Title: S17 E17: Lukas Hozda on BIP110, Bitcoin & Rust
Podcast: Bitcoin Takeover Podcast (host: Vlad Costea)
Guest(s): Lukas Hozda (Braiins; author of "Building Bitcoin in Rust")
Published: 2026-04-07
Duration: 1:38:38
Episode URL: https://bitcoin-takeover.com/audio/?name=2026-04-07_s17_e17_lukas_hozda_on_bip110_bitcoin_rust.mp3
Audio: https://bitcoin-takeover.com/audio/media/2026-04-07_s17_e17_lukas_hozda_on_bip110_bitcoin_rust.mp3
Transcript: CORRECTED. Base: local Whisper medium.en transcript with automatic speaker diarization, reconciled against a full local small.en (float32, VAD) re-transcription with word timestamps and the YouTube auto-captions of the recording (three-way comparison over the whole episode; the small.en pass hallucinated at 00:29:58-00:30:37, where the base transcript and captions were used). Because the episode has many short back-and-forth turns, the transcript was rewritten paragraph by paragraph from the three sources and re-split by hand. 12 doubtful stretches were re-transcribed locally from short audio clips with medium.en (int16, VAD). Speaker turns were checked with a full-episode local speaker-embedding scan (TitaNet reference voices for Vlad and Lukas, 1.5 s windows at 0.75 s hop), including the turns at 00:01:39, 00:02:40, 00:12:33-00:12:43, 00:18:52, 00:29:30, 00:40:46, 00:42:54, 00:48:11, 00:50:51, 01:05:30, 01:09:14-01:10:00, 01:10:42, 01:20:02-01:20:31, 01:28:22-01:28:57 and the giveaway exchange at 01:29:20-01:31:22. Names and jargon were checked against the title, Vlad's show notes and the web: Braiins, Braiins OS, Braiins Pool, Braiins Hashpower, Braiins Forge, Braiins Deck, BMM 101 mini miner, Slush Pool, Jan Čapek (Stick), Kristian, Between Two ASICs, Arno's Engram layout, Bitaxe, Cake Wallet, SideShift, Breadwallet, Bitcoin Knots, Libre Relay, libbitcoin, Grafana, JBOK, BIP 32, The Big Book of Bitcoin, Kixunil / Martin Habovštiak and his TIFF of Luke Dashjr embedded via MARA Slipstream, Super Testnet, Marathon / MARA Pool, OP_RETURN and Core v30, BitMEX Research, Utreexo, LevelDB, StarkWare, Andrew Poelstra, bitcoin-dev mailing list, Rootstock (RSK), BIP 444 / BIP 110 (256-byte limit, quoted from the BIP), CSAM, SegWit2x, Mark Karpelès and Mt. Gox, DMCA and safe harbor, Hunter Beast and BIP 360, hvozd, Škoda. Light clean verbatim: ums, false starts and stutters removed; factual slips by speakers are kept as spoken. Intro music omitted. Remaining doubtful words are marked [unclear: best guess].
================================================================================

[00:00:46] Vlad Costea: Hello there and welcome to the Bitcoin Takeover Podcast. This is Season 17, Episode 15-ish, by the time I publish it. It might be 16, 17, I don't know. I'm next to Lukas, or Lukáš, who is the author of this beautiful book, which is...

[00:01:02] Lukas Hozda: Building Bitcoin in Rust.

[00:01:04] Vlad Costea: Yeah, I couldn't read very well. I wasn't seeing if the first word is Bitcoin. It could be like Bitcoin Rust or something. But yeah, it's massive. It's published by Braiins, which is, once again, the place where we are recording this. Lukas, you're working with Braiins and you have been an employee for like seven years or something.

[00:01:24] Lukas Hozda: Oh, no, not seven, like five. Five. Yeah, I went here right after high school.

[00:01:31] Vlad Costea: So you're much younger than you look.

[00:01:35] Lukas Hozda: Well, thank you. I look old. No, I don't know.

[00:01:39] Vlad Costea: No, it's what Bitcoin does to you. Made you look a lot older.

[00:01:44] Lukas Hozda: It's the years of stress. It made me an old and withered man.

[00:01:48] Vlad Costea: It's the bear market.

[00:01:50] Lukas Hozda: Indeed. I will light up and look like an elf when we go back into a bull market. We reach an all-time high and I start glistening.

[00:02:00] Vlad Costea: Fabulous.

[00:02:01] Lukas Hozda: I am.

[00:02:02] Vlad Costea: And you also do some interesting work with building your own keyboards, which I found interesting.

[00:02:07] Lukas Hozda: Oh, yeah. I have a lot of distracting hobbies, and I can show to the camera. I make these split keyboards, which are very ergonomic. I got very interested in the ergonomics of keyboards and been making more keyboards than I will ever need, just trying to get the perfect setup for using Emacs.

[00:02:36] Vlad Costea: You made me try them yesterday. You're using different layouts. This one is B-Y-O-U.

[00:02:44] Lukas Hozda: This is Arno's Engram, is what the layout is based on.

[00:02:49] Vlad Costea: And you also said that you have different pressure types according to the finger.

[00:02:52] Lukas Hozda: Yeah, because different fingers have different amount of strength in them. So, you know, you want, let's say, stiffer switches for stronger fingers and vice versa. Naturally, your thumb is the strongest.

[00:03:10] Vlad Costea: Okay. We established from the beginning that you're a nerd.

[00:03:14] Lukas Hozda: That is true.

[00:03:15] Vlad Costea: You are 100% a nerd.

[00:03:18] Lukas Hozda: Indeed.

[00:03:19] Vlad Costea: And part of the topic of this episode is gonna be your review of BIP 110.

[00:03:28] Lukas Hozda: Yes.

[00:03:29] Vlad Costea: You spent a lot of time looking at that one. Kristian, when we recorded the episode yesterday, he hyped that you're the research person who looked into that, and you tried to have a very open mind and be objective. We can save the topic for maybe 30 minutes from now, so that people, you know, we gave them the hook to keep listening.

[00:03:50] Lukas Hozda: Yeah.

[00:03:50] Vlad Costea: And then they're gonna have, like, the main conversation. Let's establish in the beginning how you got into Bitcoin, because we had some beer yesterday evening and you told me some very interesting stories about computer magazines that told you how to mine Bitcoin here in Czech Republic. It's amazing.

[00:04:08] Lukas Hozda: Yeah. Yeah. So there was this PC magazine that had, like, an article in about 2013 or so about Bitcoin, and it was presenting it as just, like, cool new technical thing. And, you know...

[00:04:26] Vlad Costea: For the record, how old were you in 2013?

[00:04:29] Lukas Hozda: Well, mostly 12. My birthday is in September, so two-thirds of the year I was 12 and one-third I was 13. And I was subscribed to this magazine, or actually, well, my grandfather was subscribed, and when he was done with a number, he would give it to me as a hand-me-down. And there was this article about Bitcoin that was, like, purely technical, presenting it as this cool new thing, and you can use it to pay for stuff. It's permissionless. And, you know, it wasn't making any arguments about if it will be a bubble or if it will have any economical viability or no, it was just technical. And I found it really interesting. So I decided to play around. I mined some bitcoin with, actually, Slush Pool. Not much bitcoin. And I was baking the family computer doing it. So I think that there is an account here in Braiins, in Braiins Pool, lost to the sands of time, because I couldn't figure out what email and password it was for the life of me. I made so many. That has some fractions of bitcoin I mined, like, in 2013, 2014.

[00:05:52] Vlad Costea: Do you still have that magazine? Is there any way to recover that article? Because I feel like it's such a historical piece.

[00:05:59] Lukas Hozda: Yes, yes. It's in my parents' flat somewhere. I know I didn't throw out any of the numbers, so it exists somewhere, and I should be able to recover it if I find it.

[00:06:14] Vlad Costea: What's the name of it? PC Magazine? Chip?

[00:06:17] Lukas Hozda: I think it was PC Magazine directly, which is, I guess, not the most SEO-friendly name, but it was a good magazine. I liked it back then.

[00:06:34] Vlad Costea: I wonder what the person who wrote the article is doing these days. I should look that up.

[00:06:39] Lukas Hozda: Yeah, that's really curious. I don't remember who wrote it, but I wonder if they got into Bitcoin themselves and if they're still in it, because the landscape is drastically different. You know, I don't think that that many people expected that Bitcoin would be as priced as it is today, back then, because people tend to be quite shortsighted. So every time there is an all-time high, "Oh, it's a bubble." Every time we go down, "Oh, it's a crash," and, you know, "It's over." And it just never tends to stick with Bitcoin.

[00:07:23] Vlad Costea: That's an interesting way of putting it, because Stick was someone who also worked on Slush Pool at some point.

[00:07:28] Lukas Hozda: Oh, yeah. Yeah. Yeah.

[00:07:30] Vlad Costea: Yeah. So you discovered Bitcoin when you were 12. I want to ask the people listening to this, what were you doing when you were 12? Don't be very graphic about it, please. But where was I going with this? How did you come from just being this teenager who discovered Bitcoin to getting a job at Slush Pool slash Braiins now today? And you're sort of a jack of all trades around the office, as far as I know. Like, you're not assigned to a specific department. You just switch from one job to the other.

[00:08:11] Lukas Hozda: I am assigned to a specific department at any point in time, but it just keeps changing quite often. So right now I'm under marketing. Before then I was under HR. I had my own education department, and I also was an embedded software engineer working on Braiins OS. I think this is my fifth team in Braiins. So it's because the things that I do are kind of overreaching any individual field. So I am originally a programmer, and I am by far mostly a technical person, but I'm quite social. And so I give talks and I write articles on behalf of Braiins, or contribute to newspapers. I also do lectures, I teach, and I try to hire the students that are promising and are excited about working in Braiins. So all sorts of things. I also like helping out with less programming things. You know, I helped out with the recording of the other episodes here in Braiins, and I hope I will help with the ones that come after me here, while you're still here. And so it's difficult to place me, especially when it comes to the budget, because, you know, out of which department's budget should my activities be? And it's not easy to decide. So they keep placing me in different departments, even though I still do the same thing.

[00:09:49] Vlad Costea: I have an actual curiosity here, because Braiins is a mining company...

[00:09:53] Lukas Hozda: Yes.

[00:09:54] Vlad Costea: ...but I haven't seen any mining happening in the office.

[00:09:57] Lukas Hozda: Oh, well, there was a time that there were miners in this very room that we are in. Turns out miners can get quite loud. People don't really like it when there are loud miners. And also this room was not really usable as a podcasting room. We built this to be a podcasting room, but then we figured out, hey, we have all this sound foam around here...

[00:10:23] Vlad Costea: So what you're saying is that you're using this to suppress the noise coming from the ASIC miners.

[00:10:28] Lukas Hozda: Yeah.

[00:10:29] Vlad Costea: Then you decided that you can record interviews.

[00:10:32] Lukas Hozda: No, no, no. We built it to record interviews. Then we found out, hey, we can put miners here. So we put miners here, and then it was hell here. It was so hot that no one really wanted to record any podcasts here, because even if you turn the miners off to record a podcast, the room would be hot for a long time, because the AC doesn't work that well in this room.

[00:10:55] Vlad Costea: I think, like, four or five years ago, I don't even remember, there was this podcast in Bitcoin called Between Two ASICs, and that was the whole concept, that the people talking on the podcast were between two ASICs that were running.

[00:11:10] Lukas Hozda: Must have been like in a helicopter, you know, where you can't hear the other person. "Hey, what are you saying?" "Bitcoin, what?" Yeah. And then we figured out, hey, maybe we probably shouldn't have the miners here. So we have a couple of, as we call it, mining rooms around Czech Republic. That is where we store the miners, and it's a much better environment for the miners. We can have quite a few miners there, and we can run tests on them and all sorts of stuff. So it works out better than having them in the office.

[00:11:46] Vlad Costea: Well, I'm curious which department complained the most about the noise, because I guess it wasn't the engineers, because the bread and butter...

[00:11:54] Lukas Hozda: No, I think it was actually the engineers, because they used to sit right there.

[00:11:59] Vlad Costea: Okay.

[00:12:00] Lukas Hozda: Braiins OS developers. And we did this in the same breath as we split off the hardware department into its own office building, because turns out that the sound of drills is not conducive to focusing well on what you're programming.

[00:12:19] Vlad Costea: So... But it's not like a drill. It's more like a vacuum cleaner.

[00:12:23] Lukas Hozda: No, no, no. The hardware department was using actual drills to make stuff and fix stuff.

[00:12:28] Vlad Costea: Oh, interesting.

[00:12:29] Lukas Hozda: Yeah.

[00:12:30] Vlad Costea: So does this hardware department also have an office here in Prague?

[00:12:33] Lukas Hozda: Yes. Yes, they have an office in Českomoravská, and some time ago...

[00:12:39] Vlad Costea: Sorry for the dox, guys. I'm really sorry.

[00:12:41] Lukas Hozda: It's public. It's on Google Maps.

[00:12:43] Vlad Costea: So can we go there and get some ASICs fixed or something?

[00:12:48] Lukas Hozda: Oh, I'm not sure. I don't think they do that necessarily that much anymore. Like, if you ask, they probably could fix an ASIC, but they focus on other hardware products. We recently split them off into its own business unit, so they only share some resources with us, and they are making hardware products straight to customer. Yesterday with Kristian, I think you maybe discussed the Braiins Deck or the Braiins Mini Miner, one of them.

[00:13:18] Vlad Costea: Yes, there was a giveaway.

[00:13:20] Lukas Hozda: Yeah, yeah, yeah. So that's made by the hardware department that we call Braiins Forge. It's like a sister company under the Braiins brand.

[00:13:30] Vlad Costea: I'm actually curious. You have the Mini Miner, the BMM 101, which, by the way, you can buy for a 5% discount with promo code BTCTKVR. And does it use original chips that are brand new and were manufactured for this purpose, or does it repurpose old chips from other ASIC miners, from Bitmain or something?

[00:13:51] Lukas Hozda: Yeah, yeah, yeah. It repurposes chips. I'm not sure, I think they are from S19s, the chips that are in there. I mean, you can pull it apart and take a look, you'll figure it out straight on the chip, what type of chip it is. But yeah, they are repurposed. It would be quite a large amount of work to design new mining chips straight up.

[00:14:17] Vlad Costea: I was always curious about this, because there is also the Bitaxe project, which repurposes chips from ASICs which maybe are not efficient anymore for large mining farms.

[00:14:28] Lukas Hozda: Yeah, yeah, yeah. It's very similar. The main difference is that the Mini Miner is like a big miner, but small. So it's running the same Braiins OS. You can use all the tools that you use to manage large ASICs, but it's tiny, it's quite quiet, and it doesn't consume much energy. It actually, I think it tops out at about 40 watts. So it produces less heat than, you know, those little candles, we call them the tea candles. So it produces less heat than that small candle, actually.

[00:15:08] Vlad Costea: That's interesting.

[00:15:09] Lukas Hozda: So you can use it as, well, it's a novelty, but you can also use it as an educational tool, where you get it, you learn how to operate the software, and it will be essentially the same if you get a bigger ASIC. But the bigger ASIC will be more investment, and it will be louder, and it will consume more energy. So this is like an easy way to dip your toes in.

[00:15:33] Vlad Costea: I want to ask you, I remember from yesterday that you seemed to be in favor of drivechains. What do you think about this idea of having merge-mined sidechains, which are enforced by miners, and miners also collect the transaction fees from them?

[00:15:50] Lukas Hozda: Yeah, so I like it, because it's a way to promote experimentation in a way that's not really harmful if it doesn't work out, because the sidechains can be added, and I'm not sure if you can remove them per se, but you can override them. There is a cap, I think it's like 256 sidechains, which is quite a big amount. And I think that it's a way to test out new ideas in a way that's more significant than if you're running things on testnet and so on. And that it's a way to bring the functionality that some people think is missing for Bitcoin. Me, I'm fine with what Bitcoin does, but a lot of more cryptographically oriented people, for instance, want to be trying out new things, and bringing them to Bitcoin can increase adoption by these types of people, I think. So that's why I think it's a neat idea. I know that there is quite a lot of debate around it, and that Paul has a lot of work to do constantly. But yeah, I think it's one of the better ideas that are being discussed.

[00:17:13] Vlad Costea: Yeah. Check it out right now. There's software that you can actually run on the Bitcoin signet. It's at layertwolabs.com/download. Can be like a fun weekend project, just to run these sidechains and play with them with testnet coins. There's also a group where people can send you coins for free if you want to run the experiment, so you don't have to spend time mining them. It's pretty cool. I also want to ask you, and this is also a sponsored question, where do you get your Bitcoin news?

[00:17:43] Lukas Hozda: Where do I get my Bitcoin news? Yeah. Oh, mostly memes and Twitter, of course.

[00:17:49] Vlad Costea: Okay, so you never open, like, a news website?

[00:17:53] Lukas Hozda: Well, when someone links to it, I sure do. But no, not that much. I just watch what people do. And of course, I read the Braiins blog whenever it comes out.

[00:18:05] Vlad Costea: Nice. Nice. Anyway, check out Bitcoin.com News. What I like about them is that now they know the difference between Bitcoin Cash and Bitcoin. They used to be this website that was saying that Bitcoin is slow and expensive and you should use Bitcoin Cash. Now they are actually orange. And they also have news which I think are striking a good balance between institutional and political. They also have some grassroots coverage of events happening in communities, and I think it's good journalism. I like that. Check it out. It's news.bitcoin.com.

[00:18:46] Lukas Hozda: That's pretty neat. news.bitcoin.com? Yeah. Ah, okay. Good.

[00:18:52] Vlad Costea: It's a good domain. And it is probably super expensive. Like, how much do you think Bitcoin.com is worth?

[00:19:01] Lukas Hozda: Probably a couple million dollars, I think, with, you know...

[00:19:05] Vlad Costea: No, it's probably a lot more.

[00:19:07] Lukas Hozda: Goddamn. Goddamn. I've seen some very, very expensive...

[00:19:12] Vlad Costea: Just think about the businesses, like exchanges or whatever, that would like to have Bitcoin.com.

[00:19:19] Lukas Hozda: Yeah, such a widely applicable domain. It's quite neat. Yeah, but as I said, I don't read that much news. It's mostly memes on Twitter.

[00:19:35] Vlad Costea: Yeah. So also the Braiins blog is excellent. There's also a newsletter. You can check that out, you can register to it. I think the URL is Braiins, with the double I, dot com slash newsletter, and you can sign up for a weekly article that you're gonna read. Very cool. And you guys also have the new Hashpower.

