WBD503 Audio Transcription

Why are Bitcoin Forks Contentious with Adam Back

Release date: Thursday 19th May

Note: the following is a transcription of my interview with Adam Back. 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.

Adam Back, an original cypherpunk, inventor of Hashcash and co-founder and CEO of Blockstream. We discuss the controversy around BIP 119: the soft fork proposal aiming to add “covenants” to Bitcoin. We also talk about the collapse of Luna.


“It’s a competition of ideas… and it’s a collaborative competition of ideas. So if you propose an organised set of ideas as a proposal, some parts of them might end up being good and used, and some parts might not. And you just have to be open to that.”

— Adam Back


Interview Transcription

Peter McCormack: Morning, Adam, how are you doing?

Adam Back: Morning, good.

Peter McCormack: New studio, never done this one before.

Adam Back: Yeah, it's nice.

Peter McCormack: Yeah.  Anyway, we've got a technical topic to get into, which is always good to do on this show, because I get a lot of questions coming in from people who aren't technical like myself, who want an explanation of things.  And over the last few weeks I would say, I became aware of a push by Jeremy Rubin, and some people who support him, to push CTV into Bitcoin, and it seems like a big change, there was a lot of drama, there's a lot of discussion.  It's the first I ever became aware of it, even though Udi had been telling me this had been worked on for years.  And when I compare it to something like Taproot, which was something even though I didn't understand it, I was fully aware of it for a long time, probably at least two years before it went live; I was aware there was broad consensus for it.

So, it opens up a whole can of worms of things to talk about that I think -- there's going to be people who are not technical, like myself, who don't really always care about the technical details, but they care that Bitcoin's being looked after in the right way, and there aren't things being implemented that maybe bring risks to their investment.  So, loads to get into.  We're going to start right at the basics, and I want to start with, what is a BIP.  I covered this before, a couple of years ago now, but I think that's a good starting point.  So, do you want to explain what a BIP is?

Adam Back: A BIP stands for Bitcoin Improvement Proposal, and it's a kind of specification for a proposed change to Bitcoin, which may be a consensus change, or may be a kind of opt-in thing that some wallets can use and others choose not to.  So, I guess the BIP1 is interesting as well.  It was proposed by Amir Taaki, who is -- there's confusion between a Bitcoin developer of a particular wallet, or a Bitcoin developer, or people looking at Bitcoin protocols in general.  So, I think BIPs are interesting to anybody technical working on Bitcoin, whichever wallet they're working on, and some people who are not even programmers, but they're interested in protocol and security.

Peter McCormack: Were there changes to Bitcoin prior to BIPs?

Adam Back: Probably.  I'm supposing in the early days that Satoshi just fixed patches and rolled them out and told people on IRC and they all upgraded.  So, I think it was less caution, probably lower security bar in the early days, then things became more cautious, more careful, more systematic as time went forward.

Peter McCormack: And then, when did the BIPs come then?

Adam Back: I'm not sure of the date of BIP1, but I think it's got to be pretty early, 2009/2010, something like that, I would assume.

Peter McCormack: I'm trying to imagine how the process became established that there would be a BIP.

Adam Back: Right.  Well, I think actually the Bitcoin upgrade process draws a lot from other internet protocols, and that evolved in the, I guess, mid-1980s, 1990s, something called the IETF, Internet Engineering Task Force, which manages a list of RFCs, which are BIP-like things, describing email protocols, web protocols, low-level internet transport protocols, all kinds of things that ensure interoperability at the internet level.

Actually, more coincidentally, there was a big political debate about internet protocols, I think, in the mid-1980s, where there was starting to be big internet companies, like Cisco and Microsoft, who saw the internet protocols, and thought they would like a vote and that they thought their companies' interests should be represented in how the protocols evolved.  So, I see some parallels with that with the Blocksize Wars 2015. 

The people working on the protocol said, "This isn't good.  Internet protocol should be for the good of humans everywhere, users first.  Companies, you're just service providers.  It's interesting, we want to hear your technical experts, but we don't want you to override internet protocols in the interests of your companies, because your companies could come and go, and the internet is for humanity, not for you".

Peter McCormack: Well, it reminds me of the whole net neutrality thing.

Adam Back: Yeah, so ultimately the users won that one, and there's a famous quote from David Clark which is something like, "We don't believe in kings and presidents, we believe in rough consensus and running code".  So, that was a famous quote making the point that they were going to use a rough consensus process, which is a kind of decision-making process from business or organisations really, just organisations, political organisations, that's hundreds of years old and predates internet and everything.  So, Bitcoin adopts that, but makes it a bit more decentralised.

Peter McCormack: Okay, so BIP1, Amir Taaki?

Adam Back: Yeah, so it describes some of the process, creating a BIP, making a draft, that kind of thing.  And usually, you would want some example code that proves that it works.  So, if there are one or two implementations, that's better than words, because words can sometimes be less precise than you thought they were.

Peter McCormack: Okay, but how many BIPs have we had now?

Adam Back: I don't know, probably 100 or something.

Peter McCormack: Okay, so someone's got an idea to improve Bitcoin, they develop a BIP, or they propose a BIP.  How do they distribute this, because this is the thing that's totally unclear to me?

Adam Back: So, there's not much of a bar to getting a BIP.  The fact that something has a BIP doesn't mean that anybody's going to implement it, or the Bitcoin protocol will be upgraded using it.  But if you propose something that passes some basic -- it makes rough sense, some people are discussing it, you can ask whoever is currently tasked with assigning numbers, can you have a BIP number, and then they add it to a list on a website.

Peter McCormack: Right, okay, so there is a central repository of the BIPs?

Adam Back: Yeah, I think it's in GitHub.

Peter McCormack: Right, okay, so somebody makes a proposal for an idea for an upgrade, they request a BIP number, maybe have some sample code to explain it, and then that goes then to who?

Adam Back: It's just available to anybody who is interested in Bitcoin's technical evolution.  And depending on the type of BIP, some of them are Core network rules, and then people are interested in security and robustness; and if it's an opcode, they're interested in --

Peter McCormack: What's an opcode? 

Adam Back: Oh yeah, sorry!

Peter McCormack: We hear about these all the time and I've never actually, in any interview said, "Somebody explain an opcode to me".

Adam Back: So, at the lowest level, Bitcoin has a kind of programming language, and a little bit like CPUs have operations at the basic level, which are very simple things, they just add two numbers together and things like that; so, Bitcoin has something like that at the lowest level.  And, over time there have been new opcodes added to add new functionality, basically.

Peter McCormack: Right, okay.  And so, the BIP goes out there, a few people might take a look, a lot of people might take a look, you will tend to get some feedback delivered in certain ways; and I guess if a BIP doesn't have much interest, doesn't have much support, then that person either has to rework it or forget about it; whereas, if something shows a lot of interest, I guess this is the rough consensus thing, right; it's a bit messy?

Adam Back: Yeah.  I mean, the ITEF has some texts and history about what rough consensus means in this process, and their model is a little bit more centralised.  But generally, in some cases there may be multiple ways to do something, so one question is, "Which is the best way to do it?"  And then, the idea is that -- I mean, consensus generally means unanimity; everybody agrees.  But you've got to place some sanity bars around that, because some people object in an illogical way just to be disruptive, not technical, regular contributors; but if you just made a hard rule, if there's any objections, everything stops.

Now, in a political version of consensus, that's basically what happens; unless everybody agrees, things stay the same.  And that's kind of what you want for Bitcoin in a way, right, you don't want people to make drama and be able to change it, because that's where we started from, the 2008 Financial Crash, the politicians being asked to make extraordinary changes to the money supply, and basically making the money unreliable.

Peter McCormack: Like Luna?

Adam Back: Yeah, recent news!  Anyway, that's the kind of process.

Peter McCormack: Right, okay, so I get it.  And then, if there's some interesting in something, it gets developed, it gets to the point where there's general rough consensus that it's wanted, the code gets developed, it gets tested and then you go through the process of, I guess, activation.  So, that's like the general process.

Adam Back: Right.  I mean, there's some historic examples, because some people commenting are looking at it ab initio without any of the history of how it has worked and that it does work, and it works very well for what it intends to do, which is to make it very hard to push changes to Bitcoin that not everybody wants.  And they might look at that as an obstacle, because they want this.  So then they'll say, "Well, we should change the process".  First question is, "Well, before you change the process, do you realise there is a process, that it works, and why it's designed, why Bitcoin uses this process?"

I mean, some historic examples are P2SH, which probably people have heard about.  It's a mechanism that's used for multisig.  At the time it was introduced, I don't know, probably about ten years ago, there were three versions.  And it's an interesting cautionary history there, which is there was one called OP_EVAL, which some people were extremely keen on and pushed hard.  And they were about to launch it and Russell O'Connor had a look at it, a developer who's still active in Bitcoin today, he had a look at it and found a horrendous flaw with it and said, "Wait a minute, you can't activate this, it will create a denial of service amplification, it's too dangerous".  So, everybody recognised that and said, "Wow, that's a terrible flaw, let's abort that immediately", at the last minute.

So, then it switched over to P2SH, which was an alternative proposal, but there was actually another one called CHV and there was a bit too much urgency.  So, P2SH won over CHV, even though people were on the fence, or thinking maybe CHV was slightly better, but they didn't care enough.  But in hindsight, given a few years and seeing how it evolved and what the limits were, it turns out CHV would have been the better of the three proposals they narrowly missed.

So, that was cautionary, so I think people learn from cautionary things.  So now, here we are, scroll forward to covenants, there are four or five different versions.  Again, somebody's very enthusiastic about one and making a lot of noise, but there are other variants and they're probably better.

Peter McCormack: And I guess with certain upgrades, there's certain ways to do things, which can be a discussion point; there's certain changes that are easy binary decisions to make, so if it just makes Bitcoin more efficient, either it makes it more efficient or it doesn't.

Adam Back: Taproot is a good example of that.  I mean, the payload -- well, both parts of Taproot make it more efficient and more private, potentially.  The Schnorr signatures are just more compact signatures, and of course there were discussions about specific encoding and how to make the multi-signatures, so they had lots of discussion there.  But there was no debate that Schnorr signatures were the building block.

Peter McCormack: But with something like covenants, it brings new functionality where it doesn't seem to be that everyone agrees what it actually means, or fully understands maybe what they mean, and if that brings any risk to fungibility, and blacklist, whitelist, you'll be able to correct whether that's right or wrong; but that's a more contentious issue, because if you can change what you can do with Bitcoin, rather than just make it more efficient, then there deserves to be more of a debate about what the outcome for this is?

Adam Back: Yeah, I mean it's true to say that Bitcoin's programming language is compact, concise and security first and intentionally safe and cautious.  And so, the covenant idea expands what you can do.  I think philosophically, Bitcoin likes small scripts, not big, complicated programmes.  So, I suppose there's a potential that people could write more elaborate things, which probably beyond a certain point, are not good uses of block space, or could be done in different ways.

Peter McCormack: All right, so we understand what BIPs are.  Let's get into the specifics of this then, and it's no criticism of Jeremy, I talk to him, he's a great guy; but at the same time, like I say, for me it just came out of nowhere and it might have been worked on for two years, maybe I've missed some of the discussions.  But for me, Taproot was out there in the open, being openly discussed by everybody I know, for a good two or three years.  I think I first spoke to Andrew Poelstra about it at MIT about three years ago.

The first I became aware of CTV was a few weeks ago, and people were talking about activation.  So, that was an immediate red flag to me, because I was thinking, "I'm pretty sure I should have heard about this", so I was nervous about that.  Before we get into that though, shall we explain what covenants are and why there's some contention around them?

Adam Back: Yeah.  So, covenant is, I guess, a word borrowed from legal terminology.  So, if you've ever engaged in a property transaction, real estate, there'll be some clauses on your house or your apartment that says, "You've got to maintain this boundary wall with a neighbour", or something like that, and it's a covenant because when you sell it, you've got to maintain that clause.  You can't decide you don't like it and remove it.  So, it's not within your authority to remove this rule.

That applies to these scripts.  The idea is that you can create a transaction, you opt into it as a receiver, that has a restriction on how you can spend it.  And that actually turns out to be useful for certain programmability.  A most concretely useful example people give is a time-lock vault.  If you try to do cold storage, it's quite interesting.  It would mean that you could store some Bitcoins in cold storage; when you want to take them out, you do a de-vaulting process and then, let's say for a week, the transaction could be cancelled and go back into the vault.  And if it's not cancelled within a week, then you can spend it wherever you were trying to spend it.

That gives you some safety, because you could give the cancel button to your friends, extended family, colleagues, that kind of thing; and if they see a lot of coins moving and they can't reach you, they could get suspicious and cancel it.  So, it has that kind of -- that's the basic idea for the vaults.  And it's possible to build a version of that in today's Bitcoin scripts by deleting private keys, but it's not the most elegant way to do it, but it would work.  Nobody's actually implemented that.  And of course, it's a flexible mechanism, so people could build other things with it, potentially UTXO sharing, which is a kind of scaling idea.

But I think people also lose track of the fact that a lot of things that are possible with Bitcoin scripts are not implemented and made available in wallets in a usable form.  So, there's lots of innovation that can happen today in existing script, and so that kind of ease of use, standardisation of formats, understandable and easy-to-use available in wallets is lagging often by years.  So, even once we get Taproot, it will be a while, as you see, before wallets support it.  It was the same with SegWit.  It's over 50% now, but it took some years.  And I'm sure it will be the same for covenants too, or time-lock vaults.

Peter McCormack: Is there any validity to the aTalkrgument against covenants that they affect Bitcoin's fungibility?

Adam Back: Well, there are two versions of covenants broadly, and one is recursive.

Peter McCormack: Is this what Andreas was talking about?

Adam Back: Yeah, I think so.  So, the original covenants for a Bitcoin proposal was by Greg Maxwell on the Bitcoin Talk list I think 2013, something like CSIP, which is a technical term, hilariously bad idea, and so he was thinking about the things that could go wrong.  But I think since that time, people have started to think more about it, muse on the idea, mull it over and think that certainly restricted covenants are probably safe, and even recursive covenants are reasonably safe, because you opt in, and so forth.

So, there are at least three arguments about why that might be okay, that have been developed since then.  One is, you have to opt in, as a recipient.  Now, that can be a slippery slope, because if somebody in authority, like a bank says, "Opt in or don't have business with us", how much choice do you really have, right?  So that's one argument.  Another one is, the fact that Bitcoin doesn't have covenant capability today, because it's a programming language, you can often build a huge permutation of things using the building blocks. 

So, Russell O'Connor makes the argument that it's almost an accident that Bitcoin doesn't have covenants today.  It's because some of the opcodes were removed, because they had bugs in the very early days.  If a few other features had been implemented slightly different, we would have had them by side effect.  So it's, as the functionality of Bitcoin scripts gets more general over time, it's probably going to cross that boundary, possibly even by accident.  And there might be a small possibility that it's possible today even, if somebody really put a lot of effort into finding some clever tricks to just about make it work.  That's to say, if it's almost possible, its general implication of programmability, maybe it's not such a clear, important line. 

I think the last thing is just the programmability.  So, you don't have to use it, it's opt in, and the KYC Bitcoin risk that people worry about, they're basically saying, "What if a set of banks agreed to a standard where there had to be KYC on a Bitcoin that you couldn't transfer it?"  It turns out you can implement that with multisigs, and people have proposed that before.  So, from that point of view, again you'd have to opt in, so it has largely the same characteristics, and those are generally called formally Online Covenants, which is there's a multisig; one key is held by an authority and it imposes some policy, so kind of an online covenant, because that policy can be recursive if they choose.

Peter McCormack: And, what about whitelists and blacklists; is that possible; is that a valid concern?

Adam Back: I think they mean basically the same thing.

Peter McCormack: Okay.  So you yourself, Adam, obviously I back you, no pun intended, but are you okay with the idea of covenants?

Adam Back: Yeah, I mean I think so.  Whether it starts with simple, or goes straight to a more generic version is a topic being discussed at the moment, and there are technical people on either side of that.  So, one argument for starting simple is that Bitcoin can gain some experience from that, and that can inform the generic; but the other point is that the rate of introduction of new opcodes is not very rapid, so if you're going to take the time and expense to review and get consensus on a covenant feature, you might as well do it properly once; that's the counterargument.  I think a lot of the technical people are a lot more on that line of argument. 

Rusty Russell has proposed one, which is sort of extensible within itself, so it can start simple and have future extensions over time, so that's another hybrid approach.

Peter McCormack: Okay, I'll come back to that.  The other thing I wanted to ask you about is that, and again I don't understand the detail of this, but some of the arguments regarded CTV was that it would introduce the ability to shitcoin with Bitcoin.  What was this about?

Adam Back: I did not pick that one up.

Peter McCormack: You didn't pick that one up?

Adam Back: No.

Peter McCormack: That came up quite a bit.  It was starting to give more Ethereum-type functionality with scripting, sort of smart contract stuff.

Adam Back: Well, I guess if you can make more complicated scripts, maybe some automated trading kind of functionality becomes possible.

Peter McCormack: Right.  So, would you say it's a fair way to explain where we are is that, covenants, there is broad consensus that these might be okay; we should be careful; there is a proposal from Jeremy that has been fairly well developed that he feels is ready to go, that doesn't have strong consensus either amongst Core devs, general devs, the community; there are competing proposals that those who aren't with Jeremy think need considering as well; and, Jeremy's idea isn't dead, it's just not at that point where it's reached consensus; is that fair?

Adam Back: Yeah, I think that's approximately the situation.

Peter McCormack: His proposal isn't dead?

Adam Back: No, and assuming that consensus is reached on something, it may be none of them, it may be a synthesis of the best ideas from all of them, for example.

Peter McCormack: Right.

Adam Back: So, I mean there's a sort of experience in history to it.  So for example, many people have gone through this process where they've proposed a BIP, and it's informed something else.  So, Christian Decker, who's one of the Lightning authors, and actually the co-author of APO, which is another opcode that's in flight at the moment, that's sort of competing, or ideally should come before covenants because it's been in the queue for longer and helps Lightning.

So, before SegWit, Christian was working with Pieter Wuille on Normalised Transaction IDs, so it was an attempt to fix the problem that SegWit fixes.  And, it turned out SegWit was a better way to do it, so that BIP got left in history.  So, that's an experience you have, right; you work on something, it reaches its limits and somebody finds a better way to do it.  Now, the fact that you worked on it, it wasn't lost effort, because the realisation that there were limitations there informed a new design, or maybe it borrowed some of the design ideas.

Peter McCormack: So, why do you think we've got to this point where this is this kind of disagreement, because there's so many people that support what Jeremy's been doing, think it's fine, think it's ready to go; there are people who don't.  And as I said to you, you've got it on your T-shirt, the "Don't trust, verify"; it's a really great thing that people say as a kind of moniker for Bitcoin.  But actually, it's completely impossible for me to not trust some people. 

I cannot verify wallet code; I have to trust, at some point, that somebody else has done that.  I cannot verify whether this is good for Bitcoin or not; I have to build trust, and I build trust around people.  So, I have a high level of trust in yourself, Pieter Wuille, Andrew Poelstra, Matt Corallo, Greg Maxwell, and that's trust that's been built over time.  I have less trust with Jeremy.  That's not to say I don't trust him, I just haven't had that built.  And the trust is built by a combination of how much time I've spent with somebody, how much trust they have built within the community, career history, time in the community, etc.

So, I think people do have to rely on trust sometimes.  And when Jeremy proposes and idea and I review it say, "Okay, that's interesting, I don't know much about it", and I see Adam Back is not supporting it, and I see maybe other people, like Matt Corallo, not even talking about it, that's a red flag for me.  I don't know if that's a good or a bad thing.  I think that's just a reality.

Adam Back: Yeah, I think it's true for everybody, because I mean my background is applied crypto, so I know quite a bit about internals of Schnorr signatures and things.  So in a way, the Schnorr signature part of Taproot came from IRC discussion in 2013, where I was saying, "Bitcoin should use Schnorr signatures; they're better", and people listened over time and eventually, that got picked up.

But at the same time, I wouldn't necessarily know the serialisation details of Taproot, or how it's been coded, or stuff like that.  So like yourself, I'm trusting that some very clever engineers have carefully built that, rigorously tested it, peer-reviewed line-by-line each line of code in it, and so you're sort of relying as well that there are many eyes carefully checking, a lot of mutual interest for it to be secure.  So, I think that's ultimately what people are relying on.

I mean, I think there's something about the Bitcoin change process which is maybe a bit like aircraft design modifications, right.

Peter McCormack: Well, have you seen the Boeing documentary?

Adam Back: I haven't seen the documentary, but I read quite a bit about the MAX issue, and I watch Aircraft Crash Investigation, which I find very interesting.

Peter McCormack: Yeah, I've watched I think nearly every episode of that, because I used to be shit-scared of flying, and watching Air Crash Investigation actually made me more confident in flying, strangely, because you see the A/B testing that a bunch of people have been through with flying. 

But the Boeing thing, it was when I first interviewed Austin Hill.  He said to me, "Bitcoin development should be like [I think] nuclear reactor building, it should be considered the same" or, no, "space flight, it should be the same, because one mistake on a space flight, everyone's dead.  One mistake in Bitcoin, everyone loses their money".  It has to be considered the same, the same rigour, the same testing.  That just reminds of that, and when you say aeroplanes, I was like, "Well, it all went wrong with Boeing", the rigour.

Adam Back: One of the reasons that I think that it's an interesting analogy, which is counterintuitive to many people, including technical people, like Jeremy, who come from other backgrounds, is that aircraft, yeah, there's lots of engineering; but ultimately, what makes them robust and safe is biological evolution.

So, it wasn't that aircraft engineