WBD710 Audio Transcription

WBD Live in Sydney Pt 1 with Checkmate, Daniel Roberts & Rusty Russell

Release date: Saturday 16th September

Note: the following is a transcription of my interview with Checkmate, Daniel Roberts & Rusty Russell. 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.

On September 9th What Bitcoin Did hosted a live show in Sydney. In this first of two podcasts, the guests were the Co-Founder & Co-CEO of Iris Energy Daniel Roberts, Glassnode’s Lead On-chain Analyst Checkmate, & the Open Software & Lightning Network developer Rusty Russell. Across these interviews, we discussed various topics related to Bitcoin, including its meaning, its impact on people's thinking, and its potential to reshape the financial system.

“People understand what they own now, why they own it; you don’t step in front of the FTX train without knowing why the hell you’re here, that takes serious guts and serious conviction.”

— Checkmate

Interview Transcription

Peter McCormack: All right, Checkmate everyone.  So, we sat down earlier today to have a conversation about on-chain data and then we just talked about other stuff.

Checkmate: About everything except on-chain data!

Peter McCormack: Yeah, about everything else!  But we ended on a question where I didn't feel like we had enough time to get the answer, because Danny was saying, "Shut up, we need to finish and we need to go".

Danny Knowles: "We have other things on today".

Peter McCormack: Yeah, and so we finished up and I don't feel like we got a good answer and I also felt we could have gone on for another five hours.  So, I think that's a good starting point.  So, we spent about two hours just talking about life and the difficulties of life and money and all the things we go through.  And I ended up by saying to you, you get exposed to probably more and more various sorts of information than almost anyone.  For those who don't know, Checkmate works at Glassnode, incredible on-chain data.  So, you get exposed to all of that and the opinions on how that forms you live on Bitcoin Twitter, so you see all the various narratives from the different subcultures of Bitcoin and the different memes and you also listen to podcasts, and you're generally exposed to news and corporate media.  And so I asked you, so with everything that you see, what does Bitcoin mean to you?  So, I think that's our starting point.

Checkmate: Yeah, and just something I found interesting just by watching you two, Danny, you are Peter's short-term memory.  You'll come up with a particular point.

Danny Knowles: That's very true!

Checkmate: You're his ram.  You just pick up those bits and pieces.  Yeah, no, I think Bitcoin, I mean to me, it comes back to the question, what is money, right?  Money to me is time, it is our time, and your time is the most scarce asset that you have.

Danny Knowles: Just bring your mic up a little bit closer.

Checkmate: You can't get any more of it, right?  Time is all you have and it's constrained, and none of us can cheat it.  So, it's very similar and used in that regard.  So, what does Bitcoin mean to me?  I mean, to me, it's hope.  It gives people an opportunity to think in a different way.  I think most people in this room who've spent much time with Bitcoin have probably rethought their perspective on a lot of things, probably at a personal level, at an economic level, at a societal level.  You've probably rethought the way the world actually works.  You've probably looked under some covers that you never thought to look under before and it reshapes the way that you think about things.

In many ways, the meme, "Bitcoin changes you, you don't change Bitcoin" is just as true as it gets.  So, to me it's a forcing function for thinking differently.  I said this at the Bitcoin Live Conference, "I look forward to the day our politicians own some, because they might start thinking more than two weeks into the future.  It's that low time preference and I think it's really important for society long-term, and you're not going to see this play out in the short term, right, this is not going to happen in weeks or months, it's not even going to happen in years.  It's probably not going to happen in decades, but it's very much a tree that you're probably not going to sit under in many ways, but the long-term result of people thinking about money, about time, about how systems work, about how the economy works, what their role is in the whole thing, I think it's really powerful.  And to me, it's about change, and it's not because Bitcoin changes, it's because we change around Bitcoin.  So, really to me, it's hope and it's a signal for change. 

Danny Knowles: What is it about Bitcoin that changes you? 

Checkmate: I think, I mean for me, my background is civil engineering.  And before that, even though my lecturer told me in the start of the class, first lecture I ever had, "30% of you are going to end up in finance", and my internal brain goes, "Like I need fucking finance", and now I'm in finance.  And essentially, why does that happen?  Because as an engineer, you look at the world from a problem-solving perspective.  And what you're told is not how it actually works.  And Bitcoin defies logic to the traditional economists.  They simply can't get their head around it, because it's almost too elegant and too simple for people to understand.  It's too simple, and yet it works.  And we were talking about this today.  I think at the end of the day, Bitcoin makes one assumption and one assumption only, that human beings are greedy, and we're going to operate in our own best interest.  And the great irony of making that -- I mean, greed is good.  That went through a whole phase where people go, "Is greed really good?  Look at all the havoc that that wreaks on the world".  But it's like, "Yeah, but that's the constant.  It doesn't matter what you do, you're never going to stop human beings from being greedy". 

So, because Bitcoin bases its sole assumption on people being greedy, and that's what makes it work, in a very ironic way, it delivers the exact opposite result, that people start thinking about other people, people start thinking about their impact on society, they think long term.  When you think long term, generally speaking, you're thinking about, yeah, yourself, but there's other people involved.  The longer the time goes on, the more other people are involved in your life, the more people you intersect with, and I think that Bitcoin forces you to have that lower time preference.  The scarcity is a part of it, but I think the scarcity is just the first thing that gets you interested.  It's the thing that people gravitate to initially, but then you think about energy.  I didn't expect to be thinking about energy grids, I didn't expect to learn about nuclear power.  These things come because it forces you to rethink your current model.  And I think it comes back to, it's based on greed.  And for whatever reason, that base assumption makes it work in a very elegant way and forces you to rethink everything around you.  Peter McCormack: You talked about Bitcoin acts around the fact that we are greedy and around our self-interest, and I think we all understand that the majority of the problems we have around money at the moment is down to politicians and central banks, who also act greedy and with a self-interest.  So, what is the function change that we have with Bitcoin that's different?

Checkmate: I think it really comes down to the Cantillon effect.  None of us can -- I mean I'd love to be able to print money but we can't.  And the problem is, the fiat result is that when they do print the money, it goes to a privileged class, very few of us in this room probably have any access to it, and it's just an unfair system.  And you know, there's degrees.  If you go to Argentina for example, it's just way worse.  It's a much worse level of dilution and we talked about this today.  I actually like this framing.  If money is time and time is our scarcest resource, in Australia you may pay between 30% and 50% of your income is just straight tax.  Add on GST and a couple of other things, all the fines and all the other taxes and things, you're probably talking about somewhere between 50% and 60%.  So, that means you've got to work to July.  July is the government's and then everything from July to the end of the year is yours.  That's a pretty shitty go. 

But then on top of that, they're also going to take as much as they want via the inflation tax.  And why they do it, it's not like this is just by happenstance, they write about this, and by they, I mean central bankers and the like, they tell you that, "The inflation tax is hard for us to reason about.  So, whatever's left in that time, we're going to take a chunk more of it.  Oh, and by the way, you earned some interest in your bank account for offsetting that inflation, yeah, we're going to tax that as income too".  So, you look at this whole thing, it's just really unfair.  And it doesn't dilute everyone equally.  How do you have politicians who run a $200,000 salary, worked that for 10 years, oh, $200 million.  The math doesn't make sense.  Where does that money come from?  We talked about this today.  Where does the value go; where the hell does it go?  Because it's not going to us.  It's definitely not going to us.  So, I think that perspective is, it's unfair, we all feel it, it's slow and insidious.  But we're hitting a point in society where more people go, "All right, I've kind of had enough now.  This now, it's actually bullshit.  And I'm calling bullshit".  

I think Bitcoin to me, it's actually a protest vote.  50% of what I buy Bitcoin for is because of the scarcity and the savings and all that.  But the other half it's just a real big fuck you.

Peter McCormack: I think we agreed today that we think we work to the middle of October for the government; that's where we came to when you look at every single tax that we pay, after we paid our income tax and corporation tax.  And then after 31 December, once we're dead, they still take some more.  So, I think it was middle of October that we figured it out.  I'm really interested in that point where you talk about, you keep saying you plant a tree that you don't get to sit under, which is a really nice way of framing some of the ideas around Bitcoin.  We sat down with my good friend, Michael Dunworth, who's here, we've got a show coming out with him soon, it's absolutely insane.  I can't wait for you to hear it.  But also the interesting thing is, Michael is running projects at the moment which he won't see the result of.  And so, not only are bitcoiners thinking about their time preference, but also thinking about the time preference for future generations, which is a very unselfish act, because this isn't just your own children, this is the wider world.  Why do you think that has been built into the soul of bitcoiners?

Checkmate: Yeah, that's a good question.  I mean, it's about why does it lower the time preference.  I mean, I can speak to me.  I was attracted to Bitcoin because of that.  I was always a saver rather than a spender, that was always my general position on things.  But why did I get into civil engineering?  It's because I felt like I was building something.  I mean, bridges are around for a hundred years, we design these things for a hundred years, and that is meaningful.  If you can put your mind towards building something that's going to be around long after you, and deliver outsized returns, what's the value of the Sydney Harbour Bridge?  You can't even calculate it, the amount of time that that saved us. 