[00:19:58] Lukas Hozda: Yes, we do. It's a product that we've been working on for quite some time. We really wanted to get it right on the first try, so to speak. For a time, some people here were like, "Oh, we'll never finish it." But we did finish it, and it was a lot of work, and it works surprisingly well, and it exceeded our expectations. We immediately started running out of capacity, because the demand was so high, and we have to keep increasing the capacities. It's insane. It's great fun. I think that in recent times, two blocks were found using hashpower from Braiins Hashpower. One was completely on a solo pool, and the other one was on the Parasite pool, which is like a split model where, I think, the person who finds the block gets a whole bitcoin, and the rest is distributed between all the users mining on the pool.

[00:21:00] Vlad Costea: Interesting.

[00:21:01] Lukas Hozda: Yeah. And it was like really lucky, you know. The person purchased just some, I'm not sure how much it was, but it was only hashrate for a couple of hours, and they found a block like that. So imagine getting such a return on your investment. Like, I'd be like, "Okay, I'm done."

[00:21:26] Vlad Costea: It's a lottery ticket, but you can use your bitcoin to purchase hashrate, which far exceeds what you can get with a farm of Bitaxes or BMMs or whatever. You can get, like, a thousand terahashes, as opposed to one terahash, which is what you get with the average Bitaxe. And it's not that expensive. The difference is that you can't argue that you're decentralizing mining too much. You're just renting it from someone that's producing hashrate more efficiently than you can at home. But if you're actually owning the hardware and you're running it, you're contributing to mining decentralization, even if you're like a drop in the ocean. Pun intended.

[00:22:12] Lukas Hozda: Yeah, exactly. It's also, I think it's a great way to dip your toes in, you know. If you purchase hashrate and you point it to somewhere and you see numbers on a dashboard, I think that's really nice. And it is much more effective than the Mini Miner or a Bitaxe, because it's just vastly more. But it's lottery. I recall we did some calculations about even using the Mini Miner as a lottery ticket, and your chances are way better than with an actual lottery. So, you know, if you are of the gambling type, this is a pretty good way to do gambling, because it increases your chances over buying a lottery ticket.

[00:23:04] Vlad Costea: Yeah. Speaking of, I gotta give you a gift from Cake Wallet, if you have your phone.

[00:23:09] Lukas Hozda: Yes. And through the miracle of recording the previous two podcasts with Sabina here, I have already installed Cake Wallet. So I will say...

[00:23:17] Vlad Costea: Smart man.

[00:23:18] Lukas Hozda: ...but I have not initialized it yet. I don't know.

[00:23:23] Vlad Costea: So, yeah, you got to set it up. Cake Wallet is offering one Zcash to every person who joins the show. And if you don't want to hold on to that Zcash, you can swap it inside of the wallet for anything that you want, which I guess in your case is BTC.

[00:23:42] Lukas Hozda: I guess so. So should I make a Zcash wallet then?

[00:23:46] Vlad Costea: Yes.

[00:23:47] Lukas Hozda: Very well. And let's name it "Zcash."

[00:23:52] Vlad Costea: Very good naming.

[00:23:54] Lukas Hozda: Thank you. I'm a naming specialist.

[00:23:58] Vlad Costea: It's like having a baby and naming him "Baby."

[00:24:01] Lukas Hozda: Yeah. I used to make, one of the major ways that I got into programming is that I started making plugins for Terraria servers, because my computer wasn't that good at Minecraft, so I played 2D Minecraft. And I made a lot of interesting plugins, but I chose such bad names for the plugins that people who would be redistributing them would be renaming them, so the names are less confusing. And every time I would see, "Oh, well, this looks like something that I made," and it was. So, let's see. Let me save my secret phrase and figure out which word is which.

[00:24:54] Vlad Costea: I think this is still a hurdle in the onboarding process, when you tell people, "Oh, let me send you some coins," and then they install a wallet and they have to write down 12 words, and it's the most confusing part.

[00:25:05] Lukas Hozda: And I'm done with it. So what do I do now?

[00:25:09] Vlad Costea: You show me your QR code for receiving. There we go. Now I scan it. One ZEC, which is $222. You got rugged, Kristian. You swapped it immediately. It was $200. You could have more bitcoin right now, because Zcash went up. There we go. Okay, you swipe this to send the transaction.

[00:25:41] Lukas Hozda: Very well.

[00:25:42] Vlad Costea: I'm not sure if it's on camera, but...

[00:25:44] Lukas Hozda: It's quite dim.

[00:25:47] Vlad Costea: It was sent, and you should have it in your wallet within two minutes.

[00:25:51] Lukas Hozda: Very well.

[00:25:52] Vlad Costea: Thank you very much, Cake Wallet, for doing this onboarding.

[00:25:56] Lukas Hozda: Oh, man. I [unclear: love] doing podcasts where I get paid. Imagine that.

[00:26:01] Vlad Costea: Yeah, I would not ask my future guests to expect payment, because I'm not sure if there's still gonna be the sponsorship in place. I hope it will be, but I don't want to pay for my guests. That would economically and financially ruin me.

[00:26:18] Lukas Hozda: Yeah, some guests, I imagine, could get very expensive.

[00:26:22] Vlad Costea: I bet. Like, at some point I was thinking I should crowdsource an interview with Nick Szabo, because I know he charges, like, a few thousand, a few tens of thousands of dollars for a conference appearance. And I was thinking, how much could a call actually cost? Just two hours of talking over... I was gonna say Skype, but nobody uses Skype anymore.

[00:26:48] Lukas Hozda: Yeah. Well, nobody can use Skype, because Skype was shut down last year.

[00:26:55] Vlad Costea: Even while it was still [around], it was horrible. Microsoft destroyed it.

[00:26:59] Lukas Hozda: Yeah, they did, like, a redesign, and it was so bad.

[00:27:04] Vlad Costea: No, but the quality used to be great, and then somehow Microsoft...

[00:27:08] Lukas Hozda: Yeah, yeah. Yeah, it used to be great.

[00:27:10] Vlad Costea: Like, the audio. They were very good.

[00:27:12] Lukas Hozda: Back then, with my friends, we used to play over Skype, way before Discord was a thing. And now it's all that. I saved my data before it was shut down, and I think it's still possible to export it, and just forgot about Skype.

[00:27:31] Vlad Costea: I think I was using Ventrilo before Skype. I'm not sure.

[00:27:36] Lukas Hozda: Oh, yeah.

[00:27:37] Vlad Costea: Remember that?

[00:27:38] Lukas Hozda: I do remember that. And we also used to use TeamSpeak with friends.

[00:27:43] Vlad Costea: Of course.

[00:27:44] Lukas Hozda: I really like the TeamSpeak account on Twitter, because they're funny. They found the only funny German to run the account. And I keep, well, I haven't done it in a few months now, but, like, for the last two or three years, I kept sending them marriage proposals, to the TeamSpeak account. And they never quite responded the way that I would like to. But, you know, it's, what can you do?

[00:28:15] Vlad Costea: So why would anyone want to marry TeamSpeak?

[00:28:19] Lukas Hozda: Oh, my God, it's funny, of course.

[00:28:22] Vlad Costea: How does that even work? Like, you marry a brand, or the whole...? I don't know.

[00:28:29] Lukas Hozda: Well, I like to be a nuisance at times on social media. It's fun.

[00:28:39] Vlad Costea: Yeah. Anyway, the last ad that I'll plug in this episode: check out SideShift, which is S-I-D-E-S-H-I-F-T dot A-I, which is the place where you can exchange something that you receive and you don't like for something that you actually want. So if someone offers to pay you in stablecoins or whatever, you don't need to actually try to convince them to give you bitcoin. You just put your Bitcoin address into SideShift, paste... No, copy the stablecoin address from SideShift, send it to the person that wants to pay you. And then once the payment goes from the person wanting to pay you, there will be a shift, which sends you bitcoin directly to your wallet. You don't have to think about gas fees or, you know, buying a gas token just for this transaction. You just get BTC to your wallet the way God intended. Check it out.

[00:29:30] Lukas Hozda: And I think I just got the Zcash to my wallet. It says, "Received one Zcash." Unfortunately, it needs a couple more confirmations, I think.

[00:29:41] Vlad Costea: Unfortunately, SideShift no longer supports Zcash or Monero, but they used to. I guess because of regulatory issues, they don't want the headache to deal with this. But yeah, in Cake Wallet, they actually have a few swapping service providers, and I think one of them is also SideShift.

[00:29:59] Lukas Hozda: Oh, nice.

[00:30:00] Vlad Costea: So you can exchange some of the coins. Yeah, that's why it's sustainable to fund this kind of project, because they support a bunch of coins, and you can swap between them with providers, and they also take, like, a dollar or something from every swap. And if they have this at a large scale, they're able to support the operations. It's actually a good business model. I'm surprised that, I mean, Ledger and Trezor, they have this, but I don't think they make too much money from it.

[00:30:36] Lukas Hozda: Yeah. So they're like what stockbrokers do, you know. They make money on the transactions no matter which way the stocks move.

[00:30:46] Vlad Costea: Exactly.

[00:30:47] Lukas Hozda: And that's a good business model, you know. The house always wins.

[00:30:52] Vlad Costea: The house always wins. Well, at least it has such a cute name, like Cake Wallet. It has a picture of a cake. How can you be mad at it? Like, they're making money, they're taking fees, but how can you be mad that they're taking fees?

[00:31:08] Lukas Hozda: Yeah, you can have your cake and eat it, too.

[00:31:12] Vlad Costea: Yeah. And I've heard that in the US, "cake" is actually a way of calling ass.

[00:31:18] Lukas Hozda: Yes, it is. So...

[00:31:21] Vlad Costea: I'm not sure that was the intention, but, well, guys like that.

[00:31:26] Lukas Hozda: You should get, like, one of the founders or something and put them on the spot.

[00:31:32] Vlad Costea: So I did do that, and it was unintentional.

[00:31:36] Lukas Hozda: Oh, no.

[00:31:37] Vlad Costea: No, it's how language evolves, that words sometimes get a different meaning. But it was named Cake because it's a fork of Breadwallet. So it was this open-source Bitcoin wallet back in 2016 or something, and there are a lot of forks from it.

[00:31:55] Lukas Hozda: Actually, I think that there are people still using the Breadwallet, because I made this project a while ago, a couple of weeks ago, for a talk, where I made, like, a fake node, or I called it a router. And I implemented a part of the Bitcoin wire protocol, and what I did was I just connected to as many nodes as I could, and I harmlessly relayed transactions. So not too loud, but I wanted to know the network topology. And I met some Breadwallets out in the wild that were connecting to me. It was really fun. I was like, "This is still running." And the vast majority of the network was, of course, Bitcoin Core and Bitcoin Knots. I found a couple dozen Libre Relay nodes, and then different versions of libbitcoin. But yeah, it's mostly, most of the network is Bitcoin Core. Absolutely verified with my own two eyes. I have a nice Grafana dashboard.

[00:33:13] Vlad Costea: Is this the part where we transition into the Knots, BIP 110 conversation?

[00:33:19] Lukas Hozda: I mean, we can.

[00:33:20] Vlad Costea: But first, let's talk about your book, because it's right here.

[00:33:24] Lukas Hozda: Yeah. So when I was first hired into Braiins, I was hired as a Rust enthusiast, not necessarily a Bitcoin enthusiast. Although I do like Bitcoin, but it was not the reason that I was hired. And you actually asked me about how I transitioned into Braiins, and I forgot to answer. The conversation diverged a little. So when COVID started, I graduated from high school, and we were in lockdown, and I figured, well, this might be a pretty good time to get a job. And so I put "open to work" on my LinkedIn. I filled in a basic profile, and I just saw the offers from recruiters coming in, because they were looking for Rust programmers. And even today, it's kind of hard to find enough Rust programmers, because there's a high demand. And what I did is I just responded to the ad that was the closest to where I was living at the time.

[00:34:30] Vlad Costea: So you wanted to walk to work?

[00:34:32] Lukas Hozda: Oh, no. And not only did I want to walk to work: when I got into Braiins, I had them put it in my contract that I'm actually full remote, and I only come to the office when I please. I think at the time I was the person who was living the closest to the office, and I had a clause in my contract that said that I'm full remote and no one can force me to come to the office.

[00:34:56] Vlad Costea: But you're coming to the office?

[00:34:58] Lukas Hozda: I'm coming to the office now, for unspecified reasons, as my girlfriend behind the camera, you see. Right. So, yeah, imagine that: a nerd with a girlfriend. So I was a caretaker for my grandmother. That's why I wanted to be full remote, so I could stay with her and take care of her. My mother had me quite late, and my grandma had her quite late, so she was already way over 80 back then. And I wanted to take care of her during the pandemic. So they thought that I was crazy, because I was living five minutes from the office, and I specifically requested to be full remote.

[00:35:49] Vlad Costea: But they hired you.

[00:35:51] Lukas Hozda: They hired me. And that's the funny thing. It's the only IT job that I applied to, and I got the job. And so I have a 100% success rate when it comes to job interviews. And I've been here ever since. So it worked out just perfectly.

[00:36:12] Vlad Costea: Right. So you went from being a Rust enthusiast who was only coming to the office when he wanted to, to someone who writes a book about building Bitcoin in Rust and comes to the office every morning.

[00:36:26] Lukas Hozda: Not exactly every morning, but I come more often than I do not. Yeah, I am an educator, and I'm from a family of educators. And I figured out that it's a pretty good idea to teach people about Rust and to teach people about Bitcoin. And I thought, well, we can use one to do the other and vice versa. So the idea of this book is that you will learn some technicals about how Bitcoin works on, let's say, on the white paper level, and you will learn how to program in Rust by building a small reference implementation of the white paper in Rust. Of course, there are some simplifications. We don't have scripts, for example, in the clone of Bitcoin that we develop. It's just public keys. And the wallet also doesn't have the hierarchy. It's just a bunch of private keys, you know, as it was in the olden days. What's the acronym for the just-a-bunch-of-keys wallets? The oldest ones that literally were just sets of private and public keys. I think it's JBOK. Just a bunch of keys, actually.

[00:37:45] Vlad Costea: JBOK. Yeah.

[00:37:47] Lukas Hozda: Long history, long, long past. But just a bunch of keys, I think, is what these wallets were called, because that's literally what they were. They didn't have any master seed that you would be deriving...

[00:37:59] Vlad Costea: So before BIP 32.

[00:38:01] Lukas Hozda: Yeah. Yeah. So this is essentially what this book does. And it makes a simple node and a simple CPU miner, just so you have an idea about how it works and how simple it is to implement it. Like, the original white paper is not that technically complicated, and you can make a small implementation that, of course, won't be perfect, reasonably quickly. And it's a good learning project, and you learn about Bitcoin as well.

[00:38:33] Vlad Costea: So where can people find this book?

[00:38:36] Lukas Hozda: So on the Braiins shop you can purchase a paper copy of the book, and we will ship it to you. And I think it's also listed on Amazon. It's the biggest book that we've published. I think that's, we published solely. There is The Big Book of Bitcoin, which is slightly bigger, but I think this one has more text in it. And it also has, because, you know, I had a, what would you call it, a creative streak, you know, because I'm a man of many vocations, there are pictures of animals at the end of a chapter. So this is another reason to buy it. I made them in Microsoft Paint.

[00:39:24] Vlad Costea: That's impressive.

[00:39:26] Lukas Hozda: Thank you.

[00:39:27] Vlad Costea: It's impressive. They're not, you know, stick figures, which is something that I would do very well.

[00:39:33] Lukas Hozda: Well, the further you go in the book, the progressively better the pictures get. Like, the frog at the beginning, well, not my proudest work, but it works out. It's a good frog.

[00:39:46] Vlad Costea: Microsoft Paint.

[00:39:48] Lukas Hozda: Yeah. I mean, it's really funny, too. That's why I did it. But yeah, I didn't join Braiins for, like, any noble philosophical reason that some people did. It was opportunity and sheer luck, and it ended up coinciding with the things that I was really interested in, which is why I'm still here, about five years later, representing the brand.

[00:40:17] Vlad Costea: That's really nice, Lukas. But let's talk about your research, which concerns, nowadays, looking into BIP 110. You took it very seriously. You went very deep inside. You tested all the parameters of what it's going to do, whether or not it can filter transactions effectively and prevent all the spam.

[00:40:39] Lukas Hozda: Yeah. So there's one person that went deeper, and that's Kixunil, slash Habo. He...

[00:40:46] Vlad Costea: Oh, Martin Habovštiak.

[00:40:48] Lukas Hozda: Yeah. Yeah.

[00:40:49] Vlad Costea: Yeah, he was on the show, like, two or three years ago.

[00:40:51] Lukas Hozda: Yeah. So he was making some waves recently. I had a talk recently about the effectivity of filtering, or lack thereof, at the mempool level. And essentially, my argument was that it's very ineffective, that Bitcoin is a gossip network. The nodes share information with each other without any centralized direction, which is one of the reasons that we like Bitcoin as much as we do. And, you know, the nodes are constantly telling one another, "I have this, do you want this?" "No, I don't want it." "Yes, I want it." And I think the default is 120 peers for a node. And when you have a network that's so interconnected, it would have to be a vast, vast, vast overwhelming majority for information to not be able to spread among the permissive nodes. So it's not enough if nodes would have 50%, it's not enough if they have 80 or 95. It would have to be like almost every node. And we've seen something like this with the sub-one-sat-per-vbyte summer that we had, where suddenly we started mining blocks that had transactions that had fees under one sat per virtual byte. But the vast, vast, vast majority of nodes was rejecting these transactions, because the default was one sat per virtual byte. And still, these were being mined, and so many blocks were found. I think that Core changed the default recently to 0.1 sat per virtual byte, so right now we can say, "Oh, well, there might be way more nodes that are accepting these transactions," but back then there weren't. And it still wasn't...

[00:42:54] Vlad Costea: At the end of the day, all that it takes for one transaction to get mined is one node relaying it to a miner. It doesn't need to be a large majority. Even if it's 99% filtering nodes, that 1%, or even one node that's connected to a mining pool, can actually tell the mining pool, especially if the fee is high enough, to pick it up and include it into a block. And the filters don't do anything in terms of preventing something from getting added to the block. And they also don't remove stuff from blocks. We're gonna end up storing it anyway, even if you're filtering at the mempool level. And that's something that a lot of the Knots people don't understand. They feel very pure because they're running that software, because they think they're not relaying. The best argument that I heard in favor of filters was from Super Testnet, who said, well, if you don't have too much internet bandwidth and you want to limit, you can use filters to get less of the traffic that you would get if you enabled everything.

[00:43:58] Lukas Hozda: Well, yes, but it doesn't get rid of all the data, because you have to see something to figure out that you don't want it, right? So you would still have to request the transaction from somebody and figure out that it's something that you don't want. Like, what the nodes send are the inventory messages that say, "I have this thing with this hash, and if you want to know what it is, you have to request it from me." So you still have to request it, you have to read it, and figure out, "Well, this is a transaction that I don't like," and discard it and not relay it further. There is also the thing that, you know, if you want these transactions mined, you can go directly to the miners. And if you incentivize people to more and more go directly to the miners instead of broadcasting these publicly, you will obscure the mempool. So you will have something that's in the public mempool, but if a significant portion of transactions will not be broadcasted publicly, the fee market will be much harder to gauge, because you will not know that there is such and such queue of private transactions that have these and these fees. And probably, you know, since the pools would have, or the miners would have an incentive to charge extra, and they would be big transactions, they might have the highest fees. So it will obscure the pricing. So we actually want people to submit stuff to mempools publicly, so we can have an accurate picture of what is being mined and how to price the transactions that we are submitting, such that they are mined in the time that we want them to be, roughly speaking.

[00:45:53] Vlad Costea: Yeah, that's a very good argument. Sort of related question: I know that Marathon Mining, which is MARA Pool, has this service that's called Slipstream, which enables you to go directly to them. You tell them exactly what you want to get inside a Bitcoin block, and they will do it for a price. Does Braiins have anything similar, or plan to create something similar?

[00:46:17] Lukas Hozda: Well, I'm not the right person to ask, because I never worked on the pool itself. But as far as I know now, I haven't heard any plans or anything for this type of service.

[00:46:30] Vlad Costea: It should be a good business.

[00:46:32] Lukas Hozda: Yeah, yeah, it is. But, well, it's entirely possible that we just haven't gotten to it. We've been focusing on the hashpower market, mostly, when it comes to pool-related technologies. But yeah, my worry is that we would obscure the mempool and that it wouldn't help anything there. One major misunderstanding that a lot of people had is they thought, "Well, we were rejecting OP_RETURNs that were bigger than 83 bytes, and Core raised it, and so now we are allowing up to 100 kilobytes." But they're forgetting that it is the default value of a configuration option. You could have used a thousand kilobytes before Core v30, you know. And they did.

[00:47:28] Vlad Costea: And they did. There was the four-megabyte inscription.

[00:47:31] Lukas Hozda: Yes. And they did. And so part of the reasoning by Core is, well, we want to bring the defaults more into line with reality. I'm not sure which one of the Core developers wrote about a justification for increasing the default OP_RETURN setting, but the point is, it's the default setting. It's not a policy at the level of consensus or whatever. It's just a setting. And the consensus is quite lax. I recall BitMEX Research talking about it, and the history of OP_RETURNs, and they have a...

[00:48:11] Vlad Costea: Good article about it.

[00:48:12] Lukas Hozda: Yeah. And, you know, if people say, "Well, it should be like it was in the first place," but when these things appeared, the limits were actually much higher. I think they were, way, way back then, they were one megabyte per OP_RETURN, which, you know, would fill out the block. So it means effectively unlimited. But then there is the thing where you cannot prevent spam by removing the ways of embedding data, because if you have any protocol that's sufficiently complex, it will support many, many ways of steganography, of embedding data, or data hiding even. You know, you can have very simple protocols, such as, you know, written text. If you're writing text on a keyboard, you can encode a second message by capitalizing some letters, and you don't change the meaning of the original message. Or maybe you can switch between different whitespace characters, or you can use the least significant bits in an image.

[00:49:28] Lukas Hozda: And the image will still look the same, but it will now have an encoded message. And Bitcoin is quite a complex, or versatile, protocol, so there will always be ways of hiding extra, or embedding arbitrary, information. And the more ways that you close down, or the ways that you can provably close down, which sometimes you can't, you know, you cannot really tell if something is a public key. I mean, the number could look weird, and you'll say, "OK, well, maybe this is arbitrary data," but you aren't sure. And there are ways to encode arbitrary data such that they look as a random sequence of bytes. So you cannot say for sure. And then the question is, well, will we keep closing things down so much that we end up crippling the base protocol, by slowly cutting off things that are niche but legitimate uses of the protocol? Legitimate even in the eyes of the proponents. That is...

[00:50:51] Vlad Costea: You mean, like, multisigs, or... Yeah, yeah, yeah. Some types of CoinJoins, or...

[00:50:57] Lukas Hozda: Yeah. Like, all of these can be similar to embedding arbitrary data. And, you know, you can publish arbitrary data in a way that's encrypted, so it's not obvious that it's arbitrary data, because it can be in the public keys, which is incredibly harmful, because that goes into the UTXO set very often. And then you can just say, "Well, this is how you decrypt it," and suddenly, "Hey, look, turns out I embedded a lot of arbitrary data over a period of time, and you guys did not know about it." So it's very difficult to do this. And I think if people do want to figure out a solution, it has to be on pricing out these types of things, you know.

[00:51:44] Vlad Costea: That's something that we don't talk about enough. Bitcoin does have a spam filter, and it's called a transaction fee.

[00:51:51] Lukas Hozda: Mm-hmm.

[00:51:52] Vlad Costea: And it's also called the block size limit. You can't just put as much data as you want, because there is a limit, and you have to pay a lot of money if you want to. And people complaining that there are not enough financial transactions obviously are not transacting themselves. They're not paying for the block space. But it's like this, and I don't want to offend anyone, but it's like this communist situation where people are concerned about something that they're not paying for. They feel like they own it, and it's theirs, and they should be in control of it, but at the same time, they're not willing to pay for what it is.

[00:52:34] Lukas Hozda: Yeah. And some of the concerns are vastly overblown. So people are saying, "Well, look, we don't want to blow up the UTXO set, and we don't want the blockchain to grow too quickly, because it will make it harder to run a node." And I've seen some posts by proponents which were like, "Well, if you don't stop the OP_RETURNs, the blockchain could grow by terabytes a year." Like, man, no. We have the block size. You can calculate the upper bound of how much the blockchain will grow very easily. Just find out how many ten minutes there are in a year and multiply by four or so megabytes, and, you know, that's the most amount of bytes that could be stored in a year on the blockchain, thereabouts.

[00:53:34] Vlad Costea: Yeah, this was assumed in the moment when we activated SegWit and we quadrupled the block size. Actually, we added more block weight, which is a different unit, but that's beside the point. It was assumed that the blocks would become full. And the math that they did at the time was that, OK, this is a good trade-off. We get more throughput, and it also keeps the network fairly decentralized. And I guess it would not be terrible if, and once again, unpopular opinion, if we went for 8, 16, 32 megabytes. It would not be so destructive. If anything, I think it could help a lot of people onboard layer twos for a lower price. As it seems like, there was a meme, which I shared on Twitter, where if you're using somebody else's node, like Uncle Jim's node, that's horrible. You shouldn't. You should run that on your own computer. But if you're hosting your coins with Uncle Jim, and Uncle Jim is the custodian, that's how we scale Bitcoin.

[00:54:47] Lukas Hozda: Yeah. God damn it. The second concern is with the UTXO set, where it's a bit harder to calculate how much it can grow in size for a period of time. But people consider it to be a much more pressing issue, because they say, "Well, it has to be stored in the RAM." But it doesn't have to be stored in the RAM, and it's not even being stored in the RAM. Bitcoin Core uses LevelDB, which stores some things in the RAM, but most of it is on the disk.

[00:55:24] Lukas Hozda: And even if you don't consider how it's actually implemented, you can consider simple heuristics. Well, do I need to store in RAM, with very quick availability, the UTXOs that haven't been touched for ten years? Probably not, you know. The older the UTXO is, the lower the likelihood that it will be touched at any given moment in time. The ones that are most likely to be spent are the newest ones. And maybe if you go historically and you chart out the average age of a UTXO when it was spent, maybe you will find out that there is, like, a bump at some point, but it's still, the more recent the UTXO, the more likely it is to be spent soon, because only a few very old UTXOs survive. And you can just say, "OK, well, we will store the hot ones in RAM, and we will store the cold ones on the hard drive." And there have been proposals to make this even easier, with Utreexo, if you know, where you just accumulate hashes and you cryptographically verify that the UTXO is in the UTXO set. And the amount of memory that you would need to keep this around goes down to kilobytes. It's insane how small it gets. Of course, there is some added burden on the network and on the people submitting transactions, that you have to provide the proof that it's in there. But if we decided that, well, the UTXO set is too big, and we opted for this, it would effectively fix the scaling problem, like, forever.

[00:57:19] Vlad Costea: I see some of this with ZK verifiers, like StarkWare, where they even say that you can sync a full node within seconds, and with a very small amount of data, which only represents the proof that there is validity in the blockchain files. You don't have to store everything. You just store that proof, which is verified with very sophisticated mathematics. And that's something that right now is being used in Ethereum for scaling. They want to use that for mobile nodes, for making sure that you rely much less on RPCs, which are centralizing and are not very reliable. And instead of these service providers, we're going to have cryptography and math, which I think is beautiful. That's sort of the point. That's why, I guess, we are here: to replace some of the financial services and banking and whatever it is that we don't like about the legacy financial system with smart math. But the Knots people don't seem to like it. I don't know. So I think yesterday, over beer, you made a very compelling case about why they are well-intended and they are pure at heart, and they're ideologically consistent in some way, but at the same time, they're wrong, because they don't understand how Bitcoin works.

[00:58:41] Lukas Hozda: Yeah. Most of the argumentation is value-based, or opinion-based. And, you know, the arguments that "we want Bitcoin to be primarily money," "we want to get rid of spam," it's things that at face value are admirable and morally good, so to speak. And when you appeal to emotion like that, it's also very attractive to many people. But the issue is that, let's say, the technology, or the theory, does not mesh with simplistic ways of solving the problem. It's like, yes, we can filter some things out, but you will never filter everything. And the people who want to embed stuff always have an advantage, because they are the people inventing the new ways, and you are going behind them, patching the old ways. And you also have to be careful that when you patch something, you don't break legitimate uses. So it's inherently more difficult to patch ways than it is to create new ways, for a technology like Bitcoin. And this is something that keeps happening all the time with different technologies. Even, you know, things like Minecraft: "Hey, there is this bug report, 15 years on the Minecraft bug tracker, something about redstone. Why is it not fixed?" Well, it's because if you fix it, you break other things. So we decided we will not fix it, in the name of supporting other things. "And we know this is broken, we know this is a problem, but we decided that the trade-off is worth the issue." And I would compare it to anti-cheat software. Like, you can make more and more strict anti-cheat software for a game, and the stricter it is, the more likely that you will have a false alarm and you will ban a legitimate user, a player.

[01:01:01] Vlad Costea: And you're gonna make it less fun for everyone.

[01:01:04] Lukas Hozda: Yes. And you will break the game on Linux, because "we don't support kernel-level anti-cheats for Windows on Linux," and I cannot play that game anymore.

[01:01:15] Vlad Costea: I'm not sure if that's applicable to Bitcoin.

[01:01:17] Lukas Hozda: No, it's not. It's a pet peeve of mine.

[01:01:20] Vlad Costea: In 2023, when the Ordinal inscriptions came around, or maybe it was '22. I think it was '22.

[01:01:25] Vlad Costea: The conversation on the bitcoin-dev mailing list was exactly this. Like, Andrew Poelstra, head of research at Blockstream, said, well, this is interesting, could be harmful, but there's nothing we can do about it without further harming the network. Like, we can potentially filter out this way of using Bitcoin, but it's just going to be a game of whack-a-mole. You know what whack-a-mole is?

[01:01:54] Lukas Hozda: Oh, yeah, I know.

[01:01:55] Vlad Costea: You hit the mole in the head with your hammer, but there's always another one that's coming out. So it's a never-ending struggle, and it's most likely not worth it. But whack-a-mole at least is a degree of fun. And you can earn some tickets that you spend to get some stupid items.

[01:02:17] Lukas Hozda: Yeah. A lot of people...

[01:02:20] Vlad Costea: My point with this is that this was known for, like, three years or longer. And yet you still have people who are very misguided about the whole issue. And they have false hopes, because Luke submitted the CVE report, and then he started the company, he started this mining pool, which promised to filter out these transactions, while also promising to miners that they're going to be making more money from fees, which I don't know how it works, because basically they're leaving money on the table.

[01:02:54] Lukas Hozda: Yeah. There is this issue where, you know, I had this discussion when I had my talk about filters, where someone said, well, the miners, or the pools, don't they have, like, a service role? Shouldn't they just always bow down to what... Oh, yeah, thank you. To what the community wants? And I said, well, actually not. What they are primarily bowing down to are economic incentives, because actually the pools are serving the miners. The pools have to reflect the wishes of the miners, because if they do not reflect the wishes of the miners, which is primarily maximizing profit, the miners will go to a pool that does satisfy these wishes, by and large. Of course, every agent in a market has justification, and while it is by and large rational, it is not always rational, you know. So I can mine into a black hole. Why not? But it's my decision, and I had a reason for it. But most likely I will not be mining into a black hole. And the pool has to reflect the wishes of the miners, and miners are under a significant economic pressure that most people don't realize. If you look at, like, the hashprice and so on, the margins for miners are very thin. If you don't have free electricity, the margins are very thin. And you have to factor in all sorts of other costs. You know, it's not just ASICs running in a vacuum. And so they must maximize their profits. And if they do not maximize their profits, they go out of business. And they cannot just bow down to philosophical whims if they are hurting the economic incentives of miners. And we keep seeing this, like, all the time, and we've seen it with the OP_RETURN defaults. Like, "Yeah, the miners want blocks to be as full as they can be, so they can get more fees." No. The miners, by and large, don't care about inscriptions, because it's, you pay fees.

[01:05:30] Vlad Costea: They don't care that they are inscriptions, because they're transactions that are paying fees. Yeah, yeah. That's what you mean.

[01:05:36] Lukas Hozda: Yes.

[01:05:37] Vlad Costea: Because I remember during the last halving, which was April or May of 2024, the blocks were super full, and there were a few consecutive blocks when the transaction fee was higher than the blocks.

[01:05:50] Lukas Hozda: Yeah, it was insane.

[01:05:51] Vlad Costea: And it actually gave us hope that there is enough demand for Bitcoin blocks, for block space, for this experiment to actually work. Because if you read the white paper, that is Satoshi's assumption: that as the subsidy diminishes for miners, there's going to be enough transactions to pay fees to make up for it. I think it's in the incentive chapter of the white paper. And that is still not proven.

[01:06:19] Lukas Hozda: It's not.

[01:06:20] Vlad Costea: It's still not proven. I'm not sure if it will happen, because you always have, like, a cheaper network. If you just want to send $20 or something and pay one cent, you can use Litecoin or Tron or whatever, Solana. And maybe you're not going to care that it's not the same amount of security, because all you care about is just moving the money from point A to point B. So Bitcoin does have a lot of competition. These L2s that we have today are non-mined, so the transactions that happen on the L2 don't pay anything to the miners. I think the only exception is Rootstock, RSK.

[01:06:57] Lukas Hozda: We used to merge mine Rootstock. Maybe we still do, I'm not sure.

[01:07:02] Vlad Costea: It's the only one that gives something to the miners, but it's political, in the sense that they choose, with the fees that they collect, to give them back to the miners. It's not automatic, so they can just change that anytime that they want. And it's gonna be an issue in maybe 8, 12, 16 years. I don't know exactly how long this is gonna take. So far, we were safe from any crisis because the price doubled every four years. More than doubled. And everyone was like, "Ah, it doesn't matter. Fees are not important. The subsidy..." Just, you know, it's funny, because the security model of Bitcoin is just "price go up."

[01:07:45] Vlad Costea: Not just the price, but fiat money. For as long as fiat money works as fiat money and it inflates, seems like the security of Bitcoin has been safe. But is it always going to be like this? Another concern that I have is that this past bull market, we haven't seen much of a user influx. We've seen institutions coming in, but we don't see many new users. And that's a problem, because I think a lot of people are still remembering the incidents with FTX and BlockFi and the stuff that happened in 2021. To them, crypto as a larger industry is sort of scammy, and they don't want to touch it. I think our only hope at this point is that we have the Gen Z and Gen Alpha kids who are trading Fortnite skins, and they see value in digital assets, and they're going to decide that Bitcoin has value, and they're going to buy it. And I think also the other hopium that I see nowadays is that AI agents will choose to send Bitcoin to each other.

[01:08:53] Lukas Hozda: Yeah. We'll see about AI agents. It's such a new phenomenon. And I am actually, so if you read a bit on my Twitter, I am definitely not an AI enthusiast, by and large. Oh, oh, shit, that's me.

[01:09:14] Vlad Costea: It is you. You have a Superman avatar, and your last name is Hozda.

[01:09:18] Lukas Hozda: Yeah, that's how you pronounce it.

[01:09:19] Vlad Costea: It's like a mix between Honda and Mazda.

[01:09:22] Lukas Hozda: Well, yes.

[01:09:23] Vlad Costea: If Honda and Mazda had a baby.

[01:09:26] Lukas Hozda: Yes. And it means, it's from hvozd. So originally the surname was Hvozda, but the V got lost more than a couple of generations back. So essentially my last name means "forest guy," so a guy related to the forest. Hvozd is like a primordial forest, like an old, deep, dark forest. That's what you...

[01:09:53] Vlad Costea: What kind of car do you drive?

[01:09:56] Lukas Hozda: A Škoda.

[01:09:57] Vlad Costea: Škoda. OK, you're a nationalist.

[01:10:00] Lukas Hozda: Yes. Which is the funny thing about Škoda: the word translates to "damages" or "pity." You know, so you say, "Oh, I guess that's not the best name for a car." You know, "I am driving the damages car."

[01:10:18] Vlad Costea: It's like the Volkswagen that has, like, a cheaper brand logo. Same engine, same everything, but different brand. Such a shame. It's not that bad. But to be honest, I don't like Volkswagen for the past 15 years.

[01:10:36] Lukas Hozda: I mean, they've been lying about the emissions tests. A couple years back, they had this scandal.

[01:10:42] Vlad Costea: No, it wasn't a couple. That was 2016, I think.

[01:10:47] Lukas Hozda: I have a different perception of time to most people.

[01:10:51] Vlad Costea: That was a decade ago. But yeah, they lied about that, they lost customers, and their products are objectively worse right now. I see that Europe, which used to be very French and German in terms of car choices, is becoming more Japanese and Chinese.

[01:11:12] Lukas Hozda: Well, it sure is changing.

[01:11:15] Vlad Costea: I do like the Japanese part, the Chinese not so much.

[01:11:18] Lukas Hozda: Yeah, exactly. It's less... I probably want my car to be well-made, and as anti-regulatory as I am, I maybe think that it's a good idea if a car follows European safety regulations to a T. I mean, they have to, the Chinese ones, when they import them, but I think if it's made from first principles with that in mind, I trust it a little more.

[01:11:51] Vlad Costea: Fair. Let's go back to BIP 444, slash 110.

[01:11:57] Lukas Hozda: Yeah, it's a very socially and technically interesting BIP. So it's weird in a couple of ways. The first way that it's weird is that it's being painted as temporary, and that it's a stopgap measure that responds to an urgent need. And this is all language that's in the BIP. And actually, in the first version of the BIP, there was even more of this language, but turns out people don't like it when BIPs threaten legal action, and so it has been toned down significantly. And this is not something that necessarily happens that often. When you propose a change, you generally intend for the change to be permanent, and not like, "Hey, there are these changes. They will start working here, and they will stop working there." And to me, it's an attempt to create a sense of urgency while also reducing, in some degree, the feeling of importance, to make it [unclear: as sound] as it should be, because, well, it's meant to be a temporary measure, so, you know, who cares, it will be gone in a year anyway. But to me, it's a way to get a foot in the door for the proponents. Like, "Hey, we forced this in, we'll force in other things." And I don't like that, because I think that the changes that we make to Bitcoin should be as rigid as they... So not as rigid, as rigorous as they can be. So we have to be really careful about what we're doing, and we have to have the maximum amount of productive debate about these changes that we would propose. And I think that BIP 444, now BIP 110, is trying to avoid that by trying to project itself as an urgent stopgap measure, because anyone could upload child pornography to the blockchain tomorrow. And I don't think that it will necessarily happen.

[01:14:31] Vlad Costea: I think they call it CSAM, or whatever.

[01:14:35] Lukas Hozda: Yeah, I think CSAM. Yeah, I think it's, like, child sexual adult material, or something. That's what the acronym stands for. And there are some interesting arguments. Like, I forgot what it was, but they are limiting... Oh, shit. They're limiting...

[01:14:58] Vlad Costea: You're making noise into the microphone.

[01:15:00] Lukas Hozda: Ah, god damn. They are limiting one thing to 256 bytes. I forgot what it was. Let me quickly pull it up. And the reason why they are doing it: yeah, "Why limit other data to 256 or 257 bytes?" They say, "With modern compression, it is plausible to represent images in as few as 300 to 400 bytes." And I think that's quite arbitrary. You know, I could make a small image that's 256 bytes. It will not be a good image, but, you know, you can do it.

[01:15:37] Vlad Costea: Of course. I mean, video game developers were doing it on MS-DOS games, and they could be very graphic.

[01:15:44] Lukas Hozda: They could be very graphic. Like, you could make an image grayscale, and suddenly you can fit quite a few pixels, you know, because you could have less than a byte per pixel, even. And it can be quite graphic. But there is this data contiguity argument, where we are essentially saying, well, the problem is if the data is contiguous, but if the data is not contiguous and you have to decode it somehow, then it's fine, because in the eyes of Bitcoin, it's just data. And this is an argument that doesn't make much sense to me as a programmer, because you have to decode all data, you know. You even have to decode plain text, because it could be using a different encoding. You know, it could be UTF-8, it could be UTF-16, it could be one of the cursed Windows encoding formats that existed for decades. And all data has to be interpreted before it is shown to the user. And to me, whether it is done as an extra step at the level of the protocol, or if it is done one step later with an image decoder, does not seem that relevant. And you always have to do some decoding, because, you know, if you dump out raw transaction data, you probably need to pull the actual data out of it one way or another. And so there will always be a decoding step, you know. So whether data is contiguous, or if there is an opcode every couple of bytes, who cares? And Habo made a calculator a few days ago that calculated how much more expensive the ways of embedding data that are not prevented by BIP 110 would be, and turns out, not by much. So the cost increases marginally, but what happens is that the more harmful ways of embedding data are now the more economically viable ones. And that's bad, because that's what will end up bloating the UTXO set more. And, you know, I said that I don't think that that's that much of a problem, but the point here is that it goes against the stated goal of doing something like this. "We're doing this so we prevent data bloat." But if we incentivize OP_RETURN, which doesn't go into the UTXO set, that's the thing that actually prevents bloat.

[01:18:41] Vlad Costea: I think that there are a lot of ideas inside of Bitcoin which are very bad, but we cannot know for certain that they're bad, because the market never assigned any price or value to them. So if there's going to be a fork where the Knots people activate BIP 110, and they prove how it works, they will see how the market reacts to it. And most likely it's going to be very bad. And that's going to be like a cold shower, a wake-up call for them to understand that the only reason why they had any value to their opinions was that they were part of this larger group, which encapsulates a lot of different opinions. And maybe a few of them have value, and they capture most of the market activity, and are the reason why the price is so high. But once you fork away, you're going to find yourself in a place where that's all you got. You only have a few proponents, you have a few advocates, and that's it. You're not going to scale up from there.

[01:19:45] Lukas Hozda: Yeah. And the miners will follow the most profit, because they have to maximize profit. And if miners leave from the permissive chain, that makes it more profitable for the miners that did not leave the permissive chain. So, you know...

[01:20:02] Vlad Costea: Mining is something that follows profits.

[01:20:04] Lukas Hozda: Yeah. First and foremost. And there is also some historical precedent with SegWit2x, if you know about it. It died very quickly, but...

[01:20:19] Vlad Costea: It wasn't quickly, but it was a few months.

[01:20:22] Lukas Hozda: Yeah, it was a few months. That's quickly for me. Yeah.

[01:20:31] Vlad Costea: But they gave up on it just because there was already Bitcoin Cash as a fork, and they did not want to further fragment the network. And they said, "OK, there's a big-block version of Bitcoin. Let's just get to that if we want big blocks, or just support Bitcoin if we believe that Lightning is going to be successful."

[01:20:49] Lukas Hozda: Yeah. And there was also, I think, I'm not sure if it was Binance or Coinbase or one of those platforms, tracked both of the chains separately. And soon after the split, SegWit2x started tanking. So it became less and less economically viable, because people were abandoning it. So it died, well, in my eyes, it died fairly quickly in the grand scheme of Bitcoin, on economic grounds. And that's a risk, that something like that may happen to the Knots people if they split off. Part of it is social, and we spoke about it yesterday as well. The thing that will end up being called Bitcoin after a split has the advantage, because that's the Bitcoin, and the other thing is the other thing. And you either use the thing that everybody is talking about, and I mean outside the community, or you use the other thing, and you have to have a stronger justification to use the other thing than the thing that everyone knows. And it's also more difficult to have profits on a chain that's more restrictive. And at the very beginning, there is also an advantage for the permissive chain, because the blocks that are valid for the restrictive chain are also valid for the permissive chain, but not the other way around. So if you're going by the longest chain, then the permissive one has a starting advantage, because the permissive chain can always take the blocks from the restrictive chain at the beginning, but not the other way around.

[01:22:46] Vlad Costea: For a long time, we have been saying that Bitcoin is permissionless, and that nobody cares about what's happening in Bitcoin transactions, which makes it censorship-resistant. That's another quality of it. Of course, this is all socially enforced. If we decide that we should filter out transactions, then this censorship-resistant quality can diminish. And if we decide that we can hard fork... I know that there was this proposal to return a large amount of Bitcoin, I don't recall exactly how much, by Mark Karpelès, from Mt. Gox. And basically, he said he can prove with a court order that the coins are stolen and they belong to Mt. Gox, and he wants to return them to the rightful owners, who are clients of Mt. Gox. And if the Bitcoin community is willing to cave and basically hard fork for the purpose of taking away these coins to give them to Mt. Gox... The amount is 79,956, I just checked. It's a lot of Bitcoin, chunk of change, but it's also a slippery slope. And I feel like all of these core tenets of Bitcoin that we have been taking for granted, that it's censorship-resistant and unconfiscatable, and also that it's neutral, like nobody cares how you're using it, Bitcoin doesn't care, all of these are being tested at the same time from different fronts. And we have to really resist and prove that this technology is really neutral, and a tool that anyone can use, even if we don't like the way in which they use it. It doesn't mean, if you defend the principle of, for example, universal freedom or something, it doesn't mean that you're also advocating for crime. You're understanding that crime is one aspect of it. You can deal with crime separately, but you should not limit everyone's freedom just because you want to marginalize the criminals.

[01:24:52] Lukas Hozda: And this is also, you know, quite communist, as you said. It's, you know, why do you care what other people are doing with it, you know?

[01:25:02] Vlad Costea: Well, they will say, "Yeah, it gets stored in my node." Prune your node, OK? Just store, like, five gigabytes or something, and you're not gonna have to deal with any of these problems. You don't have to buy new storage. There is no reason for you to store the entire history of Bitcoin transactions. There are mining pools that are doing that, there are exchanges that are doing that, there are other people who can afford more powerful computers that are doing that.

[01:25:28] Vlad Costea: You're not really making Bitcoin more decentralized with your Raspberry Pi. The point of the Raspberry Pi should be that you validate your own transactions, and that's it. You confirm that you received the right amount of coins, that you're at the right block heights, and whatever.

[01:25:48] Lukas Hozda: Yeah. And it sets a bad precedent, that, you know, as you mentioned with the slippery slope, it sets a bad precedent that we are willing to start rejecting some use cases of Bitcoin. We might say, "Well, we did it once, why not do it again?" And there has been a lot of, like, legal arguments with the child pornography and so on. And I have two issues with that in particular. The first issue is, well, it's not like you are storing... So services that host user content are generally subject to safe harbor provisions. So you are not a priori responsible for the content that users store in your service. Now, if you are running a business, you may get DMCA requests and similar, to remove stuff that breaks someone else's copyright and so on. And, you know, if you're hosting a file hosting service, well, you delete the files. But of course, that's something that's not necessarily possible with Bitcoin.

[01:27:00] Vlad Costea: First of all, they're not even files. They're not even files.

[01:27:04] Lukas Hozda: Yes.

[01:27:05] Vlad Costea: They're not JPEGs. Even if you can decode that sequence of letters and numbers and turn that into a JPEG, it's not a JPEG. You're not storing a JPEG.

[01:27:17] Lukas Hozda: Yeah, exactly. And I don't see why we shouldn't instead lobby for safe harbor provisions for Bitcoin, you know, so that we say, "Well, you will not be responsible." And the second thing is, we have been positioning ourselves as a bit countercultural, especially in the beginning. And, you know, we're the cypherpunks. We care for privacy. We don't like the state, and so on. Why should we suddenly care about what some politician says, you know?

[01:27:50] Vlad Costea: Oh, in some parts of the world, maybe it will become legally problematic to run a node because of the contents that you have to download. That's what the argument is.

[01:28:02] Lukas Hozda: I'm hosting a seedbox for torrenting files. I don't care. And you cannot prove that I'm hosting it, by the way. But it's, you know, we don't care about the state, and we don't care about the legal thing, because we do not believe that...

[01:28:22] Vlad Costea: Now we care. Now, because we deal with institutions.

[01:28:27] Lukas Hozda: Yes. So it's an institutional pressure. But I think that we are aligned. Yeah, we sold out. And I think that we are assigning too much weight to it.

[01:28:38] Vlad Costea: For better or for worse. We wanted mainstream adoption, but we got the worst kind of mainstream adoption. Instead of getting all the normies trying Bitcoin, we got banks...

[01:28:54] Lukas Hozda: And institutions, that we...

[01:28:57] Vlad Costea: We used to pretend to hate them. They're our friends. You can sponsor my podcast, banks and institutions. Like, I'm not gonna complain about it. I'm not gonna praise you either. You can give me money, I don't care. Lukas, you are @LukasHozda, which is exactly your full name, on Twitter.

[01:29:20] Lukas Hozda: Yes.

[01:29:21] Vlad Costea: Kristian was trying to pressure us into doing a giveaway.

[01:29:25] Lukas Hozda: Yes.

[01:29:26] Vlad Costea: So, is there anything to give away to the listeners?

[01:29:29] Lukas Hozda: Well, of course, there is my book.

[01:29:31] Vlad Costea: Your book. And I'm really curious, how do we do the giveaway?

[01:29:35] Lukas Hozda: Well...

[01:29:36] Vlad Costea: I was thinking a good way to do it, to not intersect with the people who already registered to the newsletter, would be to ask people watching this on YouTube to write something in the comments. Like a special word, or answer a question. And then we just choose someone from there, and we send them the book. Or is it gonna be like a voucher, so they can order this from...

[01:30:10] Lukas Hozda: Oh, we'll send them the book, I think. That's better.

[01:30:15] Vlad Costea: Can you sign it?

[01:30:16] Lukas Hozda: I can sign the copy.

[01:30:18] Vlad Costea: Yes, you can sign a copy. You should sign the copy.

[01:30:21] Lukas Hozda: Yeah. And if you're, maybe if you're going to be posting it on Twitter as well, the people can follow me.

[01:30:27] Vlad Costea: Yes, you want the followers. Let's do it like that. Yes, I can post about this on Twitter. You want to give away one book? What's the challenge?

[01:30:37] Lukas Hozda: What's the challenge? Well...

[01:30:39] Vlad Costea: What should people do? Because it shouldn't just be, like, "Like, retweet and follow." It should add an extra step, like write a special comment, or do a little side quest.

[01:30:52] Lukas Hozda: Well, let's see. Let's see. So I wanted to do programming challenges back then, but I think that's not too broadly applicable. So let them post their favorite quote from the podcast, that either I or you made, and I will pick out the best one.

[01:31:17] Vlad Costea: Oh, God. I can already imagine what they're going to pick.

[01:31:22] Lukas Hozda: Hell yeah. And I'm going to retweet it if I like it really well.

[01:31:28] Vlad Costea: OK, that's a good approach. In closing, I also want to add that none of this is new. Like, the conversation that we're having about putting files in Bitcoin and whether or not it's spam, it goes all the way back. The same conversation was being had in 2010 or something. Satoshi was part of it. At some point, someone decided to put WikiLeaks files into Bitcoin, and those were highly classified. They could get a lot of people into trouble. But you're storing them. The same guy who's trying to push filters right now put Bible verses into the Bitcoin blockchain.

[01:32:06] Lukas Hozda: And more interesting words than Bible verses. Like, I think that there is one block that has the message "I am eating a juicy burger" embedded in it. You know, maybe there is some, at least spiritual, weight to the Bible verses, but, well, maybe it was such a good burger that it just needs to be immortalized for future generations.

[01:32:30] Vlad Costea: I'm pretty sure there are no burgers in the Bible. Just fish and sheep.

[01:32:37] Lukas Hozda: And manna in the Old Testament, which no one knows what it was.

[01:32:42] Vlad Costea: Bread?

[01:32:43] Lukas Hozda: No, the manna came from the ground in the desert. So maybe it was some plant, but God put it there.

[01:32:53] Vlad Costea: So you can use it to cast spells.

[01:32:56] Lukas Hozda: Yeah, you can use it to cast spells. Yeah. We'll see how it pans out. But there are a lot of topics that are cyclical in Bitcoin, and we remember them every few years. Well, we start... No, no, no, we actually forget them. We forget them, and so we start talking about them all anew, and at one point, people start digging up the old arguments.

[01:33:22] Vlad Costea: So what do you think is going to come back next?

[01:33:25] Lukas Hozda: Well, the one thing that's coming back is quantum computing. Like, we've been talking about quantum computing a few years ago. I recall we even posted an article, I think it was like 2019, we even posted an article on our blog about quantum computing, and the threat it poses, and how realistic and how close by is the threat. And then it kind of died down, and now Hunter Beast is working on BIP 360, and it's once again a very relevant topic for discussion. But I think that it will die down again, because it will probably take a few decades for quantum computers to be really useful. So these are topics that are cyclical. Maybe we will come back to the block size topic again, because that also popped up a few times. I can't wait.

[01:34:26] Vlad Costea: Yeah. Oh, boy.

[01:34:29] Lukas Hozda: So maybe we will be discussing whether the 21 million cap is something that we want to keep around, and probably we'll come back to the same conclusion, and that's yes, we do want to keep it around. But some of the longest-standing decisions in Bitcoin benefit from being tested every once in a while, I think, because we need to remind ourselves why we did the things that we did, and why we made the decisions that we have made. And I don't think there is any other way. You have to discuss it all anew and come to the same conclusion. Or maybe you will not come to the same conclusion, but most likely you will. And that will reinforce the validity of the decision. And I don't actually mind it at all that we are discussing with such fervor. I do mind that a lot of the discussion is quite toxic. Like, Kristian posted about Habo's project, where he embedded a TIFF of Luke onto the blockchain using MARA Slipstream. And Kristian was being informative. He didn't really make, like, an attack or anything, but people started attacking him. They called him ugly. Like, why? Why? You know. And I don't think it necessitates this much emotion, but it's part of the thing that...

[01:36:10] Vlad Costea: I don't think he's ugly.

[01:36:12] Lukas Hozda: I don't think he's ugly. [unclear: No homo], but, well, he's my boss right now, so I don't think he's ugly. But I think it's because the argumentation is value- and emotion-based, and that's why the supporters can get quite emotional, and then call Kristian ugly, because they got angry, you know. It's one thing if you debate a technical fact, and the other thing is if someone, well, in your eyes, attacks something that you believe in on an emotional level. You also take it as an attack upon yourself.

[01:36:57] Vlad Costea: Like the whole "Andy Back" meme.

[01:37:00] Lukas Hozda: Oh, I don't know the "Andy Back" meme.

[01:37:02] Vlad Costea: "Andy Back." No? You haven't read the Epstein files?

[01:37:08] Lukas Hozda: I have seen some things from the Epstein files, mostly about eating children, but...

[01:37:13] Vlad Costea: No, no, no, no, no. It doesn't matter. It's just, Blockstream, when they were trying to raise funding, they had their first CEO, Austin Hill, go to the island, and Epstein was calling Adam Back "Andy Back" in the emails. And people started guessing, like, "Huh, why?"

[01:37:36] Lukas Hozda: Yeah, I think that Epstein was dyslexic. Like, if you read the emails, a lot of them have such egregious typos and mistakes.

[01:37:47] Vlad Costea: I think he just didn't care. He was on his island, just typing on his BlackBerry or whatever.

[01:37:52] Lukas Hozda: And possibly on drugs.

[01:37:54] Vlad Costea: It was like this, and... Yeah. And he had an assistant to deal with appointments and whatever.

[01:38:02] Lukas Hozda: Oh, well. What a note to end on: the Epstein files. How did we get here?

[01:38:09] Vlad Costea: Oh, man. The world is messed up.

[01:38:12] Lukas Hozda: Well, yes.

[01:38:13] Vlad Costea: Let's not mess up Bitcoin. Thank you very much, Lukas.

[01:38:17] Lukas Hozda: Thank you for having me.

[01:38:19] Vlad Costea: Follow him on Twitter. We'll do a giveaway of his book. Cheers.