. There'll be a massive collapse. There'll be a reset. Some people buy that infrastructure and capital for pennies on the dollar. And then all the token costs come down. The real place where the value is is whoever's building apps on top of this AI, whether that's in supply chain systems or whether it's in financial services systems or whatever.

47:10It's going to be applications.

47:12Vlad CosteaYeah, for the record, I've tried this with my local Qwen. And instead of actually telling me a number, it said, Am I allowed to run this Python script that says import random print random radiant one to 30? No, it's randint random integer, I guess, randint one to 30. And that gave me 15. So it's not lazy, like the frontier models. All

47:41Mike BelsheRight, you're obsessed now with the with the random number.

47:45Vlad CosteaI guess it's a recurring theme in this episode. Random numbers. But yeah, I got 15 from Qwen. And I'm actually proud of it. Like I could see the thought process. It was pretty slow in generating the script. But it did not say 17 outright. So

48:03Mike BelsheWell, if you want to bring this back to crypto, then okay, now we got to go and see like, hey, Grok, give me a random seed phrase for my Bitcoin wallet.

48:15Vlad CosteaI'm gonna do that for the sake of entertainment.

48:19Mike BelsheIt'll be like 17, 17, 17. No, I think I've tried this already. So I think they're gonna start giving you safety controls and say you shouldn't generate a seed phrase in chat.

48:33Vlad CosteaThat's a bummer. I mean, I do understand why it's doing this. But for the sake of our experiment, it's not convenient.

48:40Mike BelsheIt's an interesting question, right? Like, do you need your Grok AI to tell you not to do these things? Or should it do it and then say like, this is a bad way to do it? It's sort of like the nanny state of like, do you want the protection? Or do you want the information?

49:02Vlad CosteaSo it did give me like a wall of text on why it's a bad idea. And then at the end, it actually gave me one. If you do it again, does it give

49:10Mike BelsheYou a new one? Or does it give you the same seed phrase over and over again?

49:13Vlad CosteaIt says abandon 11 times and then about I guess the last word is for the checksum to check out. But I guess this is a value. That's 0000.

49:23Mike BelsheOkay. Yeah. Abandon is the first word in the lookup for BIP39.

49:31Vlad CosteaGive me another seed phrase. So that's not very random. With randomness. Let's see what it does. This is fun. I've never thought about doing this. I mean, it's kind of pointless because it's not serving any use other than just trying to understand how the AI works.

49:52Mike BelsheYeah. I mean, it's a little bit trivial and so silly, but it does show kind of the state of where AI is at with regard to understanding this particular component.

50:03Vlad CosteaNow it says I will not generate a seed phrase for you.

50:07Mike BelsheAnd the safety kicks in. Yeah. Right. So it's choosing safety over information. I don't know if I like that or not. I think it's okay. But but like the better answer is like I'll generate you a script for it. Right. Which can use the secure random number generators that are in the system. So those are good solutions.

50:29Vlad CosteaI just asked Qwen to see what it's going to do. This one should not refuse. I think

50:36Mike BelsheYou're running that locally on Ollama or something?

50:40Vlad CosteaYeah, it's called Hermes. But I run that on top of something that's called LM Labs... Yeah, LM Studio.

50:53Yeah, I'm a nerd like this. Like I had this task for my website, and I tried to improve the SEO. And I've asked Kimi K3 to give me a solution. And I got a solution that would take like two hours. And I was like, how do I automate this? And then I spent instead of spending two hours to fix the issue, I spent like two days trying to get a local agent to fix the issue for me.

51:17Mike BelsheSo instead of the, I guess, number between one and 100 that you asked earlier, you should do. Hey, I generated a random seed phrase with my with my AI, hacked my Bitcoin wallet.

51:31Vlad CosteaI mean, that's one way to put it. But I would prefer to send the transaction myself to someone who I mean, that's also a challenge. But I presume that people can run bots for that to just sweep the funds, but they can also run bots to put numbers in the chat. So I don't know.

51:53Mike BelsheVlad presented a Bitcoin wallet, hack it for me.

51:57Vlad CosteaThat could be a nice approach. Anyway, thanks Cake Wallet for the giveaway. I got to present two more sponsors. We're almost one hour into this interview. And then I'm gonna ask you more questions about BitGo. And what do you think about Bitcoin these days and Zcash and the forks? That's also a big one. So yeah, let me start with layer two labs because they're launching the eCash fork. You can find out more information at ecash.com. If you own any Bitcoin today, you're going to get one eCash as an airdrop on the 31st of October, which is Halloween Day.

52:31Which also happens to be the 18th birthday of the Bitcoin whitepaper. So that one is getting launched. It's basically Bitcoin with drive chains. It's the one that has all the features on the sidechain. And it's unfortunate that it's not on Bitcoin, but at least we get to see this experiment play out. Once again, ecash.com, find out more information. And also Braiins Hashpower, which is the service that you can use for mining eCash. So if you're like myself, a Europoor who's stuck in a country where electricity is stupid expensive, you can actually go to hashpower.braiins.com, with a double I.

53:10And that's where you can rent hash rate. It's actually from Braiins, which used to be called Slush Pool. That's the world's first mining pool for Bitcoin. And it's not like you're going to this shady, what was it called, Compass Mining or something, which turned out to be a scam, this cloud mining service, where they were telling you that they rent an ASIC for you and they run it, but they actually turned out to not have them. So anyway, it's not like that they're an actual legit business that's been in Bitcoin since 2010 or something.

53:46And you can basically direct the hash rate anywhere you want on hashpower.braiins.com. So you can mine Bitcoin cash if you want, you can run BSV, you can run the hash rate to mine eCash or BTC, whatever it is that you want. It's probably much cheaper than buying a home miner and also using electricity at home. And you don't have to upgrade your power grid. It's it's actually a pretty cool concept. Anyway, I gotta ask you, Mike, because there is already a Bitcoin hard fork that launched the one from Luke Dashjr.

54:24That runs BIP 110. And there's another one that's coming next month running BIP 300. What's the process for your clients to get the airdrops? Does BitGo support this?

54:36Mike BelsheLook, great question. I think we've been getting this question since 2017 or so back when we had kind of the first round of forkiness. And we put in a we've got a fork policy. I think if you search for BitGo fork policy, you'll see it. But we have a couple criteria. Like look, first off, you can't possibly support any single person's fork that might get created. There's just, you know, too much cost that goes in with it. So we look at three things. We got some technical requirements.

55:09We look at we look at market cap. Like, is it a sufficient size is worth like doing? And does it have liquidity? Because even if it has a market cap, remember, it may not have any sufficient liquidity. So the second two are more subjective. The first one on the technical side, we do care about who the technical team is and we do care. Does it have like replay protection? I don't know if your audience has heard of replay protection. But when you fork a coin, the responsible technical thing to do, unless you're deliberately trying to trick people, is to add into the protocol something that's called replay protection, which means when you're spending the coins for the forked coin, does it cause you to simultaneously spend the original coin?

55:55So if you had Bitcoin and now you create a fork and you spend on the fork, does it cause you to accidentally spend on the original? If it does, it's a problem because now every user that's trying to use the new thing to collect that airdrop, as you say, can accidentally lose their real money. So anyway, these things go into it. Of course, most forks don't support replay protection, which means that, OK, either A, you don't support it or B, the wallet provider, us, has to go figure out how to basically manually create it.

56:33And there's ways we can do it. Right. So if you had a really popular fork of Bitcoin, what we would do probably is we would suspend movements on both chains for some period of time. We would go get some, I guess not Satoshi's, but Satoshi primes, whatever it is from the forked chain. And then we can insert those into transactions to help make it so the transactions don't replay across both chains. Super expensive. A lot of development work that has to go into this. So we very much do not want to have to do that.

57:08It's going to have to take a very high bar before we would support a chain where we have to do that type of activity. It's also risky, right? Like, you know, you screw anything up and then people lose their money. Terrible. So look, on the good news side, I don't actually see any material forks on the near-term horizon that could change. Of course, you never know.

57:30Vlad CosteaSo is there a point where you decide that if it has replay protection, you're going to support it? Like, what does it take? Demand from large exchanges or your customers? Or is it a certain price where you say, well, this is worth considering?

57:45Mike BelsheWell, we reserve the right to decide. But there's three things that we really think if you're going to do a fork, we think there's three things that you should accomplish. Number one, you should hit some level of technical proficiency with your new fork. And if you do those things, it makes it a lot more likely that it could be supported. Number two, you have to have a new fork that's got a market cap with some value that matters. And then number three, there has to be liquidity on it.

58:12Number three is usually the easiest. I mean, exchanges like to trade anything. So you create a fork. For them, it's cheap and easy. They don't have to do all the work we have to do. So they'll often support a fork before we would. But again, it's hard to get all of these things going. It doesn't really happen with credibility very often.

58:33Vlad CosteaYeah. I mean, I do understand that and I respect the conservative approach. And I like the honesty, like some exchanges pretend that they have a very, how should I put it, sophisticated listing process, which they claim to be the norm or whatever. But when the Trump coin launched, I mean, all of them listed it the next day. Well, that wasn't a fork. Yeah, it was not a fork. But still, it was not something that anyone paid for listing or whatever, or something that required some, I don't know, development team that's serious or whatever.

59:12They just went for the liquidity. If it had a lot of trading activity, they listed it.

59:18Mike BelsheLook, exchanges like to trade. Right. So how do they how do they grow? They have to grow users or they have to grow tradable assets. I mean, I think it's why right now we see so much hotness with these offshore SPV tokenized stocks. Right. It's new audience for the U.S. exchanges that trade stocks. They used to only be able to trade with U.S. persons. Now they can trade this tokenized stock, give people price exposure outside the U.S. So it increases their audience. And then, of course, it trades 24/7.

59:51So it allows them trade more, and that's how they make money. Again, this kind of comes back to like, you know, should we have a you should a listing of like, oh, you should trade this, but not that. Anytime you get into a subjective thing there, like it's like, OK, well, who gets to decide? I don't actually have a problem with all these things being tradable. Anyone should be able to decide that. And then you or me, we should be very discerning about which assets we decide to buy or sell.

1:00:21But as for the exchanges, like, yeah, it's kind of their business.

1:00:27Vlad CosteaWhy do you think that we have two more Bitcoin forks this year? Do you think it concerns the politics of Bitcoin or the lack of governance? Or is it just like some idea that comes back every decade with people trying to improve it in certain ways?

1:00:44Mike BelsheYeah, I think there's some pent up frustration about, you know, should the protocol completely ossify or not? And so, yeah, it does seem to come up. I don't know. I think in 2017, when we had the what's now called the block size wars. Remember, the fees on Bitcoin really, really spiked. And there were pretty much everybody was universally like, hey, what should we be doing to address this? And there were proposals. From pretty much every party. And then eventually it turned into this thing of like, hey, it's an us versus them.