In fact, there was someone who did a study, I forget what bridge it was, but they looked at what happens if it didn't exist and just the sheer man hours of people driving around that same distance; it was astronomical,  we're talking huge amounts of time that's wasted.  And again, time is our scarcest resource.  We build things for the long term, make us more productive, make us more successful, allow us to influence others in a much more constructive way.  In many ways, one of the big drivers for me with on-chain data is, there's a lot of people, in fact, there's videos of people in Vietnam teaching the newsletter that we publish every week, like in a physical classroom.  And where do these analysts go?  Well, they have absolutely no chance of getting a job in Vietnam, but they apply as quants somewhere in the US and they get hired by Goldman Sachs.  And to me, sure, I mean Goldman Sachs probably isn't the right place to be if you want to change the world, but in many ways it's giving them an opportunity that they never had, and that's meaningful. 

So, in many ways, the same drive with, you know, why did I become a civil engineer is the same reason I do on-chain analysis.  It's because it helps a very wide cross-section of people.  And when I talk to people at the Bitcoin Conference about why I do the videos for Glassnode, there's two types of people.  One is, "It's a technical instructional video".  Every single week, you get to learn some of the different tools, how we're using them, how we think about them.  Some people take that, apply it to their trading, great.  There's another group who just want to know what the hell happened.  And that's really useful, because if you understand why the market does what it does, then you can survive, you can hodl through it.  The volatility is less scary when you know why.  And to me, again, it's about that narrative, but it's a data-driven perspective on the narrative.  So, I'm pretty sure I deviated from the original question, but carry on. 

Danny Knowles: I think that leads us into an important question.  Is on-chain analysis just voodoo astrology for men? 

Checkmate: Yeah, I mean it's the same as anything.  So, the root answer to that is, only if you don't do the proof of work.  If you just throw darts at a wall, you might hit some of them, but is that statistically significant?  Probably not.  There's a few analogies I use.  If you could see money flowing through the SWIFT system and you're a trader and you could see who's transferring it and when, you may not know exactly who they are, but you can see there's $1 million there, $10 million there, I mean sometimes I speak to quants and they're like, "I don't understand why this data has any value".  It's like, "You're telling me the asset that you're trading every day, solves blocks every ten minutes, has no value in the blocks being solved every ten minutes, what, are you mad?"  So, in many ways, it's about how you parse the data. 

Taking data, I mean data is useless.  You have to parse data into information and then you have to turn information into insight.  Some people use machines to do that, like quant strategies.  Some people use what I do, which is kind of understanding the mechanics of what's going on and using it to explain it in terms of narratives.  But at the end of the day, it's looking at the true reality.  Every single transaction, you have to pay a fee.  Nobody can put a transaction without paying a fee.  You have to spend some scarce resource or you have to do a hash if you're a miner.  So, in many ways, it's about how you apply those tools.  And with anything, I mean, how many technical analysis charts have you seen on Twitter?  I mean, there's not a lot of people saying, "Hey this is voodoo", it's because we've kind of become accustomed to the fact that some people are good at it, most people are not.  And that's the same with all forms of analysis. 

With on-chain, I mean it started in 2018.  We had first conceptions of it in 2011, but 2018 was kind of the first time real metrics developed.  I was saying today, when I started at Glassnode in February 2021, I was at the all-time high.  I'm given this toolbox of tools, and since then has been the learning journey.  And the analyst I was back then, night and day to now, because I've spent that market time studying it, analysing it, using our newsletter as a live-fire exercise.  The answer is that I was a shitty analyst when I started, but I'm a much better analyst now.  So, it's that progression, and in many ways, people have got to pioneer and learn this stuff first -- Willy's here -- exploring these things from first principles, the very, very origination of these metrics.  We didn't have them even six months ago, we didn't have most of the stuff we use today.  But now we're finding we're using the same tools over and over again, because we're finding they're useful. 

So, we've kind of passed, we've gone from data to information to insight.  And there's a lot of analysts who tried to use it and rightfully so, testing the boundaries, and got it wrong.  But we're at the phase now where we've actually distilled down what matters and the next cycle is going to be pretty interesting, because there'll be a lot of analysts who are still not very good, but there's some who've been paying attention the whole time.

Danny Knowles: So, are you saying definitively you're going to be right next cycle?

Checkmate: No, I'm saying I'm hoping by the next cycle I won't have to do it any more, because there'll be a bunch of other analysts who have come through and kick ass as well.

Peter McCormack: By the way, you might notice up on the screen, there's some live drawings happening.  This is from Jasper.  So, this is Jasper's business.  He does live drawings at weddings and Bitcoin events and things.  So, if you notice a weird drawing coming up, it's because there's a guy in the corner drawing.  So, just to tee up my next question, just to get an idea, how many people are going through their first Bitcoin cycle; who's in their first Bitcoin bear market?  Second one?  Third one?  Fourth one?  Fifth one?  It's a very interesting experience going through your first bear market.  It's a real test.  So, congratulations on still being here.  But we talked about this today, some of them still might sell. 

Just to kind of aid people who are experiencing this for the first time, because they have their conviction long term, but in the end it is still your personal money, what is the data telling you that would give some reassurance to these people? 

Checkmate: Yeah, so I think first things first, I mean I got in at the absolute top, tippity-top of the 2018 market.  So much so that Coinbase's fee for Australian dollars, my first buy is above the top tick.  You couldn't have bought higher than me.  And I was talking about it before, $10,000 to 10 cents, it happened.  And everything except Bitcoin, I learned that lesson.  But you have to experience a bear market, because a bear market is where you build the conviction.  So, 2018 is where I learned about markets, started to learn about on-chain.  2019 is when I really started to actually apply these skills and kind of observe your own mentality through a bear market.  And it's really important, because bear markets give you life skills that you simply can't get.  I've never seen anything quite like it, but you just can't develop these skills elsewhere, because it is a hardship.  I mean, 85% down, 70% down is brutal, and markets are non-linear.  You can't actually wrap your head around why things happen.  The media is always trying to give you reasons why, but there's forces that happen underneath the surface that when you look at the data, it helps you understand why these things are happening. 

At the end of the day, it's all about supply and demand.  When price goes down, there's literally more supply than demand, for whatever reason.  In the short term, markets are a voting machine, which means they go wherever the narrative wants them to go.  But over the long term, the fundamentals kick in and markets become a weighing machine, and that's where you actually start to see the long-term behaviour of these assets.  Now in terms of where we currently sit, the June sell-off, when Three Arrows have blown up, So, Luna's gone, Three Arrows have blown up, and we're down at $17,000, that is when the bottom started forming.  Now FTX was this funny kind of hyper -- I would call it a hyper extension, and you can see how Bart Simpson it really was, it kind of just reversed after January.  But textbook, absolute textbook Bitcoin bottom by every cycle. 

So, what does a textbook bottom mean?  Well, we look at people, I mean in every cycle, your first cycle you will buy high and you will sell low.  Everybody does it, it's just a part of the curve.  We actually have an article about this which talks about long-term holders for their first cycle, and they get rekt the whole way down.  But then they learn and if they stick around and you survive the bear market, you go, "Maybe this thing actually makes sense".  This is me in 2019 being like, "This thing actually makes a lot more sense".  You look at the macro environment, it starts to kind of click into place.  And you're like, "Why the hell am I buying this thing?  It's just red".  But eventually, the thesis plays out because the weighing machine kicks in. 

So, that bottom we hit in June, all the way through FTX, we saw two things which are really important.  We look at profit and loss.  We look at every coin; where did you acquire it on-chain, where was your DCA level and where are we currently?  That's your unrealised profit or loss.  So, most people can't carry a large unrealised loss.  The hodlers can and they get to a certain threshold and we hit that almost exact level to the dot.  The amount of pain that the hodlers can handle, because we're price insensitive.  There's a whole other side, which is how many people actually flushed out.  And what we saw is a statistically significant, an absolutely enormous amount of coins got flushed out at a huge loss.  And when we look at who they were, they were people who bought the top.  They weathered the entire bear market all the way to the bottom, and they sold the bottom tick. 

This is what happens every single cycle.  We can see it play out cycle after cycle in all data, and I guarantee you, if we had ten-minute block confirmations for gold or for Apple stock or for any asset, I guarantee you would see the exact same thing.  The difference is that Bitcoin is large enough, important enough, and significant enough, and we have the data that we can visualise these things.  In fact, there's a report we're going to release very soon where I use famous market quotes to help explain what the data is showing you, because it just happens.  Human beings are exactly the same.  Our brain has not changed.  It's 1,000-year-old hardware operating in 2023.  So, the interest rates change, Jay Powell's mood change, all this stuff changes, but our ape brain and our response to fear and greed never changes.  And that is why you can look at price charts from the 1920s and see the exact same patterns as today.

Peter McCormack: So, that's the miserable data!

Checkmate: Well, that says that the bottom's in.

Peter McCormack: Has anyone sold the bottom?  That's why you're here.  Okay, so that's the kind of misery side, but also look, we talked about it today.  You talked about some of the promising data.  You talked about the shrimps.  Explain to people the promising data that you're seeing at Glassnode. 

Checkmate: Yeah, absolutely.  So, we talk about it on the podcast, obviously, but there's one chart that's my favourite chart at the moment.  What we look at is when we talk about shrimp, right, that's people with under 1 Bitcoin.  Now that's also going to capture everyone who's like, if you're DCAing less than 1 Bitcoin every month, every single one of those is going to be a shrimp.  So, it's just a big amorphous blob of normal people.  We're not talking about Goldman Sachs, we're not talking about trading firms, we're talking about the average person. 

When we look at the 30-day change, and this is a great metric, because a 30-day change just gives you a monthly, what's the gradient, what's the delta of how much their balance is going up or down.  And what we saw is that we had an enormous spike, just an absolutely gargantuan spike at the absolute top of 2017.  People bought the top, the average person bought the top in 2017 in the largest size we'd ever seen.  It blew everything out of the water.  And then it died off and it didn't come back to that level for five years until we came back to $20,000 from the other direction in a bear market.  So, instead of buying the rip in 2017 at $20,000, they bought the dip and they sustained that level.  Then FTX blew up and went lower and they bought even more.  Then we saw the rally in the start of 2023 and they bought even more and then sustained levels that are roughly about equal to the issuance.  So, every coin that is mined, the shrimp are buying all of it. 

So, what does that tell you?  I mean, that is five years where they bought the rip and now they're buying the dip in the nastiest -- I mean, FTX, Bitcoin was dead for ten years, that was the narrative, "Bitcoin's over, it's finished".  And yet these crazy people are buying this thing in the largest size we've ever seen.  To me, that is the chart that shows Bitcoin education, the success of it over the last five years.  People understand what they own now, why they own it.  You don't step in front of the FTX train without knowing why the hell you're here.  That takes serious guts and serious conviction.  So, in my view, when you can't understand the data, if you're a price-insensitive hodler, and I ask this to myself all the time, when I can't read the data, do I see myself in the data?  And I bought on every single one of those peaks because my conviction is like, I mean $16,000, I'll take it all day.  So, do I see myself in there?  And I saw myself throughout the data.  So, to me, it shows that that education has really kicked in and people get it, people really do understand it.

Danny Knowles: I hate to be the pedantic one, but is that people buying or is that every single exchange in the world rugging them so people are going on-chain and showing up for the first time?

Checkmate: It's both.  So, in a lot of this stuff, it basically is people taking self-custody.  Now, during the FTX event, that's clearly going to be people who are panicking left, right and centre, "Just give me my coins off exchanges".  But if you're looking at small holders, shrimp, crabs, these kind of smaller entities, if you think about it, most of us, I would guesstimate that there'd probably be a handful of DCA patterns.  We all do kind of the same thing.  Some of us will buy and you'll wait for a certain amount, then you withdraw.  Some will buy and just immediately withdraw.  Some will go to Lightning, but there'll be a couple of heuristic patterns. 

When you move up to the whales, there's a lot less of them.  They're a different trading firm, they've got different strategies, they're much more bespoke.  There's less of them, more money, but they're much more bespoke.  But the small holders are kind of this big amorphous blob.  You really don't have to think too hard about it.  A lot of the nuance actually is quite irrelevant on a relative basis.  So, the answer is it's there, but it's in the 5% of stuff that doesn't really matter.  You can dig into it, but for what benefit?  So, that's where I'd put that kind of general topic.

Danny Knowles: So, you think the people withdrawing from exchange at that time would only make up 5% of that data?

Checkmate: No, what I'm saying is that the people who are edge cases that would kind of defy the logic of small holders taking custody, anyone outside that bucket would be a small chunk of that, yeah, so the other way around, 95%.

Peter McCormack: This has come to a close very quickly, these things always do.  But are there any kind of final points you want to make, anything you want to leave the people of the room here?

Checkmate: I think the number one thing is when you're looking at on-chain data, it's, I mean, what is Bitcoin?  It's a collection of our, all of our decisions.  The valuable part of Bitcoin is not actually really the blocks or the proof of work or any of that, it's the UTXO set.  The UTXO set is all of our decisions.  It's all of the collective ownership and all of that changing of hands.  That's what matters.  And if for any reason, proof of work didn't work, we would have to find another way to port the UTXO set to somewhere that can carry that forward safely.  Because the UTXO set is the decisions that have been made, the immutable decisions.  And on-chain data is just the analysis of those studies.  And we're going to go back and look at this and have studies of behavioural economics that comes out of just looking at the Bitcoin blockchain.

Danny Knowles: Is it basically another way of looking at the efficient market hypothesis?

Checkmate: It is, and we've done studies that looks at the mining side of the equation.  We've shown that by our best estimate, 50% of days they're in profit, 50% they're in loss, which is just a perfect smack in the middle market, which means the difficulty adjustment is a perfectly efficient market.

Peter McCormack: All right, round of applause for Checkmate.  Dan Roberts from Iris Energy. 

Danny Knowles: Thanks Dan. 

Peter McCormack: You got a mic?

Daniel Roberts: Yeah.

Peter McCormack: You've got a mic, okay.  Dan, good to see you, man.

Daniel Roberts: Good to see you too, Pete.

Peter McCormack: Thanks for having us in Australia.

Daniel Roberts: Mate, pleasure.  Thanks for coming.

Peter McCormack: Unlucky with the Swans last night.

Daniel Roberts: Very unlucky.  Our chairman's a Carlton supporter, who knocked us off, and first thing he said to me today.

Peter McCormack: Okay, wow, look, lots to get into.  Firstly, obviously, a big thanks to sponsoring everything we're doing, it's amazing, but we want to talk to you because there's so much going on in mining.  So, mining is something I totally ignored for the first four years I was in Bitcoin.  It was just this thing that happened in the background, I didn't pay any attention to it.  And I think the beautiful thing about mining is it essentially was this unknown unknown that's become kind of half the industry now.  It's half the things that people talk about, it's started to dominate the things we talk about.  So, there's a whole bunch of things I want to ask you about, but can you just give everybody in the room kind of an update on the state of mining, the state of Bitcoin Mining? 

Daniel Roberts: Yeah, sure, and I think our experience was similar, to be honest.  The first four years, we didn't know anything about mining either, and in fact when we tried to learn about it, no one had talked to you about it.  And you're trying to figure it out, and it was abstract, it was complex.  People I think that knew what they were doing didn't want to talk about it, but eventually I think we figured part of it out at least.  And it comes from I think our backgrounds where we're in finance, we're in renewable energy, we're in commodities, we went down the Bitcoin rabbit hole, everything Checkmate just said, I can't believe you scheduled me after him, anyway, he was fantastic.  Everything he said resonated and, yeah, mining was a really good way to go and support the industry.  And I think it's grown, it's evolving.  Every year that goes by, the mining industry is evolving, live time.  We're seeing a variety of different business models, a variety of different ways of mining, the underlying energy source, the way it's interfacing with energy markets, the renewable element, etc.

Peter McCormack:  People talk about Bitcoin as like this organism that you can't stop, it just becomes what it becomes.  And I'm going to bring up Michael Dunworth again, but I had another great thing said today that Checkmate mentioned, that Mike Dunworth said that was, "Nobody's got the technology in the world to stop Bitcoin", which you really have to stop and have a think about.  But behind that, Bitcoin has become so much more, Bitcoin Mining has become so much more than just finding blocks, and you put out this incredible tweet this week where you discussed the efficiency of the Bitcoin you're mining now, because of the work you've been doing in Texas.  A lot of people hear about this grid stabilisation, but can you talk to that; can you talk about the work you guys have been doing there?

Daniel Roberts: Yeah, look, absolutely, and I mentioned it before, our backgrounds are in energy markets.  You know, I worked in London back in the mid-2000s chasing feeding tariffs for wind and solar around Europe.  And none of us at that point foresaw the issues in grid stability that this intermittent renewables would cause, and now we're trying to fix it, batteries, pump storage, hydro, etc, and that's fundamentally the opportunity for Bitcoin Mining.  Because what you've got is, any other manufacturing industry, you're manufacturing widgets, but it's at a nanometre level, and it is incredibly flexible.  So, you can dial it up, you can dial it down.  And this isn't just, we can walk up and turn off the computer and then an hour later, we can walk back and turn it on.  Our guys have developed an algorithm which automatically looks at market pricing for power, looks at market pricing for securing electrons into the Bitcoin Network, and basically dictates the difference. 

So, we'll enter into a hedge, so we'll have a right to procure power, and very simply, every 15 minutes there's a market rate of power.  If the market price of power is higher than what we can earn flowing those electrons through our computers and securing the Bitcoin Network, then we'll give it back to the market.  So, it's a fantastic demand-side battery and a way of stabilising those grids.  So, when the wind blows, the sun shines, there's plenty of negative power prices, we'll come to in a second; then we'll just mop that power up all day long, run it through the machines, generate Bitcoin and have a really good business.  But equally when power prices peak and they go up and they go up to $1,000, $5,000 a MWh, we don't need the power.  We just have an algorithm, puts the miners automatically to sleep, and sells that power back to the market. 

So, the tweet you're alluding to was last month, for the month of August, Iris Energy in Texas mined at an average cost per Bitcoin of -$28,000 a coin.

Peter McCormack: Which if you can't do the maths, that means they mined at $56,000 a Bitcoin. 

Daniel Roberts: It's coincidental that it ended up two times, yeah.

Peter McCormack: It's a weird thing.

Danny Knowles: I think you should go into that and explain precisely how that's possible.

Daniel Roberts: So, really simply, we had a hedge on like most miners will.  We don't want to take market risk, it's not our business.  We had a right to procure power.  We then have the algorithm that every 15 minutes looks at mining profitability into the Bitcoin network, or mining profitability by selling that electron back into the market.  There are a lot of weather events, there are a lot of outages, you know, one $5,000 a MWh.  So, very simply, the power that we had procured -- and remember, entering into a hedge contract in these markets is what underwrites new generation as well.  By us playing in this space, entering into forward hedges, we give the market certainty that there are buyers of power.

Danny Knowles: So, you can promise your buying power over the next 12, 24, 36 months?

Daniel Roberts: Exactly, whatever it might be.  And at the end of the day, energy markets, like most markets, are a melting pot, different generation sources, different loads, etc.  And if you've got a bunch of industrial loads entering into a variety of hedges, whether it's for one month, three years, etc, that sends strong signals out into the market to potential builders of new generation, wind, solar, whatever it might be, to say, "There is an underlying market here, I should build that power".  So, first of all, we do that; price signal, incentivise renewables to be built out. 

Second of all, when there is an issue to do with the intermittency of it, we don't need the power.  We can say, "I know that we said we'd buy that power from you and we'd take it off the market, but you know what, the market's short, someone else values that power more than we do, so let's sell it back".  So, what we ended up doing was basically sleeping our machines, call it 15%, 20% of the month, during these high-priced time periods.  So, we only mined about, what, 80%, 85% of the Bitcoin we would normally, which corresponds with most other listed mining companies, coincidentally, which we can go into.  But in exchange for that, we sold so much power in terms of the revenue for that 15% to 20% that it not only covered our power bill, but it covered it well and truly almost again to lower that cost per Bitcoin mined.

Peter McCormack: So, you love a Texas cold snap!

Daniel Roberts: Volatility is good! 

Peter McCormack: Volatility is good!  Marshall Long, you will know Marshall as my friend from out in Houston, he's a miner as well.  Interestingly, he texted me the other day, and he said, "Texas Bitcoin miners literally stopped a blackout in Texas".  Is he being hyperbolic, is it true, miners are actually stopping blackouts?  Because there is a certain cohort of people within places like Texas, whether they're politically-led or news-led, they want to imply that billionaires are milking these networks to mine Bitcoin.  But there's various other people, such as Pierre Rochard, who's doing the important work to explain the work that miners are doing to stabilise the grid.  So, was Marshall being accurate with that?

Daniel Roberts: I think you've got to be careful not to fall for these headlines.  Like both sides, everyone loves a catchy headline, make it sensationalist, get the clicks, etc.  The truth, as most of us would know in this room, sits generally somewhere in between.  And I'd like to think that we look in between.  At the end of the day, you've got a functioning market.  You've got buyers and sellers of electricity.  The market determines a price for that electricity every 15 minutes.  If the efficient market results in people that don't need power giving power back to the market during times of under supply, to supply people that really do want that power, which they're expressing through the price signal and bidding up the price of power, then that feels like an efficient market that goes to solve problems. 

No one is giving handouts to miners.  Equally, miners are not doing this altruistically out of the goodness of their hearts, just giving power back to people.  The truth is somewhere in between.  There is an economic incentive for us as miners to sign a hedge contract to procure access to power, so we have security of power at a fixed price, just as there is an economic incentive for us to give that power back to the market when someone else values it more.

Peter McCormack: Okay, maybe a better question is then, what would happen if the Bitcoin miners weren't in Texas; what would the implication be?

Daniel Roberts: So, the Bitcoin miners, I think to address the rolling blackout question more specifically, if they weren't there and instead you had a more traditional manufacturing industry, call it aluminium smelting, call it any widget manufacturing, that didn't have the flexibility, and most don't, then you would have bigger problems, absolutely, because you have got this reserve capacity from a manufacturing industry called Bitcoin Mining, which at the click of your fingers can turn the power off and give it back to the market.

Peter McCormack: Right, okay.  Just moving on from there, I also know that Iris has purchased a whole bunch of new machines to look into AI, which is obviously a whole new area.  It's an innovative area for a miner to be moving in.  How much can you talk about that? 

Daniel Roberts: Lots. 

Peter McCormack: Great! 

Daniel Roberts: Yeah, let's do it.  So, I think like the tag, "Bitcoin miners" also, it's a label as well.  At the end of the day, we're not Bitcoin miners, we're infrastructure people.  We've gone into the data centre industry, we're building next generation computing infrastructure.  And what we're really good at, and when I mentioned the sectors evolving earlier, this is a reference to, gone are the days of old shipping containers, C-cans, old, abandoned warehouses, mopping up a few megawatts of power.  We are moving into the industrial era for Bitcoin Mining and next-gen compute. 

So, what we've become very good at across our four sites is operating energy-dense compute.  That's all these ASICs are.  So, Bitcoin Mining, let's get rid of that label; we are operating computers, and these computers consume a lot of electricity, as we know, and they consume a lot of electricity in a small area of space.  So, when you look at traditional capital city data centres, they're optimised for very different things.  It's not the same energy density.  It's live-time cloud computing, really low latency, bank systems, accessing your shared drive at work, etc, and latency is critical. 

But when it comes to energy-dense compute, things like ASICs to secure Bitcoin Mining, things like GPUs to run this generative AI, that requires a completely different type of facility.  It requires you to manage extreme amounts of heat and airflow and density of the compute, which is what we've spent the last five years doing.  So, for us, it's not a big leap to say, "Well, instead of tomorrow we're going to go and buy some more ASICs from Bitmain", we just went to NVIDIA and bought some GPUs.  So, today I spent the day on Midjourney learning all about a new sector, it's exciting.  Luckily, our CTO knows a bit more than me, but we're absolutely pumped, it's exciting.

Peter McCormack: Sorry, go on.

Danny Knowles: So, how do you see the future of Iris; how are you going to divvy this up?  Is Bitcoin Mining going to remain the main focus, or is it going to expand into other industries?

Daniel Roberts: We've always said we'd expand and there's a number of investors in the room, and appreciate all the support, that saw our investment decks from four or five years ago that mentioned high-performance computing and GPUs.  This is nothing new.  We've known the trajectory that we're on and what we can do is next-generation compute.  ASICs make a lot of sense because you've got an objective price signal, you can mathematically see the profitability.  You mine a Bitcoin at X, sometimes down here, and you sell it for Y.  It's really easy to model out, get your head around, you buy a commodity from the likes of Bitmain, etc, plug it in, you've got your P&L. 

For something like a GPU, well, what's the end product that you're producing?  You're producing a large language model, you're producing a Midjourney image.  There's a few more steps to go through, so it's a bit more nuanced, there's a bit more uncertainty around the end product.  So for us, the opportunity cost of going outside of Bitcoin Mining has been high, until what we've seen over the last six, nine months, which is this explosion in the generative AI space.  It feels like we've hit that inflection point where the market's now started to take off and we can start to look to service it.

Danny Knowles: Is that because it's another industry that's at least partially sort of location agnostic?

Daniel Roberts: It is, absolutely.  So, we mentioned latency earlier in the context of the capital city data centres.  If you're on Midjourney, and Pete, I was on it today after you showed me last night, you don't really care if you're waiting 5 seconds, 10 seconds, 15 seconds for an image to load after you've given it a bunch of prompts.  But if you're at work and you're trying to open a shared document with someone sitting next to you to collaborate on something, and every time you make a change, you've got to wait 15 seconds, that's going to be an issue.  So latency, absolutely, is something you don't need.  In saying that, I think we've still got 20, 30 millisecond latency into our facilities, so I am exaggerating it. 

I think where you see these capital city data centres, they will spend an order of magnitude more on CapEx.  They will have to acquire capital city real estate.  They are beholden to capital city power prices.  And it's the heart of what we're doing is going to the source of low cost, renewable energy.  We are 100% renewable, but we go to the source of where that energy is, build the next-gen compute facilities, and then to your point, we can just ship that over high-bandwidth telecommunication cables anywhere in the world that it's required.

Peter McCormack: So, at the time where the AI industry is exploding, they're in this fortunate position that there's already these companies like yourself who are already well positioned to service their needs, because you have the type of data centre that can provide for their needs?

Daniel Roberts: That's right, and traditional data centres can do it as well.  But our experience and conversations in the industry, they don't like it because these guys are painful.  They set up their rack space in a more energy-dense way, they don't use as much floor space in the data centre.  When we can pay a tenth of the CapEx for these facilities and give a very similar service, we'd like to think it's going to be quite compelling.

Peter McCormack: Can you put your hand up if you haven't played with any form of AI?  Like three people.  Put your hand up if you haven't played with Midjourney.  Okay, so that's quite interesting.  I assume everyone who didn't put their hand up to begin with has played with ChatGPT, which is an incredible, very incredible tool.  So, I think a lot of you, once you start playing with Midjourney, you will stay playing with Midjourney.  I've been using it constantly.  It can generate graphics in five minutes that would take days for a graphic designer to do.  And I think those hand signals, for me, indicate the size of the opportunity with AI, because there are a lot of people who haven't even discovered the power of it.  We played with it last night, you've been playing with it to today.  What have you been generating?

Daniel Roberts: Also a couple of Twitter images, that was good.  Floating around maybe some new logos for Iris. 

Peter McCormack: It's quite addictive, isn't it?

Daniel Roberts: It is, really.  I actually find the prompts harder on Midjourney to navigate than I do on something like ChatGPT.  The interpretation of the English language input that I put, or maybe it's just me, it doesn't translate as easily to the output of the image.

Peter McCormack: Yeah, it takes a bit of time.  I mean, I've been playing with it for a couple of months, you start to get used to it, but I still think those hand indications are interesting, because these are people also you would expect to be at the forefront to maybe have tried these things out.  So, there's going to be a huge opportunity with AI; if you guys figure that out the way you figured out ASICs, that's a huge opportunity for the business.  But what does that do for the structure of business like you are, because Bitcoin was super-volatile, up and down; does this almost flatten the curve, does this change the business model?

Daniel Roberts: It goes to the opportunity cost of every incremental dollar that we raise into the business and we deploy.  So, we're building data centres; now, the next dollar that we raise, do we go and buy a GPU for supporting Midjourney or do we go and buy a Bitcoin Mining ASIC?  And there's no simple answer for that, it's time and place.  We see the returns on ASICs fluctuate throughout the cycles.  There are good times to buy ASICs and there are times where you go, "You know what, it's looking a bit overshot, the return on it, you're having to wait a long time, it's probably not a great time".  Equally with GPUs, there's a whole assortment of business models that we could entertain, everything from just offering rack space for other people to bring in their own GPUs, through to us owning and operating the GPUs and providing a software level service.  Ultimately, it goes to the return on capital, and the return on capital also then needs to be linked to how we access that capital. 

I think it's no secret that multiples in Bitcoin Mining are lower than multiples in more traditional data centre businesses.  So, if that can enhance returns for our shareholders by diversifying the business, then absolutely that's something we'll do.

Danny Knowles: So, on that, recently hash price has been getting absolutely crushed.  What is this to sort of look forward to as a Bitcoin miner? 

Daniel Roberts: It's kind of funny, the halving is something that you wouldn't say you look forward to, but it's going to be interesting. 

Danny Knowles: I would!

Daniel Roberts: Yeah, exactly, that's the thing.  So, I think looking at where we now sit with Texas and the ability to drive our power prices negative, the halving comes, notionally the number of Bitcoin you receive probably goes, what, 40%, 45% down once you adjust for transaction fees.  But equally, there's this whole game theory that emerges at that point.  So, if you reference back to Checkmate's comments around maybe half the miners are in profit, half the miners are in loss at any point in time, when that halving occurs, there's going to be a huge chunk of the mining market that are uneconomic. 

You look at hosting prices, so we paid -8 cents last month in Texas.  Hosting rates in this industry, you can go online, they might be 6.5 to 8.5 cents.  Right now, that's probably not profitable at the current hash price.  The halving comes, it's even less profitable.  So, what you see is a big chunk of the hashrate withdraw from the market whenever you see a halving event, or you see a big price decline in the commodity of Bitcoin, because what happened is that bounty every ten minutes, the amount of Bitcoin has reduced in fiat terms.  And all of a sudden, if you can't afford your power bill, you're having to switch off.  So, the hashrate goes down, difficulty adjusts.  If you're a low-cost miner, you've actually got a natural hedge because you've got a volume hedge. 

Let's say the network hashrate goes from 400 exahash down to 200 exahash post-halving, so half the miners have shut down.  If we're a survivor, because we're a low-cost miner, we receive twice the number of Bitcoin that we did prior to that.  If we've got 2 exahash or 5 exahash, let's call it 4 exahash out of 100, nice easy number, 1%.  If it goes down to 200 exahash global hash rate, and we've still got our 4 exahash, now we've got 2% of the global network, we get 2% of the Bitcoin every ten minutes, our cost per Bitcoin produced has halved.  So again, it's one of those natural evolutions in the market where market forces favour the people that think long term, secure that long-term cheap power; and if you're a speculator, you get in on the way up, you've got expensive power, you'll make good money for a while, and then you've got to be prepared to switch off for a while.

Danny Knowles: So, do you think that that means that miners will consolidate after the halving; or do you think there'll be like a standoff where people may be mining unprofitably and just hoping someone else falls before them?

Daniel Roberts: Absolutely the latter.  We've seen it, yeah, it definitely happens.  And look, it's not necessarily a standoff because they're unprofitable; you'll have energy hedges rolling off, you'll have things like demand charges, where a chunk of your power bill you have to pay regardless of how much power you utilise during that month.  So, there's always a delay, there's always a standoff that occurs.  So, it could take three months, could take six months, but ultimately people aren't going to continue.  There's a limit to how deep people's pockets are, right, and how they're going to continue to mine at a loss.

Peter McCormack: Well, the thing that's impressed us so much about Iris since we've been working with you is how innovative you are.  We see the charts comparing all the public miners, we see all the different metrics on how they perform, and we see Iris basically kicking everyone's ass.  I'm happy to say that, it's pretty true.  They might be a sponsor, but you can see the charts yourself.  But what it appears to me now is as we are approaching the halving, the innovative work that Iris has done, focused not only on your operational efficiencies, the efficiencies of your mining operations, your work that you've done in Texas, where you've created your own doubling, and the work you're now looking at AI, you've essentially streamlined yourself going into this halving, so you're ultra-competitive when this standoff might appear.  So, it feels like you're pretty robustly set up for this.

Daniel Roberts: Look, I think it goes to two individuals.  Myself and Will, we set up the business, we own a lot of shares, we've got voting rights and we believe in what we're doing.  We're not here for a quick buck, we are thinking multi-decade and I don't say that lightly.  We are pre-empting issues.  This is our nest egg, this is our baby, we've got all our wealth in it.  We don't want to be rug pulled in two years because we've tried to take a shortcut, we've tried to time a market peak and we miss it.  At the end of the day, we believe in Bitcoin, we believe in things like generative AI and this whole high-performance compute where everything is digitised, everything is going online, the amount of energy we consume as a species and going into the digital sphere is not going away, and we are at the very heart of what is happening there. 

We own the infrastructure, we own the energy, and developing the systems, the data centres, and the processes to manage that compute into the new age is something that we're pumped about.  And taking a shortcut for a short term, you know, the next couple of years, just doesn't make sense.

Danny Knowles: So, especially the last bull run, we have a lot of public miners, and I'm sure private miners, hold a lot of Bitcoin on their balance sheet, which is I know something you've never decided to do.  Is taking a directional stance on Bitcoin something you would ever look at doing?

Daniel Roberts: We're very directional on Bitcoin, we're very long.  It's really simple.  And in fact, if you hold Bitcoin on your balance sheet, arguably you're less long than someone like us who's liquidating it and reinvesting it into more Bitcoin Mining machines and compounding returns for shareholders.  My view is simple.  If you want Bitcoin, go and buy it.  Don't go and pay us to hold it on your behalf.  And we're probably going to have to use some centralised custodian, half of them get rug pulled so we're probably going to lose it.  I would far rather liquidate it, buy more Bitcoin Mining ASICs, compound those returns and generate an active return for our investors, that's probably going to outpace someone that keeps a static asset on their balance sheet.  I just don't get it. 

Peter McCormack: Well, hold on.  So, that's a very low time preference approach to building a mining company for the next few decades, whereas the others are actually restricting themselves.

Daniel Roberts: Look, if you tie up capital on your balance sheet, whether it's Bitcoin or US dollars, then it's non-productive asset.  And yes, I store most of my wealth in Bitcoin; well, I lost it on the boat that I bought, as you know.  But it just doesn't make sense to me, if you're running a large, industrial computing company, or whatever you want to call yourself as a Bitcoin miner, and you're hoarding a static asset on your balance sheet, how that generates superior returns for shareholders.  You don't see it in any other industry.  I don't think I've ever seen a gold miner stockpile gold on their balance sheet or a copper miner.  I don't understand.  If your specialisation is widget manufacturing, why would you keep widgets in a warehouse and wait for them to appreciate?

Peter McCormack: Yeah, that's a fair one.  Okay, so there's a number of large public Bitcoin miners out there.  As I said, if you just look at the charts, Iris seem to outperform every single one of them on every single measure.  Why do you think you've managed to do that; and when moon for Iris?

Daniel Roberts: Look, the market's the market.  I'll give you a bit of a diplomatic response here.  We are just generally focused on making the right decisions, looking out as far ahead as we can, minimising attack vectors and building a safe business, but also one that's got extraordinary upside within.  We've got 30 exahash of power capacity that we can expand at one single site.

Peter McCormack: Try and help people understand what 30 exahash means.  Does everybody understand what 30 exahash means, because I wouldn't have six months ago?

Danny Knowles: Go on, give us a scale of the size of the building, the operation; what is that?

Daniel Roberts: It's big.  So, the whole global network today is about 400 exahash and we've got power at one of our sites that can take our installed capacity, using the latest chips from Bitmain and the other Bitcoin Mining manufacturers, up to 30 exahash.  That's 760 MW of power.  I think for reference, Australia's annual energy consumption might be, DB help me out, I reckon it's probably about 3% or 4% of Australia's energy consumption.  So, it's a big number.

Peter McCormack: Well, it's impressive stuff.  We have to move on to the next interview.  Dan, obviously thanks again for supporting everything we do.  Please do go and check out Iris, it's important, they are our sponsor, so they deserve you to check them out.  Their work's incredible, they've been very supportive to me and Danny.  Big thanks to Dan Roberts.

Daniel Roberts: Thanks, Guys.

Peter McCormack: Rusty Russell!  Go and get a drink; everyone, go and get a drink!

Rusty Russell: Nothing is going to happen.  Okay, so I'm loving these drawings, so this is going to be a challenge.  I don't say anything interesting.

Peter McCormack: Go and get a drink, man, that's fine.  Is anyone just staring at the drawings the whole time?  That's what we thought might happen, we might stare at the drawings the whole time.  I think we'll make these available afterwards, won't we?

Danny Knowles: We will.  They'll be on the video, yeah.

Peter McCormack: Rusty, it's good to fucking meet you for the first time.

Rusty Russell: Okay, so before we get started, Checkmate said the basis of the whole thing is greed, and I cringed.  I'm like, "Oh my God".  So, we're going to get into this.  Before you ask a question, do you know why nobody figured out the proof-of-work scheme that Satoshi came up with prior to the 2009 whitepaper?  Because it's shit.  Okay, let me tell you.  Cryptographers like asymmetry.  So, you should be able to do something on your computer that would require the whole universe of computers working for the age of the universe times by ten for anyone to discover; it's asymmetry.  So, your defence should be so much stronger than your attack that it will always work.  That's what cryptographers love.  When they look at cryptography, they're like, "You can do it, it takes you ten nanoseconds, and it would take the age of the universe to someone to reverse it.  So, if you want to hack my private keys, you want to guess them, go wild.  We know the sun's going to explode before it ever happens", etc.  Great, okay.  That's what cryptographers like. 

The idea that you would build a scheme where literally you've got to keep the honest people 51% ahead of the dishonest people?  That is a shit scheme, that is awful.  Every cryptographer who would look at that would go, "No way.  It's vomitous".  The only problem is it's the only scheme that works.  So, it's kind of brilliant, but it's also kind of awful.  So, as somebody who was reading the whitepaper for the first time, I was like, "Oh, this is grossly clever.  It's disgusting, and I can see why nobody did it before, but it's a grotesque hack.  And that is the basis on which we're building all this stuff.  So now, I am way more comfortable with that now than I was when I first read it.  But when you find people who go, they dismiss Bitcoin, like technical people, like cryptographers and people who dismiss Bitcoin the first time, that's why, because they looked at it and went, "Wow, this is the worst trade-off ever". 

So, in some ways, Checkmate was right because you're like, "Well, how do we know that the honest people will stay ahead of the dishonest people to make the system work?"  And that comes down to much more fuzzy, kind of greed-like incentives.  Sure, but it's not fucking maths.  It's terrifying if you relied on the fact that maths secures all this stuff and it all works.  Anyway, I wanted to get that out of the way to start with, because he said that and I went, "No, that's not the worst thing about Bitcoin, that it relies on greed; the worst thing is that it relies on this really weird assumption that makes cryptographers really fucking leave the room in panic, okay?" 

Peter McCormack: Is it something though that as a concept, gets stronger the bigger the network is?

Rusty Russell: Oh, no, the attackers get bigger and the defenders get bigger, and you hope that the good guys win.  Seriously.  I mean, there are reasons to believe that this will work, but we're not there yet, it's still an experiment.  So, okay, there's three ages of Bitcoin.  There was the first age where it wasn't money.  It was just this toy thing you could run on your computer and make it warm.  Literally, you would install Bitcoin and you'd run this thing and it would just churn for a while and then, maybe the next day you come and you'd have 50 Bitcoin.  And do you know what you did with 50 Bitcoin?  Fucking nothing, because nobody you knew was into Bitcoin, and you had 50 of these things that weren't worth anything, and so you probably upgraded your computer and didn't even save the file, and it's all gone.  That was the first stage of Bitcoin.  It wasn't anything. 

Then there was a second stage of Bitcoin, where it was kind of money-ish.  It kind of had these kind of proto-money, kind of like it was actually worth something, and you got rug pulled in Mt. Gox.  This is the phase.  But it is kind of money, but it's not really money, it's still kind of play money.  It's got its play money characteristics, and it's fucking inflationary as all hell, right.  Basically, there's all this money being printed.  And then that was the stage of Satoshi Dice and all these business models built on the fact that it was basically instant and free and you could send this money around the world on-chain, all that stuff.  But that wasn't Bitcoin's final stage either, and that's the one we're in now. 

The final stage is, you've got to pay for it.  The mining subsidy goes away, and we're supporting it with fees.  And look, we have a lot of reasons to hope that it'll all work out, and that people will be prepared to pay fees because it will be useful and all these things.  But holy shit, we're not there yet.  Bitcoin is not what it is today.  So, just a hint, I'm warning you now, this is what the Blocksize Wars was about, right?  People thought that Bitcoin is like this and all the devs knew that this is not Bitcoin in its final form.  Bitcoin in its final form is going to be this fee-supported monster that's going to be very different.  And we hope it works, but we're not there yet.  Okay, so don't worry, you've got lots of speakers after me who'll talk it up, but I just wanted to set the floor right now.

Danny Knowles: Just before we came on stage, I was talking to Rusty, and he went, "I think I've just had my rant at you; I don't know if I've got the energy" and he's come on stage…!

Rusty Russell: That was a different rant!

Peter McCormack: So, okay firstly, should we just tell people what it is you actually do!  Does anyone actually know?  He's not some random guy who just came up a stage to rant!  Rusty does work in Bitcoin.

Rusty Russell: I rocked up to heckle, Peter.  

Peter McCormack: You're our technical guy tonight, right?

Rusty Russell: Yeah, so I work for Blockstream.  I think I was employee number 10 or something at Blockstream, I've been at Blockstream for eight years.  I was the first person to implement the Lightning Network.  Oh, thank you!  I used to chair the spec meetings, but that's hard work.  But they're still at 5.30am Adelaide time for historical reasons.  So, twice a year, everyone turns up at the wrong time because of daylight savings.  So, there's people around the world now who know when Adelaide daylight saving is for this really obscure reason that the whole Lightning spec system runs on it.  Anyway, what else?  I'm still waiting for my Mt. Gox coins to come back, so I'm that generation. 

Peter McCormack: But you're a technical guy for the night. 

Rusty Russell: Yeah, and I'm a developer, so I write code, that's what I do.

Peter McCormack: And you also have two Twitter accounts.

Rusty Russell: Oh, yeah.  So, I have my Rusty Twitter account and I have my Shit Bitcoiners Say Twitter account, because as a tech guy -- so, it's all people that are finance guys and they talk about, you know...  Okay, so I feel like the technical people are like, we're building this rocket ship and it's this complicated thing and we worry about fuel mix and all this stuff.  And then I come and I listen to your podcast and people are going, "Here's where I'm going to put my cappuccino machine inside my Moon base when we get there".  And I'm not even sure the rocket's going to fly, but they are like all the way over there.  And every so often, I find that the hype, especially, oh God, in bull markets, the hype goes astronomical because everyone's competing to have this vision of like how many citadel's they're going to own, or something. 

So, just in order to vent a little bit, you can see I have a tendency to do this, I did create a Shit Bitcoiners Say Twitter account and I would just retweet people. 

Peter McCormack: It's like the Bitcoin version of Libs of TikTok. 

Rusty Russell: Yes.  But it's not so much fun in bear markets, because it feels like kicking people when they're down sometimes.  But also, they're just not as enthusiastic, they're just stupid.

Peter McCormack: I saw you put me on there once.  I was like, "You fucker, I thought we were friends". 

Rusty Russell: I put myself on there once, so you're in good company. 

Peter McCormack: All right, listen, you put out a tweet saying, "I'm doing this live podcast, What Bitcoin Did, what do you want to talk about?" and then you just listed the things you wanted to talk about, and everyone went, "Yeah, we want to talk about that".  But I'm going to pick out the -- I mean, you said Tether, you said scars of the Blocksize Wars, and then Lightning and patience.  So, I'm just going to be selfish.  I definitely want to hear about scars of the Blocksize Wars, because I think I know what you're talking about, but tell me what you're talking about.

Rusty Russell: Okay, so, who here knows what we're talking about when they say Blocksize Wars?  Yeah, okay, that's my story.

Peter McCormack: Anyone who doesn't, because there are more hands up?

Rusty Russell: Anyone who doesn't?  Yeah, okay, cool.  Okay, so gee, shit, where do we start?  There's a whole book.  Every project goes through this, right?  So, I worked on Linux back when it was very much a geek thing, and there were these governance crises like, "How's this going to work?" and you're just this group of randos, right, you're trying to figure stuff out.  And the first one always feels like it is this massive unclimbable mountain and it's huge.  And for Bitcoin that was really the Blocksize War.  That was, everyone knows Bitcoin doesn't scale, there's only 1 MB every ten minutes, kind of thing, that'll only give us, what, three or four transactions per second, obviously not enough for the entire universe; we need to grow.  And so people were like, you know, "We should increase the block size". 

Kind of think of it like a bridge.  So, there's a bridge, it's the first bridge in the world, and everyone's like, "Wow, this is great".  Not many people use it yet, because it's kind of obscure, but it's got like a 1 ton load limit on it.  And it gets busier and busier and busier, and there are people like, "Man, it's getting crowded.  Why don't we just take the sign down?"  And the engineers are like, "No, you can't just do that".  And then they go, "Okay, well fair, what is the limit?"  And we're like, "We don't know".  "What do you mean you don't know?  Surely then we could just take the sign down?"  It's like, No, we don't know exactly, but there is a limit.  There is a limit beyond which the whole thing will collapse.  And once it collapses, you can't undo that.  You can find out the limit by getting there and then going, 'Shit, that was too much'". 

But also, this was in a time where it was very much a divide.  There were the Bitcoin enthusiasts, and the Bitcoin enthusiasts had grown up in this first two stages of Bitcoin development, the economic eras, where it was instant and it was free and all these things, and it was uncrowded and you could just send shit on-chain all the time.  There was this thing called Satoshi Dice that, you could send it money, and every block randomly, you'd either get double your money back, or there were different games.  But if you didn't win, it sent you back 1 satoshi, right, because you could do that shit.  And then there was this whole dust rule that wiped out the whole business model.  But the point was that there was all this stuff.  And so there were OG Bitcoiners who were like, "No, Bitcoin is like this.  It's free and it's got to stay free and it's got to be instant and scalable and all these things, and we've just got to get rid of the limit".  And there were tech people going, "No, we're looking at the pillars under the bridge, red lights are flashing everywhere.  We need to drop, if anything.  Luke Dashjr's there going, "300 kB, we should drop it to 300 kB". 

So, when there's experts, there's disagreements.  But it turns out, okay, so hilariously, we had a plan.  It's like, "Okay, we can increase the load limit on the bridge, but your cars are going to have to change, your trucks are going to have to have these weird wing SegWit things, you can't just drive your same truck".  And people are like, "No, I don't want to change my trucks".  And I'm like, "You're kind of going to need to change them because we've figured out a way to do it, but you're going to have to change your trucks".  And everyone's like, "No, we just want to fucking drive over the bridge the way we always did".  And so there was this rift.  And part of the thing is that the experts knew, we knew all along that this was coming.  We knew that there was a limit for a reason, we had to have a limit, this was all obvious.  And so, the fact that this hadn't permeated -- because in the Bitcoin hype space, bad news doesn't propagate.  Good news, "Fucking, yeah"!  El Salvador suddenly farted Bitcoin and boom, everyone knows it.  But hey, "There has to be a block size limit, we're not going to scale" yeah, you might not have heard that before.  So, there was a communications problem, shall we say. 

Now, the devs were shocked that nobody knew this.  We were literally like, "You didn't?  Like, what?"  Obviously, they don't understand, but it turns out most people didn't understand.  So, it was this huge frisson.  And it turned out that the solution they had, this whole, "We'll do SegWit, we'll do this thing, we can get some extension, we can squeeze more stuff in there, as long as we change the way we do things and stuff", it was proposed, it was like, "Everyone will be happy with this".  But it turned out they weren't, because they were actually doing this thing called ASICBoost and they were stealing from their own customers, all bullshit.  Anyway, the point was there was massive economic incentives against changing the way we did things in the Blocksize Wars, which we didn't even know about.  But suddenly, where is all this FUD coming from?  It turns out it was really well-funded, mainly by Bitmain, who they're proudly still buying a lot of their ASICs from, I might note.  Stop doing that, they are not your friends.  But yes, so that was the Blocksize Wars.  Where were we?

Peter McCormack: Is anybody not following Rusty?

Danny Knowles: Give a summation, Pete!

Rusty Russell: Is anybody following?  Okay, break it down. 

Peter McCormack: Just as a TL;DR, Bitcoin was getting popular and there's a block size limit.  Some people wanted to increase the block size, because they thought that would be good for business, but that meant less decentralisation.  The other people rejected that and wanted maximum decentralisation and said look, "Wait for the devs, they'll figure it out with scaling".  I think that's the TL;DR.  But it was a very vicious, vicious war.

Rusty Russell: Oh, yeah, and we stumbled into it because nobody understood why.  The devs were like, "I don't understand what people's problem is.  One, it's clear that we can't scale infinitely, obviously physically clear", well it was to devs anyway; and secondly, "Well, why aren't people getting it?"  It turns out there was some motivated reasoning on the other side.  And so this is a template for future debates.  There is definitely going to be people --

Peter McCormack: Can I just add some more context?  This felt like, at the time, so I came in around that time properly and both arguments made sense, because somebody would be like, "Well, if we make the blocks bigger, we can have more transactions and that's good for business".  "Yeah, that makes sense".  But then people would say, "Yeah, but if we make the blocks too big, we won't be decentralised and then the government can close us down".  "Yeah, that makes sense".  So, it could be quite hard to kind of figure out which was right.  But what was very obvious when you were on the fence was how vicious the war was.  It was genuinely a war and some people felt like this could be the end of Bitcoin.  It genuinely felt like it could be.  And it was probably the most scary time for some people who'd been in Bitcoin at the time, and I think that leads to the point where you're saying, "The scars of the Blocksize Wars", because we lost people. 

There's a number of people who used to be in Bitcoin.  People like Roger Ver, he went off and he was a very important person for Bitcoin.  He's now gone, he's gone forever doing his bullshit BCH.  And we lost a lot of people.  And so, there are scars; there are people still scarred from that.

Rusty Russell: Yeah, so that is all true.  And look, shit, we could do a whole 20 minutes on the Blocksize Wars and all the fun stuff that came out.  But prior to this all blowing up, there was a number of people, including Pieter Wuille, who was the main architect, I would guess, of SegWit, which is the expansion we did get, he had a plan for how we would increase block size, based on some stats that I'd pulled out on bandwidth rates around the world going up, "So, maybe we can, you know, maybe we can go up a certain percentage a year to a certain amount".  For technical reasons, it turns out we could always reduce the block size without a hard fork, but increasing it would be a hard fork.  So, there's an answer that well, we could just kind of have a theoretical maximum, but then we can kind of soft fork it down.  There was a whole conference on scaling Bitcoin and all the cool things we could do and ways we could do this.  It was very much a technical kind of thing. 

Now, that's really hard to talk to anyone about, because there was like, "No, we've made that decision, we are never increasing the block size".  And people like me are going, "I really want to increase the block size just a little bit".  The problem is that any increase, from here anyway, is pretty much a hard fork, and we have never done that before.  That means that if we ever hard fork, if we ever change the rules in a way like that, then old Bitcoin software would stop working and everyone would need to upgrade, and we've never done that before.  You can take a Bitcoin version 0.3.2 from back in like, wow, like prehistory, and with a couple of bug fixes, because there are actual bugs that were fixed, and a bit of adaption for the modern network, you can make it sync all the way up to modern Bitcoin.  We've never broken old shit.  You shouldn't run that, by the way, because it's grossly insecure, because a number of bugs have been found, but in theory at least, we don't do that.

Peter McCormack: So, is the scar you're talking about, the scars from the Blocksize Wars, is that now, it's very difficult to get new things into Bitcoin?  We've seen drivechains, which I'm not that interested in, but we've also seen the work that was done by Jeremy Rubin that was kind of rejected. 

Danny Knowles: Covenants. 

Peter McCormack: Yeah, covenants.  Do you think it's almost impossible to get something through unless you are some well-respected OG Bitcoin developer; is that the scar you're talking about?

Rusty Russell: Yeah, I think it's getting harder.  People have gone like, "We don't ever want to change Bitcoin".  I have sympathy with all views.  I have sympathy with the big blockers who were like, literally, "The reason that we want bigger blocks is because if we get traction fast enough, then we can outpace government regulation, which is going to come down and destroy us".  There was this argument of, "We need to get popular, we need to get fast", which I think is valid.  I think there is a solid argument for, at some point, Bitcoin is going to be too important for us to trust any change.  There's a solid argument to say, at some point, we've got to fossilise.  At some point, we're going to go, "No, this is literally like fucking with the world constitution.  We just want to not ever touch it".

Now, however well intentioned it is, there'll be so much at play and so many weird people funding shit and misinformation, you'll just be like, "No, no, we're not changing it because I don't trust the information coming", because it's too gamed.  Hopefully we're not there yet, because there's some things we really want to fix.  As a dev, there's some stuff we really, really desperately want.  Covenants, full covenants would be one of them.  But the problem is, we disagree on how to do it.  Jeremy has this CTV thing which is kind of like...  Look, do you know there's a trick where you can use a chainsaw and a hammer to screw in a nail?  It's really impressive, but sometimes you just want a fucking screwdriver.  So, the fact that you can make these things work doesn't mean you should go there, and CTV is a bit like that. 

There are a whole heap of different ways that -- and part of the problem with Covenants is everyone goes, "Well, it's for all these".  Here's 28 things you want for covenants, but it turns out you don't want any of them.  It hasn't got a killer use case.  Someone hasn't pointed at it and said, "You need covenants for this", except for Lightning wants ANYPREVOUT, which yes, can be used for covenants, or we can do it with covenants and everything else.  And there's this idea called vaults, where you could basically have this thing that would be more secure for your Bitcoin, but that really only matters to large custodians.  You and I are never going to use a vault, it's just not going to happen, sorry.  I talked to one of the, okay, I'm not going to drop names, but somebody who is very much into vaults, and he's like, "I'm never going to use a vault.  This is for large custodians", and I'm like, "Okay, so NYDIG really want vaults", and that's not bad, but it's not like drop-the-network-and-panic thing. 

But there are scars around this whole conversation.  Like, God forbid anyone suggests that we should increase the block size.  That is still a toxic wasteland, and you cannot go there.  I try it every so often, and it is not popular.  We should increase the block size. 

Danny Knowles: On that note, let's move on to Tether!

Peter McCormack: Yeah, we'll make a show on that. 

Danny Knowles: Less controversy!

Peter McCormack: We'll do that.  Okay, Tether.  I'm just going to just say Tether and let you speak. 

Rusty Russell: Okay, so wait. 

Peter McCormack: I think I might disagree with you in a minute.  Danny does.

