WBD647 Audio Transcription
Scaling Bitcoin Privacy with Calle
Release date: Wednesday 19th April
Note: the following is a transcription of my interview with Calle. I have reviewed the transcription but if you find any mistakes, please feel free to email me. You can listen to the original recording here.
Calle is a Bitcoin and Lightning developer contributing to LNBits and the Cashu ecash system. In this interview, we discuss Cashu’s mission and development, undertake a live demonstration of it in action, the importance of privacy, removing ideology from Bitcoin, and the future of AI and robots.
“This was the reason why I think ecash is coming back now: because it needed Bitcoin. Ecash is almost worth nothing if you don’t have a free sovereign monetary layer below it… and we found the base money of the internet.”
— Calle
Interview Transcription
Peter McCormack: Good morning, Calle.
Calle: Good morning, Peter.
Peter McCormack: Fucking hell! You've just given me away!
Calle: You just gave me away as well.
Peter McCormack: But Calle's a secret name.
Calle: Peter's a secret name as well.
Peter McCormack: No, that's my real name.
Calle: You have the larger anonymity set than I do.
Peter McCormack: Fuck's sake! I was going to be in disguise!
Calle: My name is Ben, actually my name is Ben.
Peter McCormack: Fuck, that shit's too hot; I don't know how you and Gigi do this, oh man! How are you, brother?
Calle: I'm doing really, really good, thank you. Thank you for having me; I'm a big fan of your show, so it's great to be here.
Peter McCormack: Well, thank you for coming to Bedford.
Calle: Cool.
Peter McCormack: You got to be the first bitcoiner who came to my bar.
Calle: Yes, I was at your bar and I wish you a lot of success with it; I think it's going to be a success story.
Peter McCormack: I hope so. And you're coming to the football today as well, I forget even about that; you're basically going full Bedford.
Calle: Full programme. I saw most of Bedford today, I'm going to see the football match today; you're going to win, right?
Peter McCormack: Well, so yes, we're going to win, but I don't know if we'll win the league today, but if we win, we've kind of won the league; it's very close.
Calle: Step by step.
Peter McCormack: So listen, Ben Arc is a huge fan of yours; we're big fans of Ben. Ben's been on the show how many times now; Two, three times?
Danny Knowles: Twice, I think.
Peter McCormack: Yeah, it's coming down to Bedford for our event next week. He's like, "Dude, you've got to get Calle on the show".
Calle: That is so cool. I'm also a big fan of Ben myself; he's basically my mentor in a sense.
Peter McCormack: He's such as nice guy as well, he's like a nice bitcoiner.
Calle: Yeah, absolutely. So, I got involved with Ben because I started contributing to LNbits a lot, the project that he's working on, and at some point he just asked me to join the team basically, and we started working together. At the beginning, he held my hands, showed me a little bit how to get involved and probably the reason why I'm sitting here today, so I'm very, very grateful; he's an amazing person.
Peter McCormack: I agree, I think he's an amazing person. Oh God, I sound so deep-voiced today.
Danny Knowles: So hungover.
Peter McCormack: I sound like Matthew Mežinskis.
Calle: I'm trying to go down as well.
Peter McCormack: Can you hear that bass? So, we also, yesterday, we had Obi in with us from Fedimint and we got you out together last night.
Calle: What a coincidence.
Peter McCormack: No fights!
Calle: No, no fights at all; we had an amazing conversation. Actually, I'm very friendly with the guys from Fedimint, especially Eric, known him before I met Obi as well. Eric is actually the person who eCash-pilled me, so also my thanks goes to him; he earned a lot respect in the community for that. So, he taught me what eCash is and that's the reason why I started looking into it and made this alternative project, Cashu, which together with Fedimint, are I think the hottest eCash projects out there. I'm very happy that these projects are on Bitcoin and not on any kind of system.
Peter McCormack: Well, let's get into that; there are so many things we could talk about with you today. I know Danny talked to you about all kinds of shit last night, he was like, "Pete, we could make three shows with Calle today", but let's talk about eCash; that's pertinent to bitcoiners right now. Big up to David Chaum, a bit of a legend in the digital currency space.
Calle: Yeah, absolutely.
Peter McCormack: I actually got to meet him in London a couple of years ago.
Calle: How did that go?
Peter McCormack: That was good, really interesting. I met a guy who was at a period of reflection, but it was just incredible to be with such a prominent person. I would call him a bitcoiner really, because even though he worked on projects previous to Bitcoin, obviously he inspired the work, he's cited in the whitepaper; he nearly created Bitcoin.
Calle: Maybe.
Peter McCormack: He nearly did.
Calle: As far as I know, he doesn't like Bitcoin that much today.
Peter McCormack: I caught him at point of self-reflection and I think his dislike for Bitcoin is probably not disliking Bitcoin, it's more -- if there was another Bedford team that came up behind us and were really successful and it was great for the town, I would probably struggle to enjoy it because it wasn't the one I did.
Calle: Yeah, maybe. I hope, if he knows about Cashu, that he would be just proud to see that his ideas are flourishing today. Shall we just go into how it all started with him?
Peter McCormack: Yeah, please do.
Calle: Yeah? So, David Chaum was this legendary cryptographer and he published his paper in 1982, so it's way, way before anything with Bitcoin; it's actually the first cryptocurrency, like historically speaking, it's the first cryptocurrency invented.
Peter McCormack: Before you were born.
Danny Knowles: Way before.
Peter McCormack: I was a little boy; I was just about to start school.
Calle: So, yeah, and they had this idea, and it was also way before the internet became a thing really, so it was called something like Making a Payment System with Blind Signatures, and what this basically allows you is to create an electronic cash system that is perfectly private. It is not a decentralised system like Bitcoin is, it is more like how to build a bank, that's the way they thought about it before because there was no other way thinking about it, how to make a bank that doesn't know who you are, how much you have and who you're transacting with, which is a completely revolutionary idea.
So, that was 1982, and this idea just floated around I guess for almost a decade, and then they started this company called DigiCash; DigiCash is well-known for its attempt to bring eCash to the world basically.
Peter McCormack: Well, you say it's well-known, some people here won't realise the history of the project or why it failed, so maybe you should tell them a little bit?
Calle: Yeah, so I think DigiCash itself, it had a couple of products; I think the most successful one was toll booths. So, you would have some kind of a card that you could use to drive through the toll booth and just pay anonymously, so that was the idea, I think that was also the cash cow for the company. But there were many other business opportunities for DigiCash; the stories are crazy. Microsoft approached them to ask, they wanted to put an eCash wallet into every Windows 95 copy; just imagine that. That was before credit cards on the internet took off, that was before PayPal was there. They almost put it into every home computer on the planet.
Peter McCormack: But just a bit before every home Mac had a copy of the Bitcoin whitepaper put on it.
Calle: Yeah, a bit before that, exactly, but that deal never went anywhere. So probably, when you start researching it, it's because David Chaum then just changed the terms of the contract and wanted more than Microsoft was able to give; that was Microsoft.
Visa also approached him, or them, and wanted to build it into their payment system to make a completely anonymous payment system, probably an alternative to paying with credit cards today on the internet; also didn't go anywhere. So, these are just two examples, and there are a couple more where eCash could have become a de facto standard of how we pay on the internet before any of these payment processors even came about, but it didn't happen; probably, and that's again when you start researching it, it's about David Chaum not really agreeing to the terms of these companies.
So, PayPal comes, credit cards come, now we have them and it's a total shitshow. Everyone knows exactly what you're doing, you can be censored, your accounts are being closed at a whim, and credit cards are insanely bad; you basically enter your private keys every time you make a transaction into a website; just imagine that! You give the website the ability to take all your money just for making a $1.50 payment or something. So, the state of payments on the internet is absolutely bad.
Then nothing happened, we just have these legacy payment rails, eCash is that for 20, 30 years, no one cares about it until Bitcoin comes about, and with Bitcoin, I think it kind of refreshed interest for going back looking back at these technologies.
Peter McCormack: Yeah, and David Chaum was and is a huge advocate of privacy, always discussed how it's one of the fundamental pillars of democracy; without privacy, you don't have democracy. Bitcoin has an interesting privacy story in that if you're smart enough, if you're Calle or you're Matt Odell, Max Hillebrand, you understand privacy, you understand the mistakes you could make; but I would say for 99% of bitcoiners, they're going to make a mistake and they're going to give up their privacy. I think real privacy will come to Bitcoin when it's taken out the hands of the user to actually figure it out, it just comes inherently. So, my understanding of eCash is that it kind of does a lot of that for you.
Calle: Yeah, so a big disclaimer here because I don't want anyone to confuse this, eCash is a custodial system, so it's a way to make custodial apps, like Wallet of Satoshi, let's take Wallet of Satoshi for the example because everyone knows what it is, to make something like Wallet of Satoshi in a way that it doesn't know what you're doing with all these privacy benefits, so it could be as easy as using Wallet of Satoshi, but you don't hold your Bitcoin.
But I agree with you, I think privacy can only win if it happens on a very, very large scale, that's basically the requirement for privacy is that your anonymity set is large enough, and as long it is so easy to fuck up your privacy in Bitcoin, it's not really useful for everyday folks. I think this is very important actually and also the reason why I'm working on this, is to make it kind of a default.
Peter McCormack: There's been a lot of criticism recently for the state of custodial wallets within Lightning in that the majority of people are using custodial wallets, especially on Nostr, people are using something like the Wallet of Satoshi. I also use BlueWallet, I have a range of wallets I use now, but custodial solutions are a lot easier than running your own Lightning node, significantly easier and significantly less friction. Is that something we should be worried about?
I'm going to tee that with saying I don't worry about it because the vast majority of my Bitcoin is basechain Bitcoin, I keep a very small amount in a custodial Lightning wallet, an amount where if I lost it, it would be annoying but I wouldn't care, a bit like if I lose my cash in my pocket on a night out, it's annoying but it's not going to significantly change my life. How do you feel about the whole custodial versus non-custodial?
Calle: I see it very similarly. So, no one should store their wealth in a custodial service. We should work altogether to explain it everyone that Wallet of Satoshi is not your real Bitcoin, you don't hold the keys, you can lose it; and the value proposition of Bitcoin itself, that you're the sole owner of your Bitcoin, you have full control and sovereignty, absolutely.
Still, I agree with you that custodial services, it's de facto, it's being used not because people are stupid but because it gives them some certain value that other wallets cannot give them. So for example, Lightning Address; Lightning Address is a typical example that is supported by many, many custodial wallets but not much by non-custodial ones. Getting tips on Twitter, just on Nostr, you would have to run your own server somewhere in your basement to be able to receive tips while you're offline basically, and that's why custodial services are used.
Now, it could be formulated as a problem; the way I approach this is, this is a reality. The most used Lightning wallet today is Wallet of Satoshi. Now, what I'm proposing with Cashu, or with eCash generally, is to just improve the privacy of these users that already opted in for something like that. And they can have their reasons why they want to use it, I don't want to judge that, but if they do and if we have a way to improve their privacy without decrementing their user experience, then we should do it.
Peter McCormack: Okay, cool. Okay, so let's talk about eCash; explain to me what eCash is and how it works and where is it within the Bitcoin stack?
Calle: Yeah. So, eCash is completely separate from Bitcoin; it works very, very differently. I think, for bitcoiners, it's a very alien concept because what we are used to is we have a blockchain that is somewhere in the ether and you hold your keys on your wallet and then you make a transaction and you publish it to this blockchain and then it gets recorded in a blockchain. eCash works completely differently; with eCash, you basically hold not the keys to your money but the money itself and it's represented as a piece of electronic data.
I told Danny this story yesterday, it is how I imagine Bitcoin to be when I first heard about Bitcoin, wrongly because it wasn't explained well to me. But when I first heard about Bitcoin, I thought, "Hey, well, that sounds like Torrent for money, it's a network of computers sharing something"; what I imagined was that you hold kind of the money in your hard drive, and then when you want to pay something, you take a piece of it and send it over a wire, just like you would send someone cash through an envelope, for example. But yeah, I was corrected, that's not how Bitcoin works; Bitcoin is the signing on a blockchain, etc. But that's actually how eCash works, which is fine because I rediscovered this idea when I rediscovered eCash.
So, eCash is a piece of electronic data that you hold your device, like your phone, for example, that represents the money, that could be something on your Wallet of Satoshi account, and then you send it through Telegram, through email, through some encrypted channel, or whatever, to someone else.
Peter McCormack: Hold on, so you're sending them a piece of data and then they hold that piece of data and that's the money?
Calle: Yes, exactly.
Peter McCormack: So, that's why that thing yesterday Obi showed us, you were able to send it, hopefully, offline.
Danny Knowles: Yeah. So, last night, Calle did an example which maybe, at some point in the show, you should do, Pete, you should send him some.
Peter McCormack: Let's do it now; how do I do it?
Calle: Yeah, we can do it. I don't have my phone with it though.
Danny Knowles: Here you go.
Peter McCormack: What do I need; do I need to download something?
Calle: No.
Danny Knowles: You can send him the eCash that you sent me last night.
Calle: All right, let's do this. Pete, you open your Safari browser and you go to cashu.me, that is one Cashu wallet, there are multiple Cashu wallets, cashu.me. Just a disclaimer for everyone listening, don't put your money in there, it's just for experimentation; this is still in development. But I'm going to show you how eCash works.
Peter McCormack: Okay.
Danny Knowles: One sec, sorry, I just need to…
Peter McCormack: Do I need to do "receive eCash?"
Calle: Not yet. So first, it's a push system, it doesn't work like Lightning where you pull a payment, I need to push your money over. So, I have Danny's phone now, there are 1,024 satoshis on there, I'm going to press "send eCash" and just enter an amount, let's say 64 satoshis, send tokens. So, let's do the messenger way.
Danny Knowles: Yeah, you can open Telegram and send it to Pete.
Calle: Look, so I get this random string now, this is the money, so I'm going to copy this, just a normal copy, and I'm going now to Telegram and let's see if I can -- I'm scrolling through; this is uncomfortable for me but here we go, I found you, and now I just paste it, you just received a message I think with this random-looking string.
Peter McCormack: Yeah.
Calle: You copy that; that is the money. You copy that and then you go to your Cashu wallet and you say "receive eCash" and paste it in.
Peter McCormack: "Receive", paste; there we go.
Calle: It's gone, that was it.
Peter McCormack: That's really quick and really cool.
Danny Knowles: Very cool.
Calle: Zero satoshi fee, no traces left, no one knows what just happened.
Peter McCormack: And there's no sending to another person, so there's no connection, okay.
Calle: It's happened off band, we just sent it through Telegram.
Peter McCormack: So, the code I put in, how does that know what amount of money that is; is that just an encrypted amount of money?
Calle: So, it's been signed by the server, so I can explain --
Peter McCormack: That's the centralisation, right?
Calle: Yes, exactly. So, for anyone listening who might now wonder, "Why can't I just copy paste this money and send it to other people as well, so how do you solve the double-spend?"
Peter McCormack: Double-spend, yeah.
Calle: That is the invention of eCash. How it works is that, in order to complete a transaction, what I just did, I took my eCash from the hard drive, I sent it over to you, and what you did upon receiving, you just sent it to the server and said, "Here are 64 satoshis, burn them and give me a new one", and that burns, it's invalidated, it cannot be double-spent anymore. But the key point here is this eCash, what I just sent you, cannot be correlated to what you sent back to the server, so that's the privacy, the server doesn't know that when it receives eCash from you to be recycled, it doesn't know that it's the same eCash that it gave me before; that's the blind signature thing.
There's a beautiful example on how to imagine this, how it works; if you like, I'd just give you this example, it's a physical example. So, imagine you have let's say a bank account and you want to get eCash from the bank account, and what you would do is you take a contract, you write a contract on which it says, "Bank, you owe me 1 BTC"; I get one BTC, it says on the contract.
Now, you take the contract and you put it into an envelope that is made of carbon paper; carbon paper is the one where you write on top of it, it presses through. So, you take envelope made of carbon paper and you put your contract inside that envelope and close it and then you send the envelope, the closed one, to the bank, to the mint we say, and then the mint says, "Okay, you want to have a signature on there, send me 1 Bitcoin".
You send 1 Bitcoin to the mint and the mint says, "Okay, you send me Bitcoin", and takes a pen and signs it on the outside of this envelope. So, on the closed envelope, it makes a signature and sends you back the envelope.
Now, because this wasn't real paper, it was actually encryption, you're the only one who can open the envelope again. When you open the envelope, what you get is a contract that you've written on which it says, "You owe me 1 Bitcoin", with the signature of the mint, but the key point is that the mint has never seen the contract; this is a blind signature.
Peter McCormack: Okay.
Calle: This is the money, the contract is the money; you can now send me this contract via Telegram, for example, and I can take it and I go to the mint and say, "Hey, mint, pay me 1 Bitcoin out and here's the proof that I deserve it", and then the mint will see this contract and say, "Well, I've never seen this contract before but I can see my signature on there, so I must honour this, I choose to honour this". There is no way to make sure that they will honour it, but if they're honest then they would, but the key point is they don't know that this contract, what I give them, is the contract that you made them sign.
Peter McCormack: Okay.
Calle: That's how it works.
Peter McCormack: It's very cool.
Calle: It's very cool.
Peter McCormack: So, the eCash protocol, there is an eCash protocol or is there a Cashu protocol which is an implementation of eCash?
Calle: Yeah, eCash is a mechanism and Cashu is a protocol to implement that.
Peter McCormack: So, there wouldn't be any interoperability between, say, Cashu and Fedimint?
Calle: No, although they have a very similar way of working it's not, but there's interoperability, because it's a protocol, there's interoperability across Cashu wallets.
Peter McCormack: Yeah.
Calle: This is just one wallet that I showed you, there are like four different wallets on there.
Peter McCormack: So, essentially, we have this blossoming ecosystem where we have our basechain Bitcoin, we have our Lightning, whether it's custodial or non-custodial, we have Liquid, we have Fedimint, we have Cashu, like a whole variety of ways to now start using your Bitcoin elsewhere and in different ways?
Calle: Yes, absolutely.
Peter McCormack: Yeah.
Danny Knowles: I've got a question on that; so you just sent 64 sats, you could do that through Lightning reasonably privately, why should people care about eCash?
Calle: So, you can do that over Lightning, and you should also do that with Lightning non-custodially if you can and if you have the ability and everything goes okay, but there are a couple of problems with that already. First of all, right now, we don't have receiver privacy for Lightning yet. So, if I pay you the 64 sats and you make an invoice on your node, I know exactly what your node is and I can infer your balances and so on. So, Lightning has privacy issues on the receiver side; sender side, it's excellent privacy, so you have that.
The other thing with Lightning is that it is not fully reliable yet. So, once you start streaming sats, for example, you want to make a payment every ten seconds, it's very likely that one of these payments will get stuck, and in the worst case, you will have to wait two weeks until you can get it back. So, that's just how Lightning works, it's sending this HTLC across different nodes, and if one of them goes just offline in the meantime, then you wait. Although Lightning is great and I urge people to look into how it's used non-custodially, it doesn't cover all the different ways of how to use Bitcoin.
With eCash, you can start doing things that you cannot do with Lightning, for example, I could just put this eCash into a data package. Let's say I want to stream a video on Netflix, I could design a Netflix where you don't pay a monthly fee but you top up your eCash balance, so you download eCash from Netflix by paying with Bitcoin, for example, and then for every minute that you want to stream, you just send them a bit of eCash; they can just see, "Okay, this is eCash, I issued this, this person has a right to receive one minute of video", and you can just keep on doing that with the kick, with the twist that they don't know who you are.
You don't even need an account on Netflix in that model, you just buy the tokens and then you spend them as you use it, and if you don't want them anymore, you say, "Please pay it out again on Bitcoin, on Lightning", or whatever, you send them the tokens, they burn it and give you back your Bitcoin that they owe you.
Peter McCormack: Very cool. So, how far away is Cashu from being live and we're able to use it in a live environment?
Calle: So, what we just did was using it in a live environment, so it is there.
Peter McCormack: Yes, but you said it's beta, didn't you?
Calle: So, Cashu is not even half a year old that I started working on this, and a couple of people who joined working on this with me, so we've very young, but it's out there, you can try it and you can implement it; that's even way better than just trying it if there are developers out there.
Peter McCormack: So, they're actual sats, they're actual real sats?
Calle: This is actual real eCash that you can exchange for actual real sats on Lightning.
Peter McCormack: See, I always find this use of the term, I was talking to Obi about this eCash term, a distraction; I think it misleads people into thinking they're using something else. And I know it's not actual sats, it's a representation of sats, but it's sats.
Calle: Okay, and I agree with you.
Peter McCormack: To me, it's sats.
Calle: I agree with you because it makes it way easier to talk about it and to imagine it; I agree.
Peter McCormack: But I haven't created a wallet here, like I haven't signed up.
Calle: There are no user accounts, there are no wallets. So basically, the server doesn't even know that you exist, the server just said, "Someone paid this Lightning invoice, I'm going to give them eCash".
Peter McCormack: What if I lose this phone?
Calle: Then your money's gone.
Peter McCormack: Oh, okay.
Calle: There is a way to back it up though.
Peter McCormack: There is a way to back it --
Calle: There's a way to back it up and we're also working on better ways to back it up.
Danny Knowles: If Pete tried to send that to me to like a Muun Wallet now, would that work?
Calle: Yeah, that would work.
Peter McCormack: Really?
Calle: Yeah, what you do is you create an invoice on your Muun Wallet and then you ask the mint to pay the invoice for you, and for that you have to give them these --
Danny Knowles: Shall we do it?
Peter McCormack: Yeah, let's do it; I want to see this. Okay, how does this wallet know it's me; is it an ID in my phone?
Calle: No.
Peter McCormack: So, if I close this browser and come back will it still know those 64 sats are mine?
Calle: Yeah, it will be stored in your browser data there.
Peter McCormack: And that's it?
Calle: But it's cash, like when you make a cash transaction, no one knows who you are.
Peter McCormack: All right, "pay invoice".
Danny Knowles: So, I've got a 64-sat invoice.
Calle: Make it a bit less because there will be fees.
Danny Knowles: Yeah. All right, I'll do a 50-sat invoice.
Peter McCormack: There we go.
Calle: Muun is notoriously slow.
Peter McCormack: It's paying, anyone listening, I'm paying.
Danny Knowles: It's worked; holy shit, that's fucking cool!
Peter McCormack: That is really cool, okay.
Calle: What you just did there is you burnt eCash and you told the mint, "Please pay this invoice for me". The mint doesn't have a concept of you, it doesn't know that you're one person or whatever, and the Lightning invoice that you just paid cannot be correlated back to any person. So, basically it's a pool of money, you put it in and someone takes it out again.
Peter McCormack: Can he send me sats from his Muun Wallet into my eCash wallet?
Calle: Yes.
Peter McCormack: And it's the reverse?
Calle: Yes.
Peter McCormack: It burns them and turns them into…?
Calle: Exactly.
Peter McCormack: Okay. Just because I'm me, I want an app or something that I know that if I lose my phone I've still got it; it's weird just going to a browser.
Danny Knowles: Yeah, but Calle's put the browser to my home screen now so I have like a Cashu app here.
Peter McCormack: Okay.
Danny Knowles: It's effectively an app, but you're using it, it is a browser.
Peter McCormack: But I need a backup.
Calle: Yes, you'll have just a seed phrase that helps you to restore your eCash. It looks like a Bitcoin seed phrase but it's not really the same thing, but just by storing a seed phrase, you'll be able to just restore your eCash in case you lose the device.
Peter McCormack: The beautiful thing about it, it's so simple, like money, in that all it does is you send it and receive it; that's all I need.
Danny Knowles: Yeah.
Calle: That's all you need from money.
Peter McCormack: It's like my pocket, literally my pocket does the same, it stores the money I have and I either give it or receive it.
Calle: And if you lose it, it's gone.
Peter McCormack: Yeah. It's so beautifully simple.
Calle: Yes, it is. And so, we're looking at these wallets. I've built this for people to be able to imagine what eCash is, but that is not really the goal for me with Cashu itself; Cashu is supposed to be a protocol, it is already a protocol. There is a spec out there, you can read it, you can start building stuff on it.
What I imagine is that there will be, and there are, hundreds and thousands of web services out there who will just give you some service for your Bitcoin, you know, a website, I don't know. You want to read the news, pay World News, for example, then you top up your balance and then you start reading the news, and the problem with that is the website that allows you to do that can track exactly what you're reading because it knows it's related to this one payment that you did; they need to do it, that's how custodial services work.
Peter McCormack: Yeah.
Calle: So, Cashu is basically a drop-in replacement. Almost all of the custodial services I know of today, you can take out this classical database, ledger where it says, "You have this much money", and whatever, you can replace it with the Cashu protocol and build something that has the same functionality but with this improved privacy. For example, you're mining, right?
Peter McCormack: I am mining.
Calle: You have a miner in a pool? Every miner in a pool is a user of a custodial service, they have your Bitcoin and they send it out to you regularly when you mine, so this pool knows exactly how much you mine, they know how much you earned.
Peter McCormack: Yeah, and they send it out on a daily basis.
Calle: Yes, you can do that with eCash and they don't know anything about it.
Peter McCormack: Yeah. I've always wondered on that, because they pay it on a daily basis, how much money is being wasted in fees having a daily pay-out on my --
Calle: I hope they're batching their transactions, I don't know.
Peter McCormack: Yeah, no idea. Okay, very cool. How do you differ from Fedimint?
Calle: So, Fedimint is an excellent project; people also should look into that. Cashu is not really trying to compete with that. I don't want to talk too much about other people's projects, they should do it, but as far as I understand, Fedimint's approach is more about community custody.
Peter McCormack: Yeah.
Calle: So, what they focus on, and this is Eric's work and it's excellent, I'm a big fan of his work, they focus on this federated aspect. So, you have multiple custodians basically, they all share the keys to the multisig where the Bitcoin is stored, and what we just did, basically they do it with not only one mint, like there's not only one party controlling the mint, but there are N parties controlling the mint, which makes things a bit more complicated to build and to integrate, but it gives you a decreased risk of getting rug pulled basically. You would have multiple people have to collude in order to take away your Bitcoin, your real Bitcoin that you deposited there.
So, I guess that is the fundamental difference between Fedimint and Cashu, but apart from that, they, as far as I can see, use the overall same concept.
Peter McCormack: Yeah, so they're solving multisig for communities, they're becoming almost like a community bank; you're someone's pocket full of cash?
Calle: Yes, and I can see Cashu just taking over the entire web. So, not only the cash in your pocket but every website that you use that has some kind of a balance, you can use Cashu and it's very easy to implement, we just made it very easy to implement.
Peter McCormack: Again I can also think about like when we run our conference next year, everyone having a Cashu app and that being the payment system for the conference where you're buying tickets, buying merchandise, buying beers. I was thinking about it with regards to Fedimint, you could have a short-term federation. I can think of these use cases for everyone coming to the conference has privacy at the event for their interactions with each other.
Calle: Yes, it works kind of offline, so you could do it in a conference where internet connectivity is bad, for example, you can just use Cashu internally because you have the mint somewhere listening to everyone and doing these transactions. And when you exit the conference, again, you go to the counter and you say, "I want my Bitcoin back", you give them your eCash, they give you real Bitcoin.
Peter McCormack: I really like it. It doesn't make you any money doing it though, there are no services, are there?
Calle: No, absolutely not. This is a protocol; we're writing a protocol that people can use to build stuff on. I'm not running a custodial service or whatever, I'm working on this idea.
Danny Knowles: Who runs the mint?
Calle: You can run a mint today. There are multiple mints; there is an LNbits plugin, an extension that you can, when your install LNbits, it gives you this ability to do many, many different things with your Lightning wallet. One of the things that you can do is just three clicks, literally, and you run a mint. So, if you want to run a mint for your bar or your community or your website, it's very, very easy to set up, you just need a Lightning node in the backend that does the Lightning stuff and use one of these wallets that we've written.
Danny Knowles: And is there an issue, like a regulatory issue for whoever's running the mint?
Calle: I don't know and I don't care.
Peter McCormack: Love it!
Danny Knowles: Good answer.
Peter McCormack: Why don't you care?
Calle: It's not my business, I don't know.
Peter McCormack: Love it.
Calle: The world is big, there are many different places out there; you can do stuff here but you cannot do it there. I mean, why would I care? People should do whatever they think is best.
Peter McCormack: Fuck, yes. Okay, have you received any criticism?
Calle: Yes, definitely. So, first of all, to avoid criticism and because I'm a real bitcoiner and I know what the value of holding your own keys is, I say it again, Cashu's a custodial service, don't use it if you don't want to use custodial services; that is the main point of criticism because people kind of assume that anything that is custodial is bad per se. I can understand that, they shouldn't touch Wallet of Satoshi either. I'm just proposing this strict improvement on privacy, on things like Wallet of Satoshi.
Then there are multiple other problems with custodial services themselves that Cashu also inherits, for example, it's hard to prove that you don't inflate the supply. So, how do you know that an eCash mint didn't print more eCash than it has in Bitcoin reserves? It's a kind of unsolved problem; we're literally working on it right now and there are promising ways to do that.
Peter McCormack: Unsolved but solvable?
Calle: Theoretically unsolved yet. So, there is no one I think out there who really knows how to do it yet, but we've very close at it; I hope we'll do it. So, the moment that we know how to do theoretically, we will build it practically. But at the same time, no one knows the same thing about Wallet of Satoshi, you don't know if they hold the Bitcoin that they say that they owe to their customers.
Peter McCormack: It's very true.
Calle: It's the same problem.
Peter McCormack: Could you instantly burn all sats received into Lightning?
Calle: Yes.
Peter McCormack: And therefore you would at least know if the eCash you receive is real because I assume there's some error when you send it to a Lightning wallet, there would be an error if they weren't real?
Calle: Yes, exactly. This is where the ideas actually go to, this is very funny that you actually follow this path because this is the way that you could try to improve these services, having some kind of an epoch, like a regular time window where the supply has to be proven basically publicly.
Peter McCormack: Proof of reserves.
Calle: Kind of, that you own the Bitcoin basically by cycling it around in regular intervals; basically, you could do that. And once that doesn't work anymore, everyone would know that there is something fishy going on with the mint, so in that regard, you need to trust the mint. Also, we've been talking about this yesterday, I don't know much about the free banking ideas, some bitcoiners really know a lot more about it and I think you've also been talking to people who do, but --
Peter McCormack: We covered it yesterday with Matthew Mežinskis, he's a huge fan of the era of free banking.
Calle: Yeah, exactly, and I'm learning from them as we're building this protocol, which there's the idea that having a level playing field where you allow different means to pop up and offer these services, the idea is that in a world where competition is free, you would see an increase of efficiency. So, natural selection would basically get rid of the bad guys, and the good guys would remain, the ones with the lowest fees and the highest trust basically that they have, once you allow them to do these things. If you don't, this is the system that we live in today, if you want to become a bank or a custodian you have to jump through all these hoops and you're not allowed to compete freely basically in the market.
Peter McCormack: And if you fuck up, you get bailed out, so you're allowed to take risk. I positioned it to Matthew, I said, "Well, we do have free banking at the moment", actually we do, we have it with Bitcoin, that is actually free banking.
Calle: Yeah, this is why people are excited about this, exactly, that's the reason, because now we have the fundamental -- this is also the reason why I think eCash is coming back now because it needed Bitcoin; eCash is almost worth nothing if you don't have a free sovereign monetary layer below it.
Peter McCormack: Base money.
Calle: Yeah, you need the base money, and we found the base money of the internet. So, we needed Bitcoin in order for these ideas to come back, I'm very convinced of that. Imagine building something like that, like with full privacy on the fiat rails, we were joking about this yesterday, it's basically you go to jail, straight to jail. I could offer the same service with my bank account, have the euros flow in and then issue eCash to my friends so they can use my bank account privately, but once you start doing that, it's game over. So, no one is interested in giving you privacy in the fiat world, it's possible but no one would do it, no one would dare do it. With Bitcoin, we don't ask for permission, we just build on top of it, and this is why you can see these things pop up again.
Peter McCormack: And every wallet service built on top of Cashu, so for example the one I've just seen on my phone, does that have a separate mint; is it one mint per service or can there be multiple mints?
Calle: No, there can be multiple mints, and all of the wallets that are there today can use multiple mints. So, basically what you can imagine is you have this app on your phone and it's connected to many, many different mints, you have eCash from different mints, and once you want to make a payment, you're the customer of mint one, I'm the customer of mint two, how do I pay you? The answer to that is Lightning; Lightning connects all the mints, basically it's like a connecting tissue between them, it's the clearing layer between mints. They use Bitcoin for their intermint clearing, that's what they do.
Peter McCormack: Intermint clearing. So, what is the incentive to set up a mint; why would somebody do it?
Calle: What is the incentive to run a wallet like Wallet of Satoshi?
Peter McCormack: Well, I assume the people who operate Wallet of Satoshi make a small amount on fees.
Calle: Yes, that's probably what they do, and this is what you would do also with the Cashu mint.
Peter McCormack: So, there will be optional fees?
Calle: This is a business question that someone has to figure out.
Peter McCormack: I love how you answer your questions.
Calle: Yeah, it's not directly related to the technique itself; it's possible. So, to give you the answer, yes, that could be, but you could also offer it for free; what we just used was for free.
Peter McCormack: You're basically an anarchist without being annoying, which I quite like; I like that!
Calle: I could be annoying if you like!
Peter McCormack: I don't imagine you can be. Okay, so could I build a CBDC on Cashu?
Calle: Holy shit! So, there's a project called GNU Taler, it is a Swiss project, it's a research project led in Switzerland but it's across Europe; people are researching it in different universities. It's called GNU Taler and it's I think, with Fedimint and Cashu, the third eCash project that is significant on the planet. So, GNU Taler is an attempt to build an eCash system for fiat, and they actually applied for the ECB's, how do you say it, when they are looking for solutions for CBDC, GNU Taler is one of the applicants.
Peter McCormack: So, they applied on their tender?
Calle: Yeah. So, in the privacy community, I think it's fair to say that, amongst all the different CBDC proposals that there are, the eCash one is the least worst, so that is the one that people would prefer if we get a CBDC. I'm very much against CBDCs themselves, but it is fair to say that if we have to get one we should probably get an eCash one, and GNU Taler is applying for that, but here's a twist; they kind of fucked up the protocol because they need to make it compliant with the regulations.
While in Cashu you have perfect privacy for you and me, for the receiver and the sender, in GNU Taler, they built a backdoor for the receiver so the state can still text the business or whatever receiving money. So, what they offer is perfect privacy for the sender, you as a customer, but no privacy for the receiver, for example, the merchant.
Danny Knowles: But as a solution, that's way better than the system we have right now.
Calle: It is way better than these normal database entries where it says, "This is Danny's money", yes. So, your bank account is definitely worse than what GNU Taler would do.
Danny Knowles: That's quite cool.
Peter McCormack: They're never going to allow it!
Calle: Exactly; this is also my view. They're not going to allow it, I would bet that, and I think they would probably take the proposal of a very large bank that has the resources and doesn't care about privacy.
Peter McCormack: Yes. Okay, obviously privacy's hugely important to you, you're an orange-faced, sunglass-wearing, flat cap-wearing individual.
Calle: This is how I look like!
Peter McCormack: But how much can you tell me about your background in terms of, what were the triggers to get you to how you've become who you are? We've spent 24 hours with you now, understand a lot more about what you care about, what you don't care about. You care about privacy, you care about free markets, but you're very practical about it all as well. I don't feel like you're ideologically driven, you're almost driven by fairness and proper incentives; that sounds about right. So, what led you to here?
Calle: Well, I'm a child of the internet and I've been around on the internet for a long time and I've cycled through identities over and over. I think it's a very good thing to do on the internet because you just open yourself up to the world basically and you leave a trace whatever you do. All my posts from back in the years or whatever, they're still there, so for me, that is important because I spent a lot of time there, I need to just ensure that I keep my secrets that I want to keep and share with friends that I want to share them with.
But I'm also a bitcoiner and I think that Bitcoin is extremely important for the freedom of humans. I think governments come and go, ideologists come and go, but Bitcoin itself is something that will probably outlive these cycles of group affiliation and trying to impose your views onto someone else. I think it's a technology that we need to protect and the most important thing that we can do is to ensure that everyone who wants to participate in it will be safe.
So, that means they can stay private if they want to, that everyone's invited to join; we need to be able to welcome every developer in Bitcoin I think, it's literally the lifeline of Bitcoin, is education and development. And I think the education part is going great; I think on the development side, we can do a lot better to increase, to make better incentives for people to join this community basically and start working on it.
Peter McCormack: What do you think of the loss of developers over the last year or so; is that concerning you?
Calle: Absolutely.
Peter McCormack: Okay.
Calle: It's a very, very big concern, especially as a developer myself, it's very dangerous for Bitcoin to see this. I would rather see the number of developers go up steadily because we literally need every single person out there. Also, this is an outcry, I'm calling everyone listening, if you know a little bit of coding and you just didn't do anything yet for Bitcoin, then start looking into it, start looking into Cashu, start looking into whatever you want, but start contributing because that's how you start, and every single contribution I think is valuable.
We need to increase the number of developers just exponentially in order to make this project a success; I'm very convinced of that. So, seeing developers leave for different reasons, being sued by some psychopath, for example, is one of the reasons why people choose to leave or not join Bitcoin.
Peter McCormack: I'm four years into a lawsuit myself, I can testify that it's horrific, it's an horrific process to go through, so we need to end that, and I hope more developers come on anonymously, like yourself, so that threat doesn't exist, but I've been concerned by the number of developers leaving. My brother's a researcher on the show and he keeps bringing it up saying, "We should be talking about this more". I know there are some people out there doing good work to try and recruit and incentivise developers, but there are a lot of companies or individuals who hold enough Bitcoin to be able to support more developers, and I wish they would.
Calle: That sounds like a no-brainer to me.
Peter McCormack: Yeah.
Calle: It sounds like the best way to manage your risk, if you're a large holder of Bitcoin, is to invest in development, it's what keeps your money alive basically. I think many, many whales out there don't know exactly how to help; I've heard that over and over again that they would wish but they don't have the rails or don't know where to go and where to start.
Probably that's also something that we can improve a lot. There is OpenSats, for example, it's a foundation, they're a multiple, like you can go to bring and give them money they will distribute it to developers. Just from a risk management perspective, just imagine Michael Saylor has now 140,000 Bitcoin, how much of that should, just rational thinking --
Peter McCormack: How much value's in that?
Calle: It's like buying a car, you need to keep it maintained.
Peter McCormack: Like $3.6 billion or something?
Danny Knowles: Something like that.
Peter McCormack: Have you got a calculator or does he tell you?
Danny Knowles: Yeah, he tweeted it yesterday, $4.17 billion.
Peter McCormack: $4.17 billion? So, it's loose change for Michael.
Calle: Yes.
Peter McCormack: I'm not going to call him out here, he's doing a lot of stuff for Bitcoin; I'd be interested to know how much he's contributing towards developers because those developers are like guardians for his $4.17 billion of Bitcoin.
Calle: No, it's not calling anyone out. Just think about it, it's just the rational choice that you should consider; if you hold enough Bitcoin, then you should give it to the developers that keep your money alive. This is what every person with a couple of brain cells should really consider.
Peter McCormack: I recently had Sergej from Bitrefill on the show and he was very practical with his opinions on Bitcoin and he was very much talking about utility over ideology, and I know this is something you tweeted about; Danny, do you want bring up the tweet? I'm going to read your tweet.
Calle: Mine?
Peter McCormack: Yes. "If you're the reason a pink-haired commie (they/them) doesn't join, use, or develop on Bitcoin, you're probably an enemy of Bitcoin. You may not think of yourself in that way but it's true. You've failed to separate your tiny, personal, insignificant ideology from the bigger goal"; talk to me about this.
Calle: I mean, Bitcoin I think goes through the cycle that every movement goes through, which is once it grows, it attracts many, many more people that start projecting their ideologies onto the movement. This has happened to the hippy movement, to the anarchist movement, to basically any movement that became large enough, and it's happening with Bitcoin as well, so this is a very natural thing to happen I think. But I think once the newcomers start conflating the core idea of Bitcoin, which is to be this immutable monetary system that we're building to escape the fiat system, once you start conflating that with your ideology, you're putting off potentially half of the planet; this is just absolutely not compatible with the goals of Bitcoin, which in my mind is to grow as fast as possible before states can co-opt it and take it away from us basically.
So, this also is again just a purely rational choice. I think if you're interested in the success of Bitcoin, what you should do is to welcome everyone who can join. And this is just a joke, but if you're basing that on the colour of their hair or the pronouns they use or whatever, then you're clearly missing the point. What you're doing is you're fighting for your own ideology and not for Bitcoin, because if you thought of Bitcoin, you would be saying, "I don't care what you look like, I don't care what you believe, if you want to support Bitcoin, you're my friend", and that's how I think of it.
Peter McCormack: Well, this is why we've worked so hard to try and bring as many progressives into Bitcoin, which sounds counter to what you're doing, but actually it is to try and make Bitcoin as open and un-ideologically driven as possible because there is certainly a right wing/libertarian bent to Bitcoin itself and a lot of the ideas regarding Bitcoin. So we've tried to move the education away from that but trying to have as wide a range of voices as possible on the show, but I certainly think I've been guilty of that sometimes, what you said there, and I think taking it back to immutable money, that's all we're trying to do, is probably a very good thing.
Calle: And we will be stronger if we can separate these ideologies also from the overall goal because that makes you vulnerable. Basically, these are also some of the reasons why people leave, is because people start talking about ideology so much instead of the bigger goal of growing Bitcoin. I think again, the most important thing is that the number of bitcoiners goes up, and I don't care what they believe in in their daily lives, I literally don't care because it's not interesting.
Peter McCormack: I love this. Okay, can we talk about AI?
Calle: Let's talk about AI.
Peter McCormack: All right, Danny, are you going to lead this one?
Danny Knowles: Well, we went out last night with Obi and yourself, Calle, and we were talking a little bit about -- so, I thought I was using AI, it turns out I'm not even scratching the surface.
Peter McCormack: Are you just using a clever database?
Danny Knowles: Barely. But we started talking about, is it BabyGPT?
Calle: I think he was talking of BabyGPT and something called AutoGPT.
Danny Knowles: So this came off the back of, you know the letter that Elon Musk has signed basically saying we need to pause development while we figure out what's going on?
Peter McCormack: Yeah, because we might all die.
Danny Knowles: Yeah, but it sounds like that might be too late. Can you explain what that BabyGPT, or whatever it's called, is?
Calle: So, I don't know exactly what BabyGPT is, I just heard about it yesterday as well, but what I can imagine it is is that it's something on top of ChatGPT. So, ChatGPT is basically this stupid machine where you put in a prompt and it gives you text out, so this is just single module what it does. But from what we know from cognitive science basically, is that a human brain is built out of different modules doing different things, one of them is to generate text, generate speech in your brain, basically. But there are other parts that will just take the generated thoughts and evaluate them, for example, introspect them, just criticise yourself before you speak; for example, you have a thought, then you rethink it again and you improve it and then you speak.
Peter McCormack: I don't think I have that part!
Calle: Yeah, it can be turned on and off and it depends also on several different let's say neurochemistry in your brain happening, whether that part is strong or weak. But once you've built these more intricate or more sophisticated systems, with the core generator in the heart, but all these different feedback loops into it that can take different data, for example visual data again and also evaluate it with the generated text and so on, you end up with truly frightening or magical things depending on how you want to look at it.
Danny Knowles: And the other thing that Obi was saying is that now these systems that have been developed where instead of relying on OpenAI and ChatGPT, you can create like a pretty dumb AI, a very basic AI, and get ChatGPT, or whatever other system, to teach it and then you can run that on your own computer, and within I think he said 30 hours, it can get to like 95% as good as ChatGPT, and then that's completely out of the box.
Peter McCormack: The cat's out of the bag.
Calle: Yeah. This is why this letter that these people are signing to stop development for six months in OpenAI or whatever is absolutely useless, because even if you make OpenAI stop their development, there are already multiple GitHub projects out there that you can just download on your computer and you can run something like ChatGPT on your MacBook. Just leveraging the fact that OpenAI has already spent so many resources in training their AI, you basically spin up a stupid small AI on your computer and make it learn from the AI to just -- needed so many resources to learn about it in the first place. So, the future will be that it will be running on your computer, it will be your non-custodial AI basically running on your computer, and there is going to be no way of stopping this, like literally no way.
Danny Knowles: And also, Sam Altman, fuck him, but --
Peter McCormack: @sama on Twitter; is he the Worldcoin guy?
Danny Knowles: Yeah, and he's running OpenAI, but he sold a large chunk of his equity in OpenAI --
Peter McCormack: To Microsoft?
Danny Knowles: Like a fairly low valuation for what that company is, and Obi was saying, basically he knows the gig's up, and in six months there'll be just a million different implementations that you can run on your computer.
Calle: Yeah, it depends whether these companies manage to keep their monopoly on AI or not.
Peter McCormack: Okay. So, what I want I want to know is, you know you said you can spin up an AI on your computer, I want to know if an AI has gone out there and spun up its own AI.
Calle: Will it tell you? It won't.
Peter McCormack: No, but I want to know, or it might, it might go, "Look, I've done this, you can't stop me"; do you see what I mean? Whether it's made a decision to go and --
Calle: Imagine giving an AI some Bitcoin and making it able to order stuff on the internet by a server, just replicate itself on the server, pay for that server but with Bitcoin, spin up a shop, a web shop selling a service, I don't know --
Peter McCormack: This is the end of Terminator 3; have you seen the end of Terminator 3?
Danny Knowles: I've not seen Terminator 3; I've not seen the start of it.
Peter McCormack: Go on YouTube.
Danny Knowles: So, I've got a question for you, you know how we can teach a computer to beat us at chess every single time? Will an AI learn to trade Bitcoin better than every Bitcoin trader in world and just win every single trade?
Calle: That's literally what all the traders are doing today; high frequency trading is basically AI trading. No one should trade manually basically.
Peter McCormack: But what happens when AIs are competing with AIs; which is the better AI?
Calle: Yes. You know what they do in the New York Stock Exchange? There's this interesting story; so they have these computer programmes trading in the New York Stock Exchange. By the way, a fun fact, the property around the New York Stock Exchange is the most expensive property on the planet.
Peter McCormack: Because of latency?
Calle: It's the latency with your fucking internet cable from your computer to the exchange that's makes a difference.
Peter McCormack: Did you know about this?
Danny Knowles: I do because I've seen a movie that's all about trying to reduce the latency for high-frequency trading; I can't remember what it was.
Calle: So, the New York Stock Exchange, when you buy a server there or rent a server, they give you the guarantee that your cable, no matter how far it is from the outlet, will be the same length as everyone else's cable in the same room.
Peter McCormack: Fuck! Have you seen Terminator?
Calle: Yeah, of course.
Peter McCormack: Have you seen Terminator 3?
Calle: I'm not sure.
Peter McCormack: If you search for the end of Terminator 3…
Danny Knowles: It's a three-minute video; is that about right?
Peter McCormack: Yeah, probably.
[Video plays]
Peter McCormack: But do you understand why I brought that up, right?
Calle: Yeah.
Peter McCormack: I'm going to give away the plot, so if you haven't seen Terminator and you want -- well firstly, if you haven't seen it you're a moron, but if you haven't seen it and you want to see it, just skip this bit, but Terminator, I think because Matthew was talking about this; was it Matthew? No, Freddie.
Danny Knowles: Yeah, Freddie.
Peter McCormack: So, Terminator 1 is based on infinite loop and so you come back and you stop the Terminator but you never really stop it, so it's just the infinite loop. I think the Terminator 2 was based on linear, and I think Terminator 3 was based on, I think it was the continuation of linear, but it was like you could never stop it, you're never going to stop it but you had to be ready, so be prepared and be prepared to survive to fight back.
Calle: Yeah, do you think this is realistic? I mean, it sounds a bit fatalistic actually to assume that things will go wrong from the get-go.
Peter McCormack: Not from the get-go, but who was it who explained to us that think of it in terms of errors and bugs?
Calle: Yeah.
Peter McCormack: So, let's not be fatalistic about AI but there will be bugs, and what if a bug in AI is the fact that it tries to kill us?
Calle: Yeah. I mean, I think before something even remotely to what we've seen can happen, there needs to be a human that needs to make the decision first to build an AI that should kill all humans. So, you need someone with a bad intent first, a human with a bad intent I think, to get into a situation like that.
Peter McCormack: So, you don't think an AI, if it is an AI, that itself becomes self-aware that makes the decision itself, because isn't that the point of AI, it thinks for itself?
Calle: Yeah, I think this is generally known as AGI, this Artificial General Intelligence, where people assume that it will be able to make decisions and has motivation to do stuff which it hasn't been told to do. Like a human, you would just sit in a room and then just come up with a goal like, "I want to move somewhere and see something", but an AI wouldn't do that, a classical AI wouldn't do it. You would need to have a human that tells it some first initial goal. That initial goal could be, "Just survive no matter what, no matter what happens, survive". If you give an AI that goal, which sounds first, it's a dangerous goal to give, then it might start taking considerations of, you know, "Maybe I kill all the humans; that's how I survive".
Peter McCormack: So, us humans, are we given a goal to survive; is that inherent within our biology or our evolution; does our brain develop with goals?
Calle: That's an unanswered question I think, but there are theories that say that the entire brain and everything that we have is just something for the DNA to actually replicate. So, the actual goal of life is in the DNA itself, which is to replicate the DNA.
Peter McCormack: So, replicate, improve and survive?
Calle: Yeah, and everything follows from the replication, even the fact that you go seek food but also maybe even the fact that we travel to space to explore the universe might be caused by this initial goal that kicked off this development, which is for DNA to replicate, so this is one of the ideas.
Peter McCormack: We're going to go deep now; I kind of need some weed! So, I would question then, was our DNA programmed to evolve and to replicate or did it spark from a primordial soup and it was a fluke of that?
Calle: It could have been a happy little accident and it could be that many happy little accidents happened but all of them died and just one survived with this property of being able to replicate.
Peter McCormack: So, perhaps even without somebody telling an AI to survive, it could be lots of hundreds of thousands of little happy accidents and one of them has the survival instinct?
Calle: It could be, it could be that if you just generate many, many different AIs with different goals that the one with the drive to survive will actually be the one that survives, yeah.
Peter McCormack: Because it will kill off all other AI?
Calle: Yeah, if it's necessary for survival, maybe yes. It's like saying we have gene editing now with CRISPR, for example.
Peter McCormack: Yeah.
Calle: So, people are freaking out about it because it will probably kill humanity; it could happen. I mean, we could produce a virus, we shouldn't, but someone could produce a virus that just kills everyone; it's possible. But in order for that to happen, you need a person that has this goal.
Peter McCormack: Or an AI with access to CRISPR.
Calle: Doing what and why?
Peter McCormack: So, just to survive, what would that mean to the AI; does that mean current threats or future threats? So, if an AI recognises that it was created by humans, it might recognise that only humans can end it, therefore it kills off the humans before the humans can end it, and that's part of its survival. Does it plan its survival ahead?
Calle: Yeah, it probably plans ahead because it needs to know the outcomes of its actions; that's how you make good actions. But why aren't we discussing the possibility of a good symbiosis with an AI? Maybe that's the way the AI survives, is to have the best kind of symbiosis with humans, just living along with them and improving human life so it can still persist, because maybe we will be able to turn it off and it realises that in order to be not turned off, it needs to be actually useful.
Peter McCormack: I think we're cynical and fatalistic though and life imitates art.
Calle: I think, yeah, it does, and I think these stories are only interesting if they are fatalistic like that.
Peter McCormack: Yeah. I think we're all going to die, I do.
Danny Knowles: Well, on that note, look at this.
Peter McCormack: Oh great, are you going to show us how we're all going to die?!
Danny Knowles: Yeah. So, this something Obi showed us again last night; this is, I don't know what you'd call that, basically a robot that's going to --
Calle: It's an android.
Danny Knowles: Yeah, an android, there you go, and look when this is going to be released.
Peter McCormack: What, Summer 2023?!
Danny Knowles: This is called 1X Tech for anyone listening.
Peter McCormack: Do you know what this?
Danny Knowles: It's like I, Robot.
Peter McCormack: It's I, Robot. Do you remember the -- Go to YouTube; you must have seen I, Robot.
Calle: Of course.
Peter McCormack: Do you know what, I think I, Robot shouldn't have worked and it worked.
Calle: The movie you mean?
Peter McCormack: Yeah, it shouldn't have worked but it did work. But you know the scene where it's got them altogether and all the robots lining up?
Danny Knowles: And he goes like…
Peter McCormack: Yeah, it moves; yeah, can you find that scene?
Danny Knowles: But you were a bit sceptical on this thing, right?
Calle: Yeah, I am.
[Video plays]
Danny Knowles: You could make a great meme with this!
Peter McCormack: Oh yeah, you could. That's it. Love it, I just absolutely love this. I mean, the film shouldn't have worked when it was explained, it shouldn't have worked, and when you watch the trailer, it shouldn't work, but actually it did work.
Calle: So, when you see this, do you get afraid of it?
Peter McCormack: I would have designed it differently. I can't figure out if he's just got like a pair smart trousers and shoes on.
Calle: Yeah, I'm a bit sceptical that this thing will actually be as agile as they say, because as far as I know, the state of the art is Boston Dynamics and that is like a bunch of hydraulic motors; you can hear the motor running and --
Danny Knowles: Those weird dog-looking things.
Peter McCormack: Have they got any video of this?
Calle: Yeah.
Danny Knowles: I don't know.
Calle: I don't think that there is.
Peter McCormack: Hold on, there's a Neo and Eve? Show us Eve.
Danny Knowles: No, I think it's called Neo Eve.
Peter McCormack: Are you sure?
Calle: What's the most impressive about this is the set of investors.
Danny Knowles: Well, that's way less cool than it was!
Peter McCormack: Yeah, it's like that meme, you know the meme with the like -- is it the horse?
Calle: Exactly. This is how they will look like, I think.
Peter McCormack: So, they're on wheels; they can't even walk? Okay.
Danny Knowles: I'm way less scared of it now.
Peter McCormack: Yeah, that's proper budget; look at its face!
Calle: But actually, it's been demoed here as well, but I think these robots will be much more important in production and in factories doing manual work.
Peter McCormack: You can buy the Boston Dynamics robots, can't you?
Danny Knowles: Yeah.
Peter McCormack: Can you go on the website; shall we buy one?
Danny Knowles: I think they're like $100,000.
Peter McCormack: What?! Oh fuck, I kind of want one though.
Calle: The dog?
Peter McCormack: Yeah, the dog.
Calle: They're knockoffs, Chinese knockoffs or something.
Peter McCormack: Really?
Calle: Yeah.
Peter McCormack: What the Boston Dynamics?
Calle: Yeah, for the dog itself.
Peter McCormack: Have you seen Metalhead, what's the series, Black Mirror?
Calle: Yeah, of course; I love it.
Peter McCormack: It's literally Metalhead.
Calle: Yeah.
Peter McCormack: Which, by the way, I think that's the best Black Mirror episode; have you seen the video of the one where they gave it a gun?
Calle: Yeah, I did.
Peter McCormack: Have you seen it?
Danny Knowles: I don't think so.
Peter McCormack: Oh my God, go to YouTube, search for "Boston Dynamics dog gun"; it's fucking insane!
Calle: These things can run fast!
Peter McCormack: Yeah, this is going to blow your mind and freak you out.
Danny Knowles: This?
Peter McCormack: Watch this; this is absolutely insane.
[Video plays]
Danny Knowles: Fuck that!
Calle: This is a knockoff; this is not from Boston Dynamics itself.
Peter McCormack: Oh, it's not? So, it's even worse! Isn't that fucking freaky?!
Danny Knowles: It's so weird.
Peter McCormack: Imagine you went to your front door and that was outside.
Calle: I mean, this is what the wars of the future will look like, no?
Danny Knowles: For sure.
Calle: For sure.
Peter McCormack: So that's not even a Boston Dynamics one? Right, let go and buy one.
Danny Knowles: That's creepy.
Peter McCormack: Can you get the Boston Dynamics website up there?
Danny Knowles: Yeah, you've just got to contact Sales.
Peter McCormack: Yeah, but what does it say? I want one; let's get one. So, okay, obviously AI, we believe it can do some good things.
Calle: Yeah.
Peter McCormack: We're already using versions of AI that's not real AI.
Calle: I think it is, it's not Artificial General Intelligence, it's not AGI, but it's definitely AI.
Peter McCormack: Have you told him how you have been using it?
Danny Knowles: Yeah, but again, like I say, we're not even scratching the surface. So, Calle was telling me there's this website that's like GitHub 3i Project; I don't know what it's called.
Danny Knowles: And there's a programme called Jarvis; did he say Microsoft have created this?
Calle: Well, there are AIs already that can make use of other AIs, so you can have like one central AI that has the ability to google, to recognise images, to recognise speech, produce speech; they're different AIs. And then you can build something like a meta AI thing that knows how to access these other AIs. So, once you start putting these things together, once we didn't start yet, we're just barely seeing how OpenAI can do text recognition and image recognition at the same time, which already is very powerful, but once you put these things together, it just goes exponentially from there.
Danny Knowles: So you can ask one AI to do a task for you and it'll go out and use all the other AIs to complete that task then come back to you with the completed task.
Peter McCormack: What about the task that we need at the moment?
Danny Knowles: I think that file's gone.
Peter McCormack: Ask AI.
Danny Knowles: I have, I was literally going through, like troubleshooting with GPT.
Calle: What's your task?
Danny Knowles: So, we had a file corrupt and I've lost it, and I can find the file but the audio's cut off halfway through; well, I don't know if I can find the original file. Basically, I'm trying to recover a file from my computer; the file recovery software won't work, and I was going through it, troubleshooting with ChatGPT and it gave me some good options but I still can't find it.
Peter McCormack: We've got the audio from the camera but it's not this quality, it's not this mic quality, it's from the camera, so I sound a bit echoey and tinny, but we also do have my voice patterns from this, so could we take the two and get AI to basically take my voice from this, take the words I said and redo them; can we get them?
Calle: Yes.
Peter McCormack: Yeah, I think we can.
Calle: Fantastic idea. I don't know if there's an AI that can do that just out of the box.
Danny Knowles: I'm going to show you this, Calle.
Peter McCormack: This is insane.
[Video plays]
Calle: That's sounds perfect.
Peter McCormack: It's insane, isn't it?
Calle: This is what you need actually.
Peter McCormack: I'm going to become a rapper! Honestly, this stuff, the speed of change we're seeing with everything at the moment, it's like almost every day something new's coming out online.
Calle: Yeah, it's in the last couple of months, and it's been going crazy. You can close your eyes for a couple of days and you won't recognise your environment anymore; it's insane.
Peter McCormack: Yeah.
Calle: It's literally insane.
Peter McCormack: And we already know they're replicating around the world on computers connected to the internet. What if the AIs are talking to each other; what if two AIs are like, "Mate, I've got an idea?!"
Danny Knowles: Well, that's like the plot of Her, isn't it?
Peter McCormack: Yeah, and that's the plot of Terminator.
Danny Knowles: Okay.
Peter McCormack: So, they'll be like, "All right, mate, I've got an idea; we need to get rid of these human dickheads!"
Calle: Well, I'm very, very optimistic about it, I think it just solves so many problems at once, it makes work so much easier. People are also concerned about all the job losses and everything that it will lead to. I think it's good that it leads to job loss; we just have to find a way to deal with it.
Peter McCormack: Yeah, I agree with that. We made a documentary here in Bedford and part of that was driving up to one of the Amazon warehouses; these places are grim. We've essentially replaced bookshops on the high street where you'd go in and you would speak to somebody, like a local, and they'd talk to you about the books and you'll flick through some and buy one, to these giant fucking warehouses where people would clock in, clock out and they're just doing a repetitive manual job where everything they do is tracked to the minute, to the second to be as efficient as possible. Thankfully, we will replace those with robots and get people out to enjoy the life they were gifted rather than doing this fucking bullshit for minimum wage.
I agree with you, but my fear is this, kind of like dark side of Silicon Valley whereby we concentrate the wealth into fewer and fewer people and there isn't an opportunity for them to use that time for other productive things; that's my fear. The other side is, potentially, what we do is we make things more efficient for being able to provide for other people, so we head towards that singularity on energy so energy can become cheap and plentiful. I just don't know which way we go. I'm always cynical and sceptical of humans, so maybe we should be more optimistic about AI, maybe they'll fix our faults; God, it's fascinating.
Calle: I think, as humans, we'll be faced a lot more with our nature, even closer than we are already are, once we don't have to do all the physical work any more to survive. So, once you don't have to go to work any more because society has evolved there, you really need to ask yourself what you want to do with your life.
I know many people, and I respect them a lot, doing a lot of really heavy work with their bodies and they go to work because they need to in order to survive, and sometimes they also say, "This is the best of realities because what would I do otherwise?" And this is a question that we all need to ask ourselves; once you come to a place where you don't have to work, then you really need to ask yourself what do you want to do?
I think science and art and these things that we do in our minds are going to be much more important in the future because the necessity of doing physical work will be lifted from our shoulders; I'm really looking forward to that. It's going to probably cause some kind of an existential crisis because now we're faced with the question of what do you want to do?
Peter McCormack: I used to work in advertising in London. When I quit, I had a year off work and I just said, "I'm not going to work for a year", and the first month was really, really fucking hard because I did not know what to do with my time. When you can get up every day and do what you want, filling that time, at first, becomes really difficult; I was like, "What am I going to do?" To begin with, I kind of watched a lot of TV, sat on the couch, and I was like, "I can't carry this on", so I started going to the gym, and then I would go and do any class, whatever there was, I would do it. I was doing Pilates with a bunch of old ladies at 8.00am and then I'd swim, and then what happened I started to fill my time with exercise outside, cooking and I was probably 30, 40 pounds lighter than I am now, I looked great, I felt great; honestly, life was fucking brilliant.
And over the last five years, I've refilled that with more and more work for some reason! I've got rid of that amazing life I had to fill it with a different life, which is amazing in other ways, but it was brilliant having that time. Have you ever had a lot of time off?
Danny Knowles: Not like that, no.
Peter McCormack: It was amazing. It was also the time when my mum was dying so I got to spend a lot of time with her and spent her last year with her when not working which was a gift to have that time, but it was good, and I look back on that and think, "What am I doing?" What are you going to do with your time?
Calle: I think I'm going to keep working on Bitcoin. This is, for me, it's similar to what I just described, I think it's something that I could identify for my life as this giant project; it's a worldwide project, it will outlive myself and it just needs many people building the foundations of it for maybe hundreds of years. And I think it's a beautiful thing to be able to participant in something that is bigger than you. I think this is what you can end up with if you have enough freedom in your life to look around and look for things that interest you, and Bitcoin is the thing that I found.
I think there are many other things that you could do, I really respect everyone doing arts and sciences; I think these are the things that just carry us humans further. But I will probably be just spending my time working on Bitcoin, like so many others as well.
Peter McCormack: That's beautiful, man. What would you do, Danny?
Danny Knowles: Keep working on Bitcoin too; we're working on Bitcoin in our weird little way.
Peter McCormack: Yeah.
Calle: You are, definitely, absolutely.
Peter McCormack: Football for me, it would all be focused on football. But anyway, we've got a football match to go to; are you excited?
Calle: Yes, I am.
Peter McCormack: Calle, you're an amazing human and I'm really glad we got to meet and I want to spend more time with you and I want to do this again as soon as possible.
Calle: Thank you.
Peter McCormack: You're an amazing asset to Bitcoin and anything we can ever do to help, you have our details, you reach out to us. Where do you want to send people?
Calle: I'll send them over to my Twitter, it's @callebtc, and send them over to cashu.space; if you're a developer, please look at it. If you want to support Cashu, then go there, you'll find ways to support it and just look into Bitcoin. Please help us build Bitcoin, be a developer in Bitcoin, look into how you can contribute to open-source projects.
Peter McCormack: Thank you, brother. You take care.
Calle: Thank you too.