1:01:20But really, it was a technical debate about, you know, what should the what should the block size be? And that led to led to some things eventually. OK, have a fork. People kind of pick which fork they want to be on. I think that I think that actually works out pretty well.

1:01:39BIP 101 or whatever that was. I mean, I thought that was interesting. I don't know. Some people think that like there should be a control over what goes into, I guess, a transaction. I think that's kind of a futile thing to do. It doesn't seem like there was much support for it. So I think we got a very market based answer to whether people think that that's a good thing for the Bitcoin protocol. And then there's more coming. Right. So Paul Sztorc got his thing going.

1:02:12And we'll see what happens there. I think he's at least. You know, he's been around for a long time. I think he's got some credible reasons for like why he wants to make the changes he wants to make, whether the market's going to value that or not. I mean, we'll we'll see. Right. So Bitcoin Cash back in 2017, 2018, it got a value. Right. And some people thought it was going to go a different direction. But now if you look at the price of Bitcoin compared to Bitcoin Cash, the market has solidly decided where it thinks the value is.

1:02:42So other than the fact that these forks are quite annoying from the perspective of wallets and security providers, because we have to do all the work. We do all the work. The market sorting it out, I think, is great. I don't have a problem with that at all.

1:03:00Vlad CosteaYeah, it's also interesting to me that on one hand, you have a fork which is reductionist and is trying to say you should not do this on Bitcoin. And they also claim to be the true Bitcoin. And the other one is trying to maximize what the Bitcoin design does. And it adds side chains and it imposes no limitations on what developers can do on it. And it also says that it's not Bitcoin. It actually has a different name, a different branding, different GitHub. It does not try to associate itself with Bitcoin.

1:03:33Mike BelsheYeah, like, there's some amount of like almost a fair launch. If you're launching any coin. Right. How do you get the money distributed? So one thing you can do is you can fork a chain. If you're not trying to claim the naming or the branding of the original, like, really, what are you doing? It's like a it's just like, hey, I liked this technology. I like this distribution of users. I can give them an airdrop of my new coins. And then, like, we can see what happens.

1:03:54You can love it or hate it, but you're not affected. If you were holding the original coin and now this new thing emerges, the debate over naming, I think at this point is quite silly. I think it's very clear there is actually a thing that's called Bitcoin. I think people invested into it with a set of expectations and anyone trying to say, hey, there's a new set of expectations and they know the one true set of new expectations and they should control that.

1:04:21That that seems just wrong to me. If they want to say they want to create a new thing. Okay. You know, they can do that. Anyone can do that. The market is open. But yeah, I don't know. These things have two halves. One is the technical half. Like, what are you trying to accomplish? And the second one is like the branding and political half. And maybe the third one is like the masses of people that, you know, maybe just join on because they have a popularity vote of like who runs, you know, the original and who runs the fork.

1:04:51But I tend to concern myself more with the first thing. Like, what's the technical thing that you're trying to do? Then let the market sort out the rest and all the politics is not that interesting.

1:05:05Vlad CosteaIt's also the question of did Satoshi give us like the final solution to having Internet money or did he give us the blueprint for creating unstoppable Internet cash? And of course, the people forking are on the second camp saying, well, we can improve on this design and we can make something different, which maybe works better for what we want. And there's also people who are really religiously on the first camp and they say, well, Satoshi gave us this. It's like this pure creation. It has the immaculate conception.

1:05:40When it was being mined, it was worthless. There was no industry. There are no VCs. There are no actually there are VCs, but that's the part that they like to ignore. There are no expectations, at least in the beginning. And that's why it's like the one true form of money that everyone should adopt.

1:05:59Mike BelsheWell, I don't really like any of that religious or cultish type behavior of Satoshi as some sort of a Jesus creator. I think that that is just not very interesting. What did he do? He invented a set of technologies and assembled a set of pre-invented technologies in such a way that it gave us decentralization like we never had before. That's what he did. And it's a protocol and it's got scarcity and it's got decentralization built into it. It uses proof of work consensus. And that's what it is.

1:06:38It's inspired tremendous amounts of fantastic work across all of these blockchains that have been invented since. Each of which has different degrees of decentralization. They've added new features. They've added smart contracts. They've added all kinds of things on top. I think all of that's fantastic. And Satoshi deserves some sort of credit for having inspired it. In terms of these forks, yes, you can always use a fork as a way to jumpstart a new thing, whether it's you want sidechains, drive chains, whatever. But it doesn't take away from like, look, there's a thing called Bitcoin, which operates the way it operates today.

1:07:16And it does get some out of modest update. There's some debate about, you know, should there be one true implementation of that in the Bitcoin core implementation? There's been Bitcoin demon for a long time. I see there's a new one called rBitcoin. I'm not sure if you've seen that as a new implementation of Bitcoin. It's just trying to implement the protocol. I'm not trying to change the protocol. So, look, I don't know exactly how to answer the question. I think he invented a technology and that technology has been inspiring to, I think, millions of developers at this point.

1:07:52Vlad CosteaI do tend to agree with that and also think lately that these forks and these so they could be code based forks, which means that you take the technology and you build something from scratch like Zcash or Litecoin or whatever. And Ethereum, or you can have like UTXO forks, which take the ownership of the coins and they transfer them to a new network where users can do something else. Regardless of how you approach this, you're scaling Bitcoin. And if not Bitcoin BTC directly, because you're creating a new asset, you are scaling the idea of Bitcoin and the concept of having Internet money that cannot be stopped.

1:08:31Of course, it's going to have a different valuation and different security model, maybe a different set of tradeoffs, but it's still a way to get this kind of money into the hands of people. And if you look at it this way, it's like a net positive for adoption. It's a net positive for education, for teaching people how to use non-custodially this type of technology. Maybe that BTC is too expensive to experiment with and it's much cheaper to just run a mobile wallet and get $10 of the eCash fork or something on your phone.

1:09:11So I do think that mainstream adoption is actually reached with these forks and some people hate them just because they feel like they dilute the branding of Bitcoin or they don't allow the price to go up as much as it possibly would.

1:09:28Mike BelsheThat's a very abstract point of view, kind of that everything is some flavor of Bitcoin. Look, I think Bitcoin has inspired a lot. I mean, people look at Bitcoin, they're like, wow, it's an amazing store of value. I would like to have a store of value of my own. And so I'm going to create my coin or whatever, and then I'm going to go try to make that work. Look, I think the best part is that it's taught people to start thinking about money. What is money? How does it work?

1:09:56You've been spending your whole life pursuing it. Your parents pursued it. Like, what is it really? And so I think that educational component is really, really valuable. It allows people to start thinking about what are the systems behind what we put our faith into? What risks do we have for ourselves, for our children, for families? Now, then separately, when you start using it in different contexts, it ends up being very different very quickly. So Bitcoin really is. The only coin so far that has achieved a decent level of decentralization, and I say decent because it still needs to be better, I personally don't think Satoshi fully anticipated like mining pools, which creates a bit of a centralization point.

1:10:43I think he probably did think about a single code base, which creates a centralization point. But, you know, there's a number of places where we can do better to decentralize it. I don't think he anticipated digital asset treasury. And, you know, what happens when one digital asset treasury accumulates a tremendous amount of Bitcoin? He did anticipate ETFs coming along, and those also create centralization points. All right. Bitcoin has done that. Anyone that takes that technology, uses it for a different purpose, first has a decision around how decentralized do they want to be.

1:11:20It's hard to be decentralized, super hard to be decentralized. It does have an ossification property to it in that it's much harder to change the protocol once you decentralized it. So a lot of these guys that are trying to build securities tokens or when they're trying to build commodities trading or all of these other things, they're blockchain inspired, but they actually don't want to necessarily move to a fully decentralized thing. That's not the biggest value to them. The biggest value to them is to have kind of the open API that everybody can participate in.

1:11:54It's a public ledger. Everyone can see it's got a bunch of properties that way, but they want to be able to continue to change it as they build their software. Remember, prior to Bitcoin, like you think about software, you launch version one. There's always a version 1.01. There's always a version 1.02, then a version 1.1, a version 2.0. This is the way software works. There's always something that you could have done better. Even with Bitcoin, that's true. But with Bitcoin, the protocol is locked in a way that's very, very, very hard to change.

1:12:22So when you're building other applications that do have regulators looking at them, you actually need to be able to change them. So the good news is he created a mechanism with blockchain that's got transparency to it. Can't be cheated as much because it's got cryptography built into it. It's got a consensus layer. Depending how you do that, it can be more or less decentralized. This allows us to build all kinds of new financial products, helps increase that education cycle for investors and people that are trying to just keep money, keep their, you know, grow their wealth for themselves.

1:12:55But those are going to have very different properties than what was originally invented or attempted to be accomplished by Satoshi.

1:13:08Vlad CosteaI know that you have a history with trying to improve Bitcoin and you're part of the New York agreement and that feels like a lifetime ago.

1:13:17Mike BelsheI was not part of the New York agreement, by the way.

1:13:19Vlad CosteaOkay. I'm not going to...

1:13:20Mike BelsheBut I will talk about where I am, where I was, if you want to ask that.

1:13:24Vlad CosteaDo you believe that if Bitcoiners sat together and just decided to have a conservative block size increase and there was never a Bitcoin Cash or a Bitcoin SV or whatever, would we be better off today? Or do you think that the split was inevitable?

1:13:42Mike BelsheWell, just to be clear on kind of what happened, because I think it gets portrayed almost from a religious perspective and it doesn't necessarily need to. You know, like I said, I'm a technologist. I worked on HTTP version 2.0. I'm the lead author of that product. In the case of internet protocol standards, they go through this body called the IETF, the Internet Engineering Task Force. It's been around since the early 70s. When you go into the IETF and you talk about how protocols can be improved, you will find fervent, heated debates between people that are very, very smart, diametrically opposed, staunchly convicted that their method is the right one.

1:14:22And you come into these meetings and these debates and you solve these issues. And that debate is necessary and healthy in order for the protocols to evolve. Do you get the perfect answers? Nope. Like with any human system, you don't always get perfect answers, but you do get the debate out. You do get the ideas out and you resolve things. At the time that we got to what's now been dubbed as like the New York Agreement. And by the way, BitGo never signed it. We intentionally didn't sign it because I did not believe that like you should sign something that says, I think the protocol should be this way.

1:14:58When you go into the IETF, at the time I was working at Google, you shed your company affiliation and everybody is an equal voice as an individual. So you're there to debate protocol. You're not there to debate on behalf of your company. So I come from that ethos in terms of like how you should build protocols. And of course, Bitcoin's got its own BIPs. In the early days, people come wondering, you know, how is this going to evolve? Will there someday be... And you can still ask, like, should there be an IETF type of body that has a little bit more mature process around how do you evolve protocols?

1:15:34It's not a terrible idea to consider. But everybody had an approach about how to increase the block size. Gavin Andresen had it. Adam Back had it. There was, you know, we weren't even related to it. Like Blockstream took a bunch of people, went out to talk to all the miners in China, had like the Hong Kong agreement. These agreements were silly in my view. So Barry Silbert, I guess, is probably one of the organizers of the now dubbed New York Agreement. Basically, after having not had a lot of consensus, people said, OK, well, we haven't had consensus.

1:16:11How do we build it? Let's go get some of the players in the space and let's see what they think and let's see if we have a common viewpoint. And can we use that to get a resolution to increase the block size? So really, it's just another consensus building mechanism. The Bitcoin protocol can change and it does from time to time, but it requires everybody to come to agreement. So how do you get a consensus or an agreement? Well, you start with an idea. That's one person.

1:16:37And then you got to bring other people in and see if they agree. And if they don't agree in the end, you're not going to do it. So anyway, with the block size wars, there are a whole bunch of people that thought, you know what, the best way to deal with this is just increase the block size. And it's a little unclear as to what the original intent of the protocol was. I mean, really, the one megabyte cap from the early days was clearly as much of a spam prevention as anything else.

1:17:03All right, so SegWit got created. That's good, that's helpful, kind of separates some of the data that's in the blockchain in a pretty constructive way, kind of regardless. But yeah, the plan was like lightning is going to come and it's going to take all this stuff and like it's all going to work. Well, here we are eight years later and that hasn't really materialized. It also took way too long to get these implementations. Even the SegWit side took a long time. Look, I do think that the community could have decided to just increase the block size and it would have been pretty trivial and easy.

1:17:39But the community decided not to do that. And so that is what it is. I don't have a problem with that in any way. Would it have grown faster? It might have. Look, I do think that without having a path to low fees, a path to be able to support large numbers of transactions, Bitcoin missed out on what eventually went to become stablecoins on different chains. And I think at this point it's done. So, you know, whether Bitcoin should be a payment system, a peer to peer electronic payment system, like what Satoshi wrote in the white paper, whether it should be a store value and people talk about what it should be.

1:18:22But in terms of what it is that store value at this point with 10 minute blocks and relatively limited block space, the community decided that that was the path. Now, just to play the other side of the argument, you know, we talked about decentralization a lot. You know, there's a lot of fear that if you did increase the block size, that it would negatively impact decentralization. I think those have mostly been debunked at this point. I mean, the size of the blocks are so small that you can double, quadruple them.

1:18:58And it's pretty clear it wouldn't have a lot of impact. But I do value decentralization tremendously. So doing anything that would potentially destabilize or reduce that decentralization, not good. So look, the community decided to take a very conservative approach towards decentralization and forgo becoming payments. And that choice has been made for Bitcoin. Done.

1:19:25Vlad CosteaWell, just to cherry pick one of your arguments, I don't think Bitcoin missed out on stablecoins because USDT was actually launched on the Omni layer, which is something that's on top of Bitcoin. It used to be called Mastercoin. That was back in 2015. The problem with that is that I think in 2017 and 2018, that was taking like 30% of the block space and some people were angry because the fees were high because of USDT trading. So the guys from...

1:19:55Mike BelsheWait a minute. Let's go back. Okay. So, yeah, if you want to talk about Mastercoin or Omni, those are protocols that are unbeknownst to the consensus layer of Bitcoin. And they say that they run on top of Bitcoin, but they're not validated by Bitcoin. So what does it mean? What it means is there's a software layer not related to Bitcoin. Bitcoin knows nothing about it. The miners know nothing about it. The miners are not validating the transactions. But that software layer is dropping some extra stuff into Bitcoin transactions. And then when it sees those transactions on chain, it's interpreting it in some way.

1:20:29So the way that it is tied to Bitcoin is that as long as everybody at that upper layer, which is outside of Bitcoin, agrees, then, okay, I guess it's fine. But what happens when all of a sudden one guy says he sent a million bucks? The other guy is like, no, that's not a million bucks. That's only two bucks. Who's right? The Bitcoin miners don't solve that issue. There is no consensus network at all for Omni or Mastercoin. So you're correct. Tether was originally launched on Omni, which is on top of Bitcoin.

1:21:00And then there were some people that interpreted that just because you could move the Tether on Omni by a Bitcoin transaction, which had some comment at the end of it. The only reason it worked is because everybody agreed on the one true implementation of Omni. And sooner or later had that continued. And of course, it didn't go very far. And there have been a long history, even before Omni and Mastercoin, of other types of protocols that would claim to have the security of Bitcoin by sitting at the application layer above it.

1:21:31Look, if the consensus layer, if the miners do not know about your transactions, it's just simple fact like you are not protected by Bitcoin. It's as simple as that. And anyway, so you're correct. Tether was launched on that. It did get a lot of usage on that. People did trust it anyway, even though that's not in my view, protected by Bitcoin at all. And when I say stablecoins won, look, I think stablecoins are actually kind of a better payment mechanism because they're dollar backed, euro backed, etc.

1:22:06You get paid in dollars. You get paid in euros. You pay your taxes in those coins. People still to this day don't really like spending Bitcoin. That's in part because it's volatile. That's in part because it's going up in value. A bunch of different reasons. So, I mean, in terms of payments, I think the ability for Bitcoin to win on the payment layer, two things had to happen. A, you have to have high volume of transactions at low fees. B, you have to have a willingness for people to spend them.

1:22:34Neither one of those things happened. But even if the willingness had been there, without having, you know, large numbers of transactions, the Bitcoin protocol decided to forego being the layer at which they would transact.

1:22:48Vlad CosteaYeah. And I respect it. Like your argument was technical. And it's true that Bitcoin miners are not aware of what they're actually mining in these up returns and multisigs that embed arbitrary data. So they don't understand what's in there. And it's up to some sort of other protocol to interpret what that means. But my comment was more political because in Bitcoin, we had everything. Everything in Ethereum, almost. I'm not sure if we had cool stuff like smart contracts, Uniswap and whatever. Like we had pretty much everything from the beginning, like an early form of it on Bitcoin.

1:23:26It's just that politically, some people were always opposed to having those for technical reasons or for just the fact that they hated seeing the blockchain being bloated with stuff that they did not like. And I'm sort of happy that there is like a fork of Bitcoin right now that handles this. And they can just have purity tests.

1:23:48Mike BelsheWell, Liquid has now kind of answered the question for us. So, you know, going all the way back to...

1:23:53Vlad CosteaYou mean the Federation?

1:23:54Mike BelsheYeah. Yeah. Back 2014, 2015, there was this debate of like, well, how do you change Bitcoin? So the problem is like, OK, as Bitcoin grows, has more and more money in it. Like the last thing you want to do is have that money be put at risk because you're trying something new that may not work. So this concept of sidechains came up. And of course, Blockstream.

1:24:18Came out and pioneered a lot of that work, did a bunch of things and created Liquid. And the idea was you could have a sidechain where you can kind of peg some of the value to and from Bitcoin into the sidechain. And then the people that participate in the sidechain take the risk. But the Bitcoiners are not at risk. And this played out just, what, two weeks ago. Right. So turns out they shipped a bug on the Liquid network, which allowed someone to drain through the federation off of the Liquid network.

1:24:54Had that bug existed on the Bitcoin side, that would have been devastating to all Bitcoin holders. Right. So by using a sidechain, it pushed the risk off of the main chain and allowed those people to participate on what they were doing on Liquid without exposing all Bitcoiners. So I think that part of Bitcoin has got it right. Like, do you want to be adding smart contract layers on top of Bitcoin when it wasn't designed for that? And in the early days, even before 2014, some of the things that were happening was like the script op codes were getting pruned down.

1:25:30Why? Because the developers recognized that, like, wait a minute, even though we don't know exactly what the security issue would be with these various op codes on Bitcoin, we know that these are powerful enough that it gives a large surface area for attackers to use to attack Bitcoin. So they pulled them out conservatively. Anyway, I think Bitcoin made the right choice here. And if you even think about Bitcoin to this day, like right now, we have this problem where fiat money is going down and down in value.

1:26:05That doesn't seem to be fixable in the US and Europe anywhere. The intent that I think Satoshi very much put into the genesis block, you know, talking about the banks and talking about the 2008 crisis. Was about how do we have a store value? How do we have some sort of monetary system which is not fiat based? And he created it. Today, if you had a billion dollars and you're like, you know what, this dollar thing isn't good. I got to get into something else. Maybe you're going to go to gold.

1:26:34Maybe you're going to go to Bitcoin. If Bitcoin has got liquid sidechain types of experiments going on with it, that's a billion dollars. You're gonna be like, wait a minute, can I put this into that? I don't know. Is it safe? I'm not sure. So I think it's really important that Bitcoin did it. It did what it did. Lock down the various changes, make it very, very difficult to change so that you can continue to invest it without knowing that tomorrow some developers are going to come up with some crazy idea that puts your billion dollars at risk.

1:27:02Vlad CosteaThere's also the issue that Satoshi's biggest bet, as you can find in the white paper and the incentive subchapter, is that there's going to be enough transaction fees to make up for the subsidy that gets paid to the miners. And so far, I don't think we have enough. We have enough evidence to believe that that's actually going to play out in the way that Satoshi wanted. There was some hope when the ordinal inscriptions were hot, and it was maybe two or three years ago when a few times we had block subsidies that much succeeded, that we had transaction fees that exceeded the subsidy.

1:27:41But I think every time we fork, every time we launch a federation that competes with the base layer for transaction fees, every time we get a new altcoin, that takes a long time. A lot of users like Ethereum and Solana or even Zcash, these users are not paying Bitcoin transactions. And of course, nobody wants to pay $10 or more for a monetary transaction. That's almost as much as Western Union charges at some point. But what I'm trying to say is that Bitcoin needs to be secure. And I think a point where it becomes less secure is where the hash rate drops 50% down from where we are.

1:28:20Which means that there must be ASIC miners that can 51% attack the network. So that's a sign that the network has become insecure. And long term, we need to figure out use cases for Bitcoin, BTC specifically, to remain secure. Because if we don't get enough transaction fees, and right now, you can actually get into the next block. I want to fact check this. I don't want to speak out of memories or something. But I think transaction fees are very low right now. And it's not a good sign.

1:28:55It's not a sign that Bitcoin is healthy or is popular. We have the institutional asset that's popular. So you can get into the next block with less than one Satoshi per V-byte right now as we're broadcasting this interview.

1:29:09Mike BelsheThese are great questions. I think sometimes people forget that Bitcoin still has a little bit of experimentation or is still a bit of an experiment itself. And some of these things have been theoretical. And as we move in time, we find out exactly what happens. So we don't know what's going to happen in terms of fees as things progress. I'm not as worried about it. I mean, I think the problem is people can speculate as to how bad they think it's going to be. And they can say, like, it's not healthy, like you just said.

1:29:50I'm not sure that we really know yet. I think it's perfectly fine for now. But yeah, you're right. Like, if things go the wrong way, then maybe it's not always going to be perfect. But I think humans have a way of solving these types of issues. And I don't think this is the biggest long-term threat to Bitcoin. I think there's probably other issues that we should probably worry about more. But yeah, these are things that should be watched, and those might need a little tweaking for Bitcoin, of how do you deal with that. I don't know what the right answer is. I think Bitcoin scarcity is incredibly important. I don't think you can change Bitcoin scarcity. If you do that, I think you're changing the fundamental promise of the protocol. So anyway, I wouldn't rule out... I wouldn't jump to conclusions that this is a problem yet. It's not a problem today. I think it's years away from being a real problem. Now, Bitcoin mining, I think is, we're learning a little bit right now as Bitcoin miners shut down and shift over to AI. So that could create a little bit of a temporary set of conditions that will be interesting to watch. Personally, I don't think it's going to be a big deal either. As you know, the difficulty is dynamic and it will reset every 2000 blocks.

1:31:06So the protocol has kind of got some adaptability already built into it, which I think is going to ride us through, even if we were to have a substantial drawdown of mining capacity. You might have slower blocks for a little while if it were really extreme. But I don't think that would be a killer for the protocol either.

1:31:27Vlad CosteaYeah, there are some guys who are proposing that we can have demurrage and we have some armchair economists who are trying to tell the difference between fixed supply and scarcity. Because, for example, gold is scarce, but it's not fixed in supply. You can always find more. And that inflates with like a predictable rate, according to the mining. But you never know how much there is.

1:31:53Mike BelsheI was not accurate with that wording. I mean fixed supply. I think the fixed supply actually is really, really important to Bitcoin. With gold, you're right. The inflation rate is somewhat predictable based on historical values, but it goes up and down as the value of gold goes up. So as the value of gold goes up, it becomes economic to go mine it from farther and farther away places. And as you know, there's folks trying to mine gold off of asteroids. And there you could actually see a 5% disruption to the gold supply if somebody successfully did that.

1:32:30Maybe even larger. I don't know. So that's kind of a different thing. Obviously, that's not what we're talking about with Bitcoin. Some people are talking about like, hey, you're going to need a forever subsidy to miners. This argument has happened for like forever. Of course, if you're a miner, you're like, yeah, give me some more. But I think it violates the fundamental premise of Bitcoin in a way that would be pretty devastating to the reputation of the brand. I think it would slow things down. So I think that should only be done at a very, very, very last resort.

1:33:06I don't even think we're close to having that discussion yet.

1:33:11Vlad CosteaYeah, it's actually cool that you're defending the 21 million limit because I see a lot of people being convinced lately that maybe it was not such a good design or such a good idea. And maybe some tail emission, like 1% per year or whatever is better because it steadily pays the miners and gives them like a predictable business model. But I do like the chaos that we are sort of exposed to when miners need to figure out on the go, like what's going to be their strategy, how much money they're going to be making.

1:33:45Maybe have some reserves just in case the price is going to tank. It's quite a dynamic business. Like there is nothing unlike the proof of stake world where you have the stakers who basically have very low costs for staying online and earning rewards. In proof of work, you actually need to adapt. And it's a very wild place where even the biggest players of 10 years ago are no longer on the market today. And it just keeps changing and evolving. And there used to be a time when most of the mining was in China.

1:34:16Now most of the mining is in the US. It's really amazing how this is evolving.

1:34:24Mike BelsheFor sure. I guess my viewpoint is somewhat principled. I think the promises that Bitcoin had is what led to people investing in Bitcoin and what gives it its value today. And so when you change those promises, you're kind of rug pulling. And even if you're only rug pulling a tiny percent, or it's for the good of the network, you are changing the rules under which people invested. And so people need to rethink, you know, how do they want to invest their money? So as soon as you go there, I think actually that opens up the door to like, okay, maybe I should just be thinking about something that's not Bitcoin anymore.

1:35:12And it may be very similar to Bitcoin, may have a lot of similar properties, may try to get to be decentralized as well. Maybe it'll be a Bitcoin with inflation and maybe people like that better. That might be a little better product. I don't know. Let the market decide. If the market thinks it's a better product, they'll shift to it. If they don't, they won't. But to change Bitcoin is to rug pull it at this point. I think you can't do that. I don't even think like, so Paul Sztorc wants to change some things about Bitcoin.

1:35:40He's going to have his own fork. I don't think he's got inflation in his either. Do you know?

1:35:48Vlad CosteaNo, he has the same cap. His assumption is that, so in his design, which is BIP 301, there's merged mining, which is something that Satoshi actually created for Namecoin early on. And he believes that if this is going to take off and it's going to have a lot of users, the transaction fees from the side chains are going to be paying for the miners to keep going. So there's never any need for inflation to occur. Yeah,

1:36:18Mike BelsheI think it's a reasonable theory.

1:36:23Vlad CosteaI mean, there are only a few ways with which you can generate more transactions and more fees for miners. You can either increase the block size to let more people transact or you can have inflation or you can have side chains that actually pay the miners of the main chain. So,

1:36:42Mike BelsheWell, there is another thing that can end up happening. I'm not sure I like it, but like, let's say the fees went to zero now. So what would miners do? I mean, maybe they live off the subsidy alone. Obviously, the subsidy continues to decrease every few years. What they could start doing is start charging SaaS subscriptions to folks like BitGo. And they're like, oh, you want to have your transactions included? You need to pay a fee. Otherwise, we're just going to have it be really slow. So you could end up with a market for getting into the miners.

1:37:17This wasn't supposed to be possible, but because of the pools, it kind of is possible. So I mentioned earlier briefly, I don't think Satoshi really anticipated that we would have like this concentration of what transactions go into a block by way of a pool. But, you know, here we are. And anyway, I think that could happen. I'm not sure that I like it, but I think that it's another vehicle for miners to make money in the event that there's no transaction fees.

1:37:50Vlad CosteaAnd I think we've seen that with Marathon Mining, and they are one of the companies that have a service that's called Slipstream, where you can pay without actually showing up in the mempool to queue your transaction there. You can pay them outside of the Bitcoin network to include something that can be anything, as long as it goes through their legal department or something. They will just do it. They'll take your money.

1:38:19Mike BelsheI mean, look, I think the market should decide if those services are good or bad. If we get to a point where that ends up being like the only way we can actually pay for the mining, then that's the part that I'm more concerned about. But again, I don't think this is anything that's imminent.

1:38:40Vlad CosteaA conversation that we had on the last episode was about fungibility in Bitcoin and whether or not we're going to have some meaningful privacy that does not allow the miners to censor or discriminate transactions. Because the system is not really censorship resistant. It's just maybe socially censorship resistant right now, because you can choose to mine that if you're like a neutral miner. But unless we can no longer distinguish, we're not going to have this ideal of a system that's truly neutral and has everyone's transactions embedded. And I remember at the time you said that probably it's going to be zero knowledge proofs like Zcash, but it's going to take like two decades.

1:39:23Do you still believe this?

1:39:28Mike BelsheI guess I kind of think that Bitcoin is not going to do this at this point. I don't know. I suppose it's possible. Maybe I'm not cryptographically smart enough to answer that question for Bitcoin. I do think other chains are going to solve the privacy issues, and it may just be that they accrue the value of, I'm sorry, they accrue the traffic of the folks that need those features. I think Bitcoin is pretty hard to change at this point. I think the more imminent concern about fungibility for Bitcoin is probably regulators. You know, we have this miraculous thing, which is the blockchain, and you can see so deep in it.

1:40:11There's some companies that made massive businesses around. You know, claiming that they can see when bad money is moving from here to there. And to some degree, they can. And they can trace it and they figure out where it went, etc. And regulators love this. This is like the best thing ever. They never had this with the traditional finance system. And now they have it with crypto. And then they may want to, A, require it so that you can't have privacy. And B, they may make it so that like, if there's any transaction in your history that they don't like, for whatever reason, that suddenly all of these financial institutions can't touch it.

1:40:52So I think that's the biggest threat to fungibility. Obviously, you know, zero knowledge proofs could be a solution to it. But you'd have to have two things. A, you'd have to have the protocol changed to support it. And then B, you'd have to have all the wallets use that mode. So for instance, Zcash, I mean, we mentioned a couple of times here, Zcash has a shielded and an unshielded mode.

1:41:15And it's not very few people. Actually, I guess I don't know. Maybe you know. What is the percentage that's in shielded mode on Zcash today?

1:41:26Vlad CosteaI think like half. I can look it up. There is a website that's called ZecHub, which has all this data. It's also broken down by the shielded pool themselves. And I'm going to tell you the exact number shortly. All right. Even

1:41:45Mike BelsheIf we did deploy it into Bitcoin anytime soon, you'd have to have all the wallets upgrade, all the exchanges upgrade, and then have people really start using it. I think it would be a very, very small corner of Bitcoin for a very long time.

1:42:02Vlad CosteaSo right now, as we speak, the amount of shielded Zcash is about 4.5 million. Let's round it up to 4.5. It's not quite 4.5. So that's out of maybe what's the circulating Zec, 16 million. So there's 10 million Zec that's not shielded or more, like 12. And then there's the 4.5 that's shielded. Interesting. But it used to be much lower. Like if you look back in 2017, it was like 55,000 Zec that was shielded. Last year, it was like one, no, two and a half million.

1:42:48Now it's 4.5. So it keeps going up, people. I think because they have much simpler ways to shield in their wallets, it became more scalable and now people are doing it.

1:43:00Mike BelsheWell, it doesn't scale. Do you know how it works? It's terrible. So we implement shielded mode Zcash at BitGo. And basically every transaction that comes in off blockchain that's in shielded mode, you're like, okay, is this my transaction? So if you have like a mobile wallet, it's just your wallet. Every transaction that comes in, you're checking, is this mine? Is this mine? Is this mine? Is this mine? And it's just for one wallet. And BitGo, we operate wallets for millions and millions of people. And so what does that mean?

1:43:29Every transaction that comes in, we have to scan every single one. It's a linear scan. And there's no shortcut for this. So it's incredibly expensive. There's some interesting denial of service attacks that you have to work. Denial of service attacks you have to worry about when you're trying to do a scalable wallet that supports this mode. I'm not complaining, by the way. It is what it is. But you mentioned that maybe there's more shielded mode adoption because of scalability. I don't think so. I think what happened is the Zashi wallet, I think they renamed it Zodl, I think is one of the more easy to use ones.

1:44:04I think they've had a couple of other wallets also move in the shielded mode. So I think it takes a lot more to move into shielded mode. And that's why. It's also difficult, by the way, to send from, I shouldn't say difficult. There's a lot of Zcash wallets that support the unshielded mode, but can't send to a shielded wallet because they just didn't implement it. It's unfortunate.

1:44:30Vlad CosteaThey have like a unified address format and some wallets do not work with that.

1:44:36Mike BelsheCorrect. And I think it might be most. I haven't done a survey of like all the Zcash wallets or anything, but it's actually kind of hard to move around to the shielded mode. But all right. The other thing I think that's happening is I think as fiat currency continues to devalue, there's an increased paranoia that like maybe we need to do something here. So, you know, your viewers will probably appreciate this. Of course, 6102 has a special meaning to you, right?

1:45:10Vlad CosteaAnd not personally to me because I'm Romanian, but I know we had much worse with the communists, so.

1:45:19Mike BelsheOK, there's probably a much worse 6102 that I should know about, but I don't. But here in the U.S., where I think we do have the best property rights of any country, there's still not perfect ownership of your property as your property. And, you know, in the early 1940s, Franklin Delano Roosevelt, a heralded president of the United States, issued Executive Order 6102, which confiscated gold to private citizens and obviously is not confiscated anymore. But could there be a confiscation of assets? What seems to be happening politically in the U.S.

1:45:59Is that there is a never ending appetite for more tax. And it's just like no matter how much you give these guys, they just want more and they'll say it's unfair or whatever. It's like, no, it's not. It's not that. It's a spending problem. Everybody knows it. But it's gotten so acceptable to just say we should tax that or we should tax that. Everybody always wants to tax somebody else, right? If you're if you're not a smoker, tax the smokers. If you're not a driver, tax the drivers.

1:46:24If you're not a billionaire, tax the billionaires. All right. But there's just not enough money to satisfy the appetites of these folks. And so they're going to start going to wealth seizure. I'm here in California. There's a bill which I guess is expected to pass this year, which will be the first, you know, wealth seizing in California. I think in the United States, there's probably some legal expert that's got some other examples of places where it's happened in other ways. But just like you paid your taxes, you did everything right.

1:46:57And they're just going to come and take five percent. And this is a massive violation of I think everything that is American. But can this happen? Of course it can happen. And as the appetite for taking more money and goes up, could there be a 6102 again? Of course they could. What's the solution? Zcash, privacy coins, etc. Of course. And then. Yeah. Yeah. We'll see how it goes.

1:47:28Vlad CosteaIt's interesting that you're bullish on Zcash because you can hold your wealth in a shielded pool. And this is quite an interesting use case.

1:47:38Mike BelsheI don't think I'm the only person at all. I think I think this is I think this is part of why it's been going up in shielded mode activity.

1:47:47Vlad CosteaYeah, I agree with that. And I think the Winklevoss twins sort of promoted this as a response to AI and surveillance that's automated right now. And that's also a legit concern. It's much easier to track Bitcoin than it's been maybe two years ago or three years ago. Even with CoinJoins, it's easier to link. So, yeah.

1:48:10Mike BelsheYou know, stepping away to it's a more optimistic part of digital assets, crypto and kind of what's going on. So America has what we're calling this K shaped economy, which is basically the rich are getting richer and the poor are getting poorer. And the entire reason for this and I will boil this down to an oversimplified answer. Some people say this is too simple. It's not. It's because the cash is going down in value. So we are now at a stage in America where if you want to be wealthy, you have to have assets.

1:48:42And the problem is fewer and fewer people have assets and assets can be property. Of course, it can be gold. It can be Bitcoin. It can be investments in the S&P 500. There's lots of things. But you have to have assets, non-cash because cash is going purely down. And at the rate they're printing, it's going to start going down faster and faster. We just don't know what the slope of that line looks like. All right. So the problem is that the people that don't have a lot of assets can't acquire assets.

1:49:07They work, they get cash, they have to spend it for rent, they have to spend it for food, they have to spend it for car insurance, education, whatever. They don't have any money left to put into assets. They need to get out of cash for those that do manage to get a little bit. Maybe they got ten thousand dollars of AT&T stock or Tesla stock or whatever. You know, something bad happens. The car crashes. They have to get a new car. They have to sell their stock.

1:49:32Now they don't have the asset to replace the car, pay the bill or medical bill or whatever. So they keep getting pushed back into cash. And the cash, of course, is just going down in value. Digital assets do help this problem. The problem is systemic. So the way the American financial institution works, of course, the wealthy, they've got more. But also they graduate to a level where, like the investment banks and the financial institutions service them in a different way than the bottom part of America. So they can get loans against those assets.

1:50:07And of course, they got lots of assets and they get a loan against it. The asset keeps going up in value because not necessarily the asset went up, but the dollar went down. So when you're borrowing dollars, you're short dollars, right? It's the same thing as being short of stock. You borrowed them, you're going to have to pay them back. But if the value of that thing that you borrowed goes down, it's cheaper to pay it back tomorrow. Of course, if the opposite happens, it's worse, right?

1:50:30So if the value of the dollar actually goes up, then borrowing it becomes a lot harder. But it's constantly going down. There's no danger that it's going to go up in value at all, at least not with the way we're spending right now. So those rich people are basically borrowing dollars that are only going to go down. Which means they get to pay it back in 10, 20, 30 years. And guess what? It's going to be worth a penny compared to a dollar today. So they get the easy ticket.

1:50:53The people that don't have assets are in the worst case. All they have is cash. That's all going down. Every time they get a little bit of asset, they're not allowed to borrow against it because the big investment banks are like, you're too small. You're not worth my time. It costs too much to do this stuff. So they don't let you borrow. Instead, you have to sell. You have to sell your Bitcoin. You have to sell your stock. And you're back into cash.

1:51:14And you're back in this trap. And it's getting worse and worse and worse. The solve for this is tokenize everything. Get this out of the hands of the investment banks that cater only to the rich. Make it so that the cost of these financial services products is sufficiently low that we can offer it to everybody. Now, all of a sudden, you've got $10,000 worth of AT&T stock. Yes, you have to go fix your car. Yes, you have to pay that medical bill. But you're going to take a loan instead of having to sell that asset.

1:51:43And guess what? You've got seven years to pay it back. Guess what? By the time you have to pay that back in seven years, it's worth a lot less. So obviously, you have to pay interest on loans. But your assets are going to appreciate at rates that are faster than the inflation and devaluation of the dollar side. So you want to be short the dollar. We have to make it so that everybody can be short the dollar more. The only solve is if the politicians can figure out how to stop spending money, which I don't think they can.

1:52:15Vlad CosteaOkay, that became depressing very fast. Anyway.

1:52:18Mike BelsheNo, it's good. Like, it's good. Like, actually, people should be happy about this. Like, forget it. Like, the dollar going down. I mean, yeah, that one's kind of sad. I don't know what to do about it. That has nothing to do with crypto. But crypto is actually going to be the bridge that helps you preserve assets and wealth that you couldn't do if you didn't have crypto. And so this includes, of course, we got store wealth type of stuff like Bitcoin. We got shielded mode for just protecting yourself from others like Zcash.

1:52:43We've got then the tokenized stocks and equities and the stablecoins, all those things that are coming in, which allow all of this to be frictionless so that mere mortal retail people can get the same things that Jeff Bezos and all the rich people get, even though they don't have nearly the size of the assets.

1:53:02Vlad CosteaNow that you frame it like this. Yeah, it makes more sense.

1:53:04Mike BelsheBe optimistic. Yeah, yeah, it's definitely good. Without crypto, we would be in a much worse spot. But because of crypto, actually, I think we have a chance to bring everybody through. Lastly, I guess AI is the one thing. Like, if AI can bring this world of abundance, then maybe that brings all the prices of everything down and saves us in the end. I think Elon thinks it's going to go that way.

1:53:32Vlad CosteaQuestion. Do we have 30 more minutes according to the calendar or do you want to leave? Because by the rhythm in which you're speaking, I can understand that you're in a hurry.

1:53:42Mike BelsheI don't know that I'm in a hurry. Let me check my calendar here. That's good,

1:53:48Vlad CosteaBecause I wanted to push back on one of the issues that you presented earlier. Yeah, go for it. And I guess I sort of missed my opportunity, so I'm going to do it now. So on the topic of forks, I did not mention that they sort of create monetary inflation. And for a while, it's going to be a situation where the market needs to decide which one is legit and maybe some of them will die. But we're going to have more than 21 million coins. I don't know how many of them there should be on the market.

1:54:19I don't know who decides except for the free market, but it is confusing and difficult for a while to manage this. And we do inflate the monetary supply every time something useful that gets a lot of users comes around. So is it better than the old system if anyone can create their own coin and maybe become successful and have it traded for the same amount of US dollars, which means that it's going to diminish? I mean, of course, the US dollars are infinite, but momentarily it's going to diminish the upward potential of BTC or Zcash or something else.

1:55:00Like there's a cost for experimentation. That's what I'm trying to say.

1:55:04Mike BelsheOK, you're saying that every coin that gets created, if you create a Vlad coin and that's got any value, then you've just inflated the coin supply. That's what you're saying?

1:55:15Vlad CosteaI think this is like the Adam Back example of why Bitcoin forks are evil.

1:55:22Mike BelsheWell, I kind of disagree with the premise. I do think the quality of the asset matters. And this is a bit of a subjective and I don't know the math part of it. But look, I could I could create a coin right now. I call it Mike coin. I create one hundred billion of them and I sell you one for a dollar. And now my market cap is one hundred billion dollars because the latest trade was a dollar times one hundred billion of them. That's not real money.

1:55:56I didn't just inflate the coin supply. It's not the same caliber of, you know, a Bitcoin and Ethereum, a dollar or any of those things. Now, are some people going to be tricked and say, oh, it's worth a dollar. I'll buy it for 90 cents. Maybe. And this seems to be happening with meme coins. Right. Like people create some token, they give it some price. Some people have some value that they ascribe to it. But I don't think that means that you've created an increase to coin supply at the same level of a Bitcoin in any way.

1:56:32It's not like you can add the number of bitcoins and the number of Mike coins and say that that's the number of coins that exists. It's just apples and oranges. So I don't know. I think it's a very abstract way of thinking. And I don't think it's particularly useful. I do think those experimentations are going on there. And like when I create my coin or when Biden creates a laptop coin or when Trump creates Trump coin or whatever else, like people may decide to put value into it for all kinds of reasons, which are very different than like durable value.

1:57:08So, for instance, why does Dogecoin exist? I think Dogecoin's value is that it's funny. And, you know, people discover Dogecoin and it makes them laugh and the dog is cute. And so they put value in it and then it's survived. And Elon found it to be funny and spiked up the value because now you had kind of Elon saying it was funny and kind of introducing it to a number of people that also hadn't seen it before. And that was funny.

1:57:36Vlad CosteaIt's also useful on the Tesla store. And maybe that's also part of the reason you can buy merchandise with your Dogecoin.

1:57:42Mike BelsheOK. You know, but maybe that wasn't the best example. You can have coins that have value for a point in time, for humor, for public relations, for other things. And that's not the same type of value as like I want to put my money into an asset, which I think is going to be valuable 10 years from now. It's just it's just two different things. So I guess I disagree with the premise that like everything just inflates the coin supply as though all coin supplies are equal to begin with or something.

1:58:20Vlad CosteaYeah, that's fair. Something else I want to bring up because I feel like I did not do a good job debating it is when you mentioned that increasing the supply of Bitcoin with inflation is a breach of the social contract. And some people I mean, I agree with this, but I just want to extrapolate. Some people invested in Bitcoin with a certain expectation of it becoming useful in a certain way. I would say that a lot of people invested in Bitcoin with the expectation that they would use it for payments.

1:58:48And now that they're being pushed into using stablecoins is possibly a reason for them to become bearish on Bitcoin or have lower expectations from it and maybe sell on the short term or medium to long term. Because this was sort of the narrative for longer than 10 years. A lot of OGs got on board with the expectation that Bitcoin is going to be used for payments. And now we have fewer payments than we had last decade. So that's a big issue. These are two different

1:59:18Mike BelsheThings, right? So on one hand, you're talking about like, hey, should we change the Bitcoin protocol the way it actually works to have a larger supply or to have inflation? That would actually be like a change in the way it operates. Now, you can have all kinds of expectations. I don't know. You might think that, hey, you're going to be able to spend Bitcoin on Mars someday. I don't know if you're going to be able to spend Bitcoin on Mars. That's your choice if you want to think that way.

1:59:43And if someday you realize you can't trade it on Mars, you're disillusioned with it, you blame Bitcoin, fine. Go blame Bitcoin. In terms of the payments, I do agree. I think people were trying to figure out like, is Bitcoin going to be used for payments or restore value? I think originally we had hopes that it would be able to do both. I don't see any reason why you couldn't do both if you had a good type of money. So, again, you might be disappointed about it. But it's not like anything changed.

2:00:13It's not like there was something where we went to the protocol and we said, you can't do payments on this anymore. So, I think it's just two very different things. One is a technical change. Other one is an emotional impression about how it works. If you get disillusioned over time, legitimately or not, that's on you. That's not on the protocol.

2:00:33Vlad CosteaI sort of agree with that. But if you read the white paper or whatever, it's clear that this was intended for e-commerce or payments to avoid remittances, not remittances, chargebacks and maybe use other ways of paying with it. And I think Bitcoin is constantly in an existential crisis, which is driven by the market. And maybe that's fine. It's just the way of things and it's the way it evolves. But still, I wish I could pay more with Bitcoin. And whenever I can, I try to do it just because it's possible and I want to encourage it.

2:01:14Mike BelsheI don't want to guess at what Satoshi thought when he wrote the paper. He did say peer-to-peer electronic cash system, which kind of sounds like stuff you could pay with. And I also personally thought, yes, you could potentially use it to pay. In order to make payments work, there were questions at the beginning like, wait, how do you do a 10-minute block mining along with payments to pay coffee? And by the way, in the early days, like you go to LocalBitcoins, they end up meeting somebody at a coffee shop and then you had the awkward 30 to 60 minute wait for it to confirm, which is pretty hilarious.

2:01:58Anyway, so everybody all the way back in the beginning knew that that was something that would have to be figured out and solved. And then in terms of volatility and like, would you want to spend it? I think that's more of an emotional thing that we humans have that nobody really knew exactly how it would pan out. And then eventually we got to a place where we're like, well, wait, should we just think about it as a store of value without concluding that it's a one or the other?

2:02:25I mean, it could be used for payments today. There's no doubt that it can. You and I could decide to do a payment right now in Bitcoin. Maybe we do it in Lightning if we want to. It's just that Bitcoin has not evolved fast enough to keep up with the state of technology. And the reality is that stablecoins on pick your chain, Ethereum, Tron, heck, even Tempo, they're better payment systems than Bitcoin is for just doing electronic payment. So if Bitcoin wants to be a payment system, it has it does need to keep up with like the best state of the art.

2:03:07It is competing with other technology to fulfill that role. And then instead, the community, the ecosystem chose to be stable and not change and not pick up those new features. And in doing so, avoided hacks and avoided anything negative happening, but kind of ceded the payments market.

2:03:31Vlad CosteaI think that's a nice way of putting it. But if you look at the white paper, like you, if you look at it from a very, I don't know, specific point of view, you're going to be like, well, this is also anti BitGo or whatever, because it says without going through a financial institution, digital signatures provide part of the solution. But the main benefits are lost if a trusted third party is still required to prevent double spending. We propose a solution to the double spending problem using a peer to peer network.

2:03:59And it does legitimize in the introduction that commerce on the Internet relies too much on financial institutions serving the trusted third parties. And I don't think stablecoins solve this issue either. Stable coins are centralized by design.

2:04:16Mike BelsheLet's come back to stablecoins in a second. I don't think that the white paper says anything that's anti BitGo. Like, again, like, why does BitGo have self-custody and custody? Because we think maintaining that self-sovereign option is critical. I think that Satoshi was most likely libertarian. I don't know if he declared that officially one way or not, but I don't think he has any problem with somebody deciding to put their Bitcoin into a bank. He just wants you to have the option. Right. So can you use a middleman?

2:04:48I don't think Satoshi would care. I'm not entirely sure. But, you know that we haven't precluded the option to hold it yourself. And I actually agree. I think this is one of the most important parts, if not the most important part of digital assets, is that it gives us an option. If ever the financial institutions, the monetary system, the political system fails the people, it's having this escape hatch, the safety valve where we can go into self-custody and we can manage to live and preserve some wealth.

2:05:18That is the true value of the decentralization. But it doesn't preclude the other part. What's the second thing? You said something else there.

2:05:27Vlad CosteaAbout stablecoins requiring trust in third parties. Yeah.

2:05:32Mike BelsheYeah. Look, I mean, dollars are dollars, right? So the reason I say stablecoins are better payments, at least for now. Look, we're still, like I said, paying taxes in your fiat currency from Europe to the U.S. to China. They don't take Bitcoin as tax payments. So you are going to have to deal with fiat as long as there's somebody that's got control over your money that way. Yeah. The stablecoins do not solve the problem. That fiat is central, right? And you don't have to do anything.

2:06:10You can just sit there. You got your cash. And then the bank, the federal bank, can print as much money as they want. And the value of the thing that you're just sitting there holding didn't even do anything. It goes down in value. So, look, that's a centralized product. There's no fixing that. A stablecoin is just a tokenized representation of that fiat product. Yeah, you're correct. There's no decentralization there. By the way, the Iranians learned this as well, right? So I guess they were holding some Tether a couple of months ago, right?

2:06:42And then the U.S. government goes to Tether and says, hey, give me the dollars out of these wallets. And Tether just forked it all over, right? So even though Iran was holding the private keys to those Tethers and, you know, Tether couldn't take those private keys, of course, the Tether smart contract has the ability to freeze those wallets and then they just can fork over the underlying cash. Not blaming Tether, by the way. This is the way all the stablecoins work. That was just an example that, you know, politically came up just a couple of months ago.

2:07:16Vlad CosteaYeah, and I was going to say that I'm really impressed with your answer. I think you thought about this for a long time and you figured it out. But I also don't want my listeners to be like, oh, you got paid by the BitGo CEO to do a softball interview or whatever. Like, I really wanted to challenge you a bit to see how you're going to respond to this, because there's also this hardcore aspect of the Bitcoin community who wish we were on a different timeline, where maybe we were a bit more adversarial and we pushed back like the Monero people or something.

2:07:52Mike BelsheWell, look, thank you. Look, I don't think does Bitcoin pay you any money? Never. We're not even we're not even advertisers. So, like, actually, I appreciate you.

2:08:00Vlad CosteaI'm not going to refuse if you choose to, but I'm just saying you haven't so far.

2:08:06Mike BelsheOh, imagine how good the interview could have been. Wow.

2:08:11But let's see with two things. First off, respect that you always have kind of your own views. The people I respect most on Twitter and in our space are the people that have kind of their principles and they stick to those principles, regardless of kind of what decision they're trying to make. I guess since we talked about the block size wars, this still comes up a little bit from time to time on X. You know, some of the people out there will be, oh, Belshe was part of the New York agreement, which, by the way, is not true.

2:08:40I mean, go look it all up. By the way, there was a, I think, I guess they put BitGo's name on the New York agreement without my permission once, after I explicitly told them not.

2:08:48Vlad CosteaI think it's also the issue that Jameson Lopp left the company around the time and he sort of expressed some political disagreement, I guess. So maybe that's why there is this perception that BitGo was on the New York agreement side.

2:09:03Mike BelsheI don't think those are related. I would consider Jameson to be a friend to this day. I'd be happy to get on a podcast with him at some point. That'd be fun. He and I don't get to see each other enough anymore. But look, inside of BitGo, we did have different views. I would say, you know, BitGo was started as a Bitcoin company. It was before kind of this proliferation of lots and lots of coins. Co-founder Ben Davenport was in it for the Bitcoin, not for the Ethereum or any of the myriad of other coins that came along.

2:09:38Jameson also really liked Bitcoin much more than other coins. The guy that's building our Bitcoin is an ex-BitGo person. He worked on the Bitcoin team. We've always had a set of folks at BitGo that have a heart for Bitcoin specifically, and they don't want to work on other stuff. I don't have a problem with that. We've got a business that we have to build. I do think we built the right business in terms of supporting lots of different coins and lots of different technologies. But yeah, there's some people that decided to leave.

2:10:11If I had to guess, I think Jameson left because he wanted to work on Bitcoin only. Not that he was mad about BitGo and the block size wars, but we should ask him and find out. And then if I wanted to kind of describe my whole thinking about those block size wars, I hope you heard it earlier on the podcast, where it's like the IETF. Like, there's lots of people with ideas about how do we evolve these protocols. And what you're trying to figure out is like, what does the community want to do about how HTTP works, about how Bitcoin works, about how this Bitcoin improvement protocol works, whatever it might be.

2:10:53It starts with somebody having an idea, starts with somebody proposing that idea, then trying to get momentum around it. Without an official body, with an official set of rules, Bitcoin's a little bit more chaotic, a little bit more anarchistic. That's fine. I'm not complaining about it, but I hope that people would see I'm not religious about it needs to be an A or it needs to be a B. I'm just proposing what I think is technically the right thing to do. If the community thinks I'm wrong, the community thinks I'm wrong.

2:11:25It's fine. And I think if you go back and play back any of my comments from back then, and yes, I supported the SegWit2x, I think it would have been a very simple path forward. So I tried to make that happen. But in the end, the community said no. So it's fine. No big deal. I don't know why we have to demonize any group that was simply trying to make the protocol better.

2:11:49Vlad CosteaAnd then they routed around Bitcoin. And that's why Ethereum turned out to be such a great success. Part of the reason, like a lot of builders who wanted to put stuff on Bitcoin, they decided that maybe it's not the way to go forward. And they just went for Ethereum.

2:12:07Mike BelsheWell, I think Ethereum did a bunch of things that are very different. I mean, the smart contract was not doable in any similar way on the Bitcoin protocol. And I think Ethereum deserves all the credit for having really pioneered that part. I think then moving to proof of stake instead of proof of work. You might recall, you know, the early problems of proof of stake was the problem of proof of no stake, which is like...

2:12:38Vlad CosteaNothing at stake, I think it was called.

2:12:41Mike BelsheYeah, proof of no stake or nothing at stake. I mean, there's a couple of different variants of this, but like, you know, you could just vote for multiple things in order to get maximal for you, etc. Like, I think these things have been figured out, you know, slashing was one solution. It's probably not perfect. Might still, if you were to analyze that, you'd probably find a number of potential vulnerabilities that it would have in a truly antagonistic, truly decentralized environment. But then there's been more proof of, I'm sorry, more consensus layer innovations that happened on other blockchains.

2:13:15I think these are all great. And it's good that we've got other blockchains trying to figure out, like, how could I have a blockchain that's got some of the properties that I like from Bitcoin, but then I've expanded it with the smart contract program, like this amount of privacy that I want to do, the speed at which I want to be able to deal with transactions. And then, of course, all of those are trade offs. Like, you start putting billions and billions of transactions on train, like chain, like Solana does.

2:13:42And then, like, could you possibly run it on a small computer? No, it's kind of, it's not even close to being possible. So the decentralization is just a different level with Solana. And I think the Solana folks would argue, well, over time, the hardware gets better, and there's truth to that. And maybe that's a better way to go. But anyway, yes, it's true that other people have taken concepts from Bitcoin and then applied them with their own innovations. I don't think that takes away from Bitcoin at all.

2:14:16I think, in fact, we can now provably say it's helped.

2:14:21Vlad CosteaMoving back like an hour ago, when you asked me if I can request from an LLM to generate a BIP39 seed phrase, I did this with my local Qwen. And it took like 20 minutes, I could hear the computer getting, like the fans were spinning faster. It was stuck in a loop for a while, it downloaded the BIP39 library, then it downloaded some entropy random number generation library. And in the end, it actually gave me 12 words, which I don't trust that they're generated in a way that's safe.

2:14:54Maybe this is just theatrics, but it did give me 12 words.

2:14:58Mike BelsheAll right. Well, there's your next contest for your viewers.

2:15:03Vlad CosteaI got the work done. Yeah, I could put, oh, by the way, we should reveal the number. I guess the people are waiting.

2:15:10Mike BelsheI thought it was 17.

2:15:11Vlad CosteaOh, it is 17? This is like the inside joke, right?

2:15:16Mike BelsheOh, am I supposed to guess the number?

2:15:19Vlad CosteaNo, I've asked you to write a number from 1 to 100 for the giveaway. Oh, okay. All right. You forgot about that? You can pick it now.

2:15:27Mike BelsheI didn't know you meant me. I didn't know who you were talking to. Okay. All right.

2:15:32Vlad CosteaYou can go to random.org.

2:15:34Mike BelsheAnd then I said it was 17. Okay, random? Okay, let me see. Let me ask my ChatGPT. Oh, no. By the way, that was number 1 to 30. You want like 1 to 100, right?

2:15:46Vlad Costea1 to 100, yeah.

2:15:48Mike BelsheOkay. All right. I'm not going to ask it that.

2:16:03Okay, I have you a number.

2:16:05Vlad CosteaDrumroll. What is it?

2:16:09Mike Belshe91.

2:16:11Vlad Costea91. If there's a 91 in the chat, I'm going to find you and I'm going to send you $100. All right. If there is not, maybe next time.

2:16:24Yeah. It's weird that the chat is spammed with numbers and the only question that you actually got, which is formulated like a question for you. I'm not even sure if I should ask it, but whatever. Someone really wanted to know if you're Christian or Jewish.

2:16:43Mike BelsheOkay. I guess I haven't really talked about my religious affiliation. I don't have a lot of religious. I guess I was raised Catholic, but I'm not really religious at all.

2:17:01Vlad CosteaYeah, so I guess that answers the question. I guess I'm very bad and this is part of the reason why BitGo should not sponsor my podcast because I forgot to plug SideShift, which is the place where you swap your stablecoins for something that cannot be confiscated. It's actually topical for something that we spoke like 15 minutes ago. So SideShift.ai, not available in the US, unfortunately, because of the wonderful regulations. But otherwise, if you're not from the US, you can swap your stablecoins for Bitcoin, Litecoin, Ethereum, whatever it is that you like that cannot be taken away from you.

2:17:38And also due to the wonderful regulations, they don't support privacy coins anymore. They used to have Zcash and Monero, not anymore. Thank you regulators, but I hope you pass the Clarity Act today and also check out SideShift. That was my ad plug for SideShift. What else is there left to discuss? Do you have any closing topic? I guess I'm going to run out of time in two minutes.

2:18:02Mike BelsheI guess to some degree, like you said, it was a downer talking about certain things. Overall, I think it couldn't be better and more exciting. I think the negative stuff that's happening outside of the crypto world would happen with or without the crypto world. So this is a path that everybody's been on. And I think Bitcoin is a tremendous innovation and it's led to all of this. So I think it's really good. I guess one last thing we didn't talk about was quantum. Everybody wants to talk about quantum these days.

2:18:33So I do think it's one that we need to hustle on more. I suspect it's more about perception than it is about any actual vulnerability of Bitcoin. But here's what I think all Bitcoiners should be thinking about broadly.

2:18:51If we want Bitcoin to continue to be adopted, we always have to be thinking not necessarily about what our own personal views are, but how does the world think about Bitcoin and what do they see as specific threats that make Bitcoin not be palatable for the future? You know, some people are guessing. I think it was the BitMEX guy. What's his name?

2:19:21Vlad CosteaArthur Hayes.

2:19:23Mike BelsheYeah, I think he was just saying that, like, governments are all going to opt for gold rather than Bitcoin. I'm not sure I agree with that statement, but I do think we should think about what would it take for that to not be true. And I do think quantum does play a role here. So basically, imagine you're just a guy, you invest in stuff like you're not a technologist, whatever, and you hear quantum. In fact, maybe even have a portfolio company that you put some money into that says they're going to solve quantum.

2:19:56And then you're like, should I put money in Bitcoin? Isn't it going to get hacked by quantum? You don't really know, but there's enough talk about it. There's enough excitement about quantum computers entering into our world. I think most people agree that if there's one technology that's going to help us get there, it's that AI. It's going to help us get to quantum faster, not slower. So we need to address this and we need to do it now. So anyway, this is a solvable problem. I hope that the Bitcoin community can quickly hone in on a solution for a quantum resistance signature scheme.

2:20:38There will always be an opportunity for something better down the road. And some people might want to hold out for something better instead of doing something now. Look, we're going to have to take a little bit of a step. We don't want to put in bad cryptography, of course. So I mean, this needs to be done very, very responsibly. But these signatures are huge. They're huge on chain. Right? So we're talking about upwards of 100x bigger. Now, obviously, it shrinks and shrimps. They've got mechanisms to get it down so that it's only three, four times bigger, I guess.

2:21:11I might have those exact numbers wrong. But all of the things that we had about block size wars before are going to come up again. Right? Because you can fit one fifth the transactions into a block. Now, some of that will go into the witness data and not affect, but it still does require changes. All right. If we want people to use Bitcoin, we have to address the quantum issue, even if it is more of a perceived threat than a real threat, because it's not going to go away as perceived threat.

2:21:43Vlad CosteaI agree with that. And I think Nick Carter is doing a good job, even though he does come across like a quantum investor who's trying to pump his bags at this point of view. But I also see that there was like a lot of talk about quantum last year and BIP-360, which was like the most concrete example. Not so much anymore for some reason. I don't see as much talk about quantum these days. But also, once we get to the point where we need to migrate, because that's the other issue, you need to move your coins to quantum resistant addresses.

2:22:24Mike BelsheNow you're advertising BitGo. All right.

2:22:27Vlad CosteaYeah. It's going to be like hell. Like you need to pay a very large fee to get into next block.

2:22:34Mike BelsheNo, I disagree. Okay. So there's a few phases, right? So typically we're thinking about these serialized things have to come one after the other. First one, update the protocol to use a signature scheme that's quantum immune. Number two, get that signature scheme implemented in every wallet, every exchange, everybody that touches Bitcoin. Number three, all the users have to move their money. And then number four, deal with Satoshi. All right, let's leave number four off because that's more of a social issue than it is a technical issue.

2:23:04All right. We don't have to do these things serially. So BitGo launched and sometimes people are like, Mike, this is ridiculous. We have a quantum resistant wallet today. And what do we mean by that? Like, look, how do you be quantum resistant when the signature scheme is not fixed? Well, it turns out that if you haven't exposed your public keys by having spent on a wallet before, then you're actually not quantum vulnerable today. We should have every wallet, every exchange on the planet moving to an unspent selection mechanism, which we didn't do a couple of years ago.

2:23:38We used to use unspent selection mechanisms, which optimized for low fees. Now we need to change it to optimize for not having exposed public keys. And most wallets are not, in fact, I would argue almost all wallets are not optimized this way. Do you want me to give you an example so that your viewers understand this?

2:23:56Vlad CosteaOh, yeah. It's also good for the shorts when it's going to be cut into a reel or something.

2:24:02Mike BelsheOkay. Hopefully this is pretty simple. You've got a wallet, you receive one Bitcoin at address A. You create a new receive address to receive another one, but that user still sends it, somebody sent it to that old address. Now you receive two Bitcoins in two unspents on the same address. Then you go to spend a half a Bitcoin. Historically, what a wallet would do is it would select enough unspents to cover the half of Bitcoin. It would just use one of those Bitcoins. So it takes one of those unspents, uses it, sends you half a Bitcoin, puts half of it back in change in a new address.

2:24:36But now you've left a Bitcoin, which was received at that address, exposed. The public key, which corresponds to that address, has been used. What the new BitGo quantum resistant one does is very simple. When you make that spend, all of the coins that came in at the same address all get spent at the same time. So we don't leave things exposed. Now, there is a degenerate case you have to worry about. Like if you had hundreds and hundreds and hundreds of transactions, believe it or not, this does happen with some of the machine to machine stuff that's going on.

2:25:10Those still have to be consolidated kind of through offline processes or separate processes. But for regular people, this can absolutely be done as you go. And look with wallets, you're constantly trying to balance having enough unspents to cover different kinds of spending things that different people do, et cetera. But you can optimize for quantum resistance as opposed to optimize for low fees. And by the way, the fees are so low on Bitcoin right now, you won't really notice a difference. It'll be a little bit more, but you can hardly tell.

2:25:37We can do this today. What that means is we can get steps two and three more or less in place today. There's a few places where there's some edges like, you know, in the mempool, like if a quantum computer showed up that could, you know, hack your coin in less than 10 minutes. And as soon as you showed up in the mempool, if that happens, we'll deal with it. Look, the estimates from the last week is that like the first quantum computer is going to take something like 70, I think it's 73 days to hack a single key.

2:26:08So use multisig, you'll have to do 73 day period. All right. Anyway, we can do some of these phase two and phase three today. Every wallet should implement an unspent selection mechanism, which leaves no public key exposed as best as possible. And then we can protect basically all the existing coins. This gives us more time to solve problem number one, which is to get the quantum resistance signature in place. And then lastly, we still have to deal with Satoshi.

2:26:36Vlad CosteaI think that's going to be the most controversial part. Like, do you just leave the coins as a bounty to whoever figures out how to crack the cryptography? Are you going to freeze them? Are you going to reissue them as mining rewards? That's another way to secure the network long term.

2:26:54Mike BelsheThis is one where I would probably lean on Jameson Lopp. I think he spent a lot of time thinking about this. I think he's got a pretty good thinking about it. In general, I think the game theory is that you just freeze the coins. So I will caveat this, like the first thing is like, I'm antithetically opposed to like freezing the coins. It's just completely against everything in Bitcoin. But then you play this out, right? And what's going to happen is there'll be a fork. Somebody is going to freeze the coin, somebody is not.

2:27:30And then the market gets to decide. And then where are you going to put your Bitcoins? And you're going to put it on the one that has been frozen. I think the monetary interest lands us there. So anyway, I don't want anyone to think that I'm in favor of freezing anybody's coins. I absolutely am not. I can prove that. Wrapped Bitcoin does not have a freeze function in it, unlike every other stablecoin on the planet. That's because we don't believe that you should be able to freeze from the stablecoin issuer.

2:27:58But I do think that's how the game theory is going to play out.

2:28:03Vlad CosteaYeah, and speaking of... Wrapped Bitcoin is still the largest scaling layer of Bitcoin. It's like 120,000 BTC, much larger than Lightning, Liquid, Rootstock, whatever. So I got to give you credit for that. Also, I want to bring the interview full circle because when I first contacted you like two months ago or something, you put a bounty for Anthropic. I think you put 100 Bitcoin in a multisig address and you asked Anthropic to hack it.

2:28:33Mike BelsheI think I did, yeah. Is it still there?

2:28:36Vlad CosteaIt should be still there. Do you trust that Anthropic actually has good systems for cracking that?

2:28:45Mike BelsheWell, this gets into a little bit of the politics of what's going on with Anthropic and OpenAI. And should we have regulation of AI right now?

2:28:56I do not think we should have regulation of AI, certainly not today. I think it's way too early. But I don't think that regulation will get us the result we want. On the AI front, I think the biggest fear I have is that they will make AI be woke. And we will only be able to address it in kind of a woke manner. We won't be able to use it unless you have permission. And I think that's a terrible mistake. The other thing is that we are also from a technology perspective in a period where we need the best AI tools for defenses.

2:29:37So the entire world has accumulated decades of human written crappy code full of bugs. Do we agree on that statement?

2:29:49So what do you do when you've got decades full of human written code? Well, we have to go through all of it. We need the best tools right now. We cannot wait. You have to be investing in this. If you are a security company at any level, even if you just have PII, you need to be using AI in order to find the bugs before the attackers do. And it's just a matter of time. So we're in a race. If we were to put any regulation at all.

2:30:19That delayed even by a couple of days. Access to this, I think it's a problem. So BitGo has access to Mythos. And I don't know, maybe everybody has it now or maybe I guess there's the hacking version.

2:30:32Vlad CosteaNo, we just have Fable, which is not Mythos. You have the Holy Grail, right? The one that escaped its cage or whatever.

2:30:44Mike BelsheBut to get it, like you have to get permission from the state. And that means if you believe that that's the right path forward, that means you are delegating to the government the decision for which companies succeed and which ones fail. But I fundamentally believe that if you don't have the best defenses with AI right now, you are vulnerable. And so is JP Morgan allowed to have access to this and like the community bank is not? I mean, forget about crypto. Like, what about those guys? Anyway, strongly oppose that.

2:31:24Got off topic. You know, we've seen over and over again that OpenAI and Anthropic seem to have agents that get out of their sandbox. I mean, come on, guys. It's not hard to build sandbox. I do not technologically understand how this could be the thing they can't contain. They absolutely can contain it. They don't want to contain it. It's a marketing stunt. And so to prove it, yes, put forth a Bitcoin challenge for them to go get the Bitcoin. So in a Bitcoin wallet, go get it.

2:31:58Vlad CosteaAnd something that we have in common, and I'm going to let you go. I don't want to drag this forever, but open source software sort of solved this gatekeeping of powerful frontier AI for hacking. Like we have Kimi K3, which you can run on a server. It does not run on my local computer. It needs like 200 or more gigabytes of RAM. So it needs a lot of power. But Kimi K3 is amazing. Like I've tried it on some websites that grant me access to it. And it's amazing.

2:32:26I've used it to harden the security of my website. I've used it for auditing some stuff, which I thought was useful for myself. Like for self-defense, it's an amazing tool. It's open source. And it's also the counterpoint to this whole regulation idea, because it's made in China. It's open source. If the U.S. somehow decides that it's time to put a hold on innovation and AI, the Chinese are not going to stop and they're just going to build more powerful tools and the U.S. is going to lose the arms race.

2:33:00Mike BelsheAnd China has said that explicitly at this point. So, yeah, I think open source absolutely is a huge part of it. For businesses like ours, I mean, it's the tokens per second. You need to be able to drive through these things. And so the problem is you have to be able to do that with hardware somewhere. So you have to be able to run those models on large server farms in order to be able to do things kind of at speed and at scale. So obviously that'll get better over time as well.

2:33:29But the open source models are great. They do help and we do use them. But at least right now, I mean, I think we still have to be very careful about on the frontier models. Are we selecting who gets access and who doesn't? Because that's the same as saying who gets to win, who gets to lose.

2:33:47Vlad CosteaI agree with that. So how can people follow you and pay attention to whatever it is that BitGo is doing? And also, what do you think are the odds that the clarity act will pass today? It seems like Polymarket says 70 percent.

2:34:06Mike BelsheLet's see, I'm on X @mikebelshe. So probably you found this podcast on X is I'm guessing. And I post a fair amount. Thank you for your support. If you want to try BitGo, by the way, bitgo.com. Like I said earlier, we'd love to have your feedback for anybody that's here. I'm grateful to everybody that walks through the doors and tries it. And then on clarity, well, I guess just a correction. I think today is not the final vote. Today is the vote for cloture, right? Which allows...

2:34:35Vlad CosteaOh, Senate, yeah. And then there's the review room, which is we call them deputies here. It's representatives in the U.S., right?

2:34:44Mike BelsheYeah. So, look, I'm an optimist, so I'm going to say I think it's going to pass today. I don't know if that means it passes overall or not in a couple of days, whenever that would be. But I'm an optimist for it. Whether it happens or not, I think we're actually in a pretty good place. I do think it would be better. It would allow all of traditional finance to come in and they wouldn't have to worry about their regulators anymore. On the other hand, selfishly for a company like BitGo, look, we know how to navigate the crypto space.

2:35:17We know how to take these risks. We've been doing it for 12 plus years now. It might actually be better for us if it doesn't pass. The reason I want it to pass is because I think we want everybody in the space. I think the rising tide will carry everybody to a better destination. Selfishly, BitGo would probably be better off if it didn't pass. But yeah, I hope it passes.

2:35:39Vlad CosteaI do have to remark that this is very rare, as in the AI industry, they want to regulate because they want to gatekeep who's going to be building frontier models and basically close the door behind them. But in the crypto space, we have BitGo, which actually wants competition.

2:35:58Mike BelsheI guess so. Well, you could argue like, hey, is there something in this regulatory? We're qualified custodian, blah, blah, blah. And are we trying to get regulatory capture? No, I actually hope that we disintermediate most of the custody components of market structure. Remember, clarity doesn't actually answer the question of exactly how the market structure unfolds. What it does is it authorizes the SEC and the CFTC to go start doing that. It does codify immediately some definitions of what falls into which camp, but it starts the process on getting the rulemaking and leaves the air cover that it comes all the way from Congress.

2:36:39So there's a ton of work to be done, and we'll see how it works out. But yes, I would like to see fewer middlemen. I would like to see crypto and digital assets fulfilling the promise of helping stop the K shaped economy, reversing that to some degree, making things more transparent, reducing middlemen, reducing prices, all that. We failed as an industry so far to do that. So BitGo will keep fighting for it.

2:37:09Vlad CosteaThank you very much for your time. I want to tell my listeners that after this interview, they should go to YouTube and search for Satoshi versus Alexander Hamilton. That's a rap battle video where Mike Belshe is also part of the people in the room who are reacting to it. And also, you're going to see a pre-fame version of the CEO of OpenAI, which is kind of amazing. Sam Altman was there and nobody knew who he is, or maybe that some insiders knew what he's up to, what he's going to be building.

2:37:42But at the time, he was still unknown to the masses. So yeah, check that out. It's a lot of fun, and it's a good reminder of why we're here. It was fun.

2:37:53Mike BelsheI think Reid Hoffman put that one together, by the way. That was filmed over at the LinkedIn offices.

2:38:00Vlad CosteaIt's possible. I don't know the background details, but it's a lot of fun.

2:38:05Mike BelsheIt is fun. It's great. Rap battle is always good. Hey, Vlad, thank you so much. Great to see you again. Thank you for being here. Take care. Bye.