Rusty Russell: Right.  I am a tech nerd, I'm in here for the tech, I'm not the finance dude.  But you did this great three-part series with Lyn Alden and she talked about the incentives for fiat currencies, and I'm going, "Tether, yeah".  The incentive is to print, the incentive is always to print, why wouldn't you print?  And they have printed in the past.  So, Tether is a stablecoin, they claim to be backed.  Only one day, a whole heap of their stuff got seized, so they very quietly just stuffed a few Bitfinex shares in there and some other assets and some sticky tape and some gum and went, "We are backed by equivalents now".  They literally just changed the wording and they went, "No, it's all good", and it kept working.  And then they printed $1 billion of Tether to give to Celsius.  You may have heard of them, right?  So, they give $1 billion.  That wasn't backed by anything either.  It was backed by securities, but I don't think they actually transferred it, they just had a legal agreement that they could take their crypto assets.  And it was not $1 billion of crypto assets either. 

Now, okay, so we know in the past they have not been fully backed, and we know that when they weren't fully backed, they lied about it.  But of course, what else are they going to do?  You don't want a bank run, so the incentives are all very clear.  If something goes wrong, they're going to lie about it.  Now, the thing I hate about Tether is that, so imagine you're running Tether.  You've got a few billion dollars, you've lost a little bit, but mainly you've got most of the money.  Let's just assume for a moment it's not all a massive fraud, and they've got most of the money sitting in various places and they have access to it all, okay.  You could be totally transparent.  You could, every quarter, release, "Here's all our investments and everything else".  But why would you?  That just opens you to risk, because when you lose money, it'll be obvious that you're fractional. 

You're much better off just keeping that as a bit of a black box.  And what do people rely on in a black box?  They look around them and they go, you just need some celebrity endorsements, you need some major bitcoiners who are going to say good things about you, and most of us will take their word for it.  So, that's what you do.  And it's really fucking effective and it really pisses me off.  I feel that as bitcoiners, we should have higher standards.  And one of the things...  So, the Blockstream motto is, "Don't trust, verify".  But people think of Tether and Bitcoin as the same thing, but from that point of view, they are complete opposites.  Bitcoin is this "Don't trust, verify" and literally Tether is, "Trust me, bro, you can't verify" as a business model.

Now, right now with bond rates at like 5%, it's an incredibly profitable business model.  They could totally be pissing away a lot of money and making a lot of wrong bets.  By the way, they've never lost money on any of their bets, which is a pretty amazing.  They had Chinese paper and shit and they had like $1 billion in Celsius, for fuck's sake, and they didn't lose a single cent, according to them which is, look, they're just the best traders ever!  Their timing is immaculate; they have never lost money, man. 

Danny Knowles: So, I'm going to play devil's advocate here.  I didn't have defending tether on my bingo card, but Tether is a one-to-one backed stablecoin --

Rusty Russell: Allegedly.

Danny Knowles: Well, yeah.  They allege they have $1 in reserve for every $1 they issue. 

Peter McCormack: Or equivalent assets.

Danny Knowles: Or equivalent assets.  So, we know in the past that they've used corporate loans, they've used treasury for a long time, I think, and then maybe two years ago, they decided to move almost entirely to US Treasuries.  So, they played the bond trade better than almost anyone, and I don't think that was necessarily intentional, it was just timing.  So, I can completely buy that in the past, they had huge holes in their balance sheet.  But I think the Fed has most likely saved Tether, and I assume they are now fully backed and probably in massive profit because of Fed fund rates. 

Rusty Russell: You'd think making $4 billion a year, they could really lose a fair bit. 

Danny Knowles: Yeah exactly. 

Rusty Russell: But there's two things here.  Bitfinex got hacked, like the exchange that is basically Tether got hacked, and they did this weird thing where they kind of bailed themselves out by selling effectively stock in Bitfinex, and you'd take that in lieu and then you could trade it and everything else.  They did the same thing with Tether, because it is a cash cow.  So, they could totally, if it turned out they only had half the money, they could probably totally buy their way out of that, because that would only take them what, like, ten years to get back.  So, they could do that if they did get caught.  So, they obviously know this, so how cautious they're going to be with the money that they can piss away? 

But here's the other thing that really disturbs me.  It used to be Tether was used by traders and traders will get in and out of positions really fucking fast.  Like, if it de-pegged, they would vanish so quickly.  Now, what you really want, you want dumb money.  You know what you really want?  You want Alex Gladstein of the HRF to start backing it, so that all your Nigerian farmers and people like that, who are going to be the last ones to get out, you want them holding your Tether.  Because if they hold 30% of Tether, they can never redeem that.  So, you can basically then go fractional.  You can go fractional because you know 30% of holders can never get out.  And that's what really concerns me.  The traders, the insiders, the people who know, the people who are watching, they'll get out, they'll get out at 90 cents to the dollar or something if it collapses.  But the Nigerian farmers who thought they were holding USD in this app, because it had worked for however long, they're fucked.  They're getting nothing.  Those people are going to zero and they will not understand why.

Peter McCormack: You could be wrong though, right?

Rusty Russell: Okay, yes.  This is very important.  I have no inside knowledge about the internals of Tether and everything else.  But I have been in Bitcoin for a while, and every time there's been a case where people have had a giant bag of money and have made promises based on it, it has not ended well.  These guys could be the exception.  Sure, they lied and cheated in the past, but maybe they don't do that any more.  And the incentives are for them to do this, but maybe they don't do that any more and maybe they're not doing that and maybe it will all be fine.  You want to take that side of the bet?  

Danny Knowles: What do you think about the idea that they are now so integral to the entire financial system in the US?  I think they're the 20th biggest buyer of US bonds. 

Rusty Russell: They're too big to fail? 

Danny Knowles: Potentially. 

Peter McCormack: Get bailed out? 

Rusty Russell: Oh, man.  I do wonder if that's the endgame, right?  If they can get big enough, then they can negotiate.  They can draw that line between being friendly enough with seizing assets and stuff, but still --

Danny Knowles: They're the central bank now.

Rusty Russell: Convenient, yeah.  They become the CBDC.  They basically finagle their way through that, so they're friendly enough.  But I mean, they can't KYC everyone, because that's half the use of Tether, so they can't.  They've got to thread that needle.  And look, maybe they will do that.  But the incentives for this whole system are pretty bad.  And the problem is that Tether works, it actually works.

Peter McCormack: It works really well. 

Rusty Russell: It works really well.

Peter McCormack: I had to use it in Argentina recently.  It works really well. 

Rusty Russell: Absolutely.  So, the longer it works, you could yell from the rooftops about, "Look, you realise that this could all go one day".  And look, I don't blame anyone.  You criticise the lifeboats all you want, I'm not going to criticise people for jumping into them.  And in some of these cases, absolutely.  If it's Tether or nothing, fucking go for Tether.  But the problem is this complacency inevitably sets in.  If it's worked for five years, if it's worked for six years, if it's worked for a decade, you leave your savings in it, you just trust it, you just leave it there.  But the problem with stablecoins is they are 99% stable and 1% they go to zero, and you never know when that's going to be.  And this has been the history of all of these instruments.

Peter McCormack: It feels like this is easily solved though.  You recognise the value of Tether if it works, if it is honest, if it is backed, like you think it is, if there's no weird financial engineering.  And so, they've obviously created a very good business as well, very profitable, so the incentives line up for them to be fully backed. 

Rusty Russell: Wait, hold on, wait, there's the fiat incentive, right?  The incentive is in theory for you to make sure your fiat, you don't print money and you keep it going and everything.  Oh no, Jesus, what are you going to do, print money?  Yeah, no, you fucking print money.  The incentives surely are, you look at it and go, "Well, they wouldn't do that.  That would be self-defeating, that would destroy their own currency".  They always destroy their own currency.  So, you can sit here and say, look, they've got a profitable business, surely they won't do anything stupid.  They will do something stupid, Peter!  And you won't know, and that's the problem.  You won't know when they've done something stupid.  It will come out five years later.  It will go for an awfully long time with them being partial, and maybe they try to make it back with some desperate trade and everything else. 

Maybe they're scrounging behind the scenes, the rats leave the sinking ship, some people abscond with a few more billion here and there, and it goes off to Cayman Islands and everything else, but they're still papering over, and then one day it doesn't.  It can go for an awfully long time.  Seriously, I reckon you could probably run -- I mean, how much did they drawdown?  They had that massive drawdown and they drew down less than 30%.  Well, there's their benchmark.  Worst case, they can blow 70% of it and no one would ever know.  So, it is an incredibly profitable business model, so in theory, it should all work.  But, oh man, it's going to suck when it doesn't.

Peter McCormack: That was like the Red Bull of interviews. 

Danny Knowles: Do you have a separate Twitter handle called Bitfinexed?!

Rusty Russell: No, but this is the problem.  I'm going to be wrong so much, people are sick of hearing about this.  The incentives are very one-directional here. 

Peter McCormack: You're sounding the warning, just in case. 

Rusty Russell: I'm trying to.  But the longer it keeps working, Rusty's just this nut.  And everyone who dismisses everyone who's the Bitcoin truthers, there's these Tether truthers, and I'm like --

Danny Knowles: When it blows up, I'm going to look like such a dick!

Rusty Russell: Yeah, that's right.  So, while I feel that the dialogue is very one-sided, so I'm being completely extremely fucking watch out, treat it like a hot potato.  As long as you understand the risks, I think you can probably safely handle Tether.  There is definitely a custodian risk here.  Understand that you could be rug pulled, and everything else.  The danger is that inevitably, we're people.  The more it works, the more it's smooth, next thing you know, "Oh, I didn't keep my wealth in Bitcoin because it was too volatile, I kept it in Tether because it was stable".

Peter McCormack: We have a break now. 

Danny Knowles: Buy Bitcoin, fuck Tether!

Peter McCormack: Thank you, Rusty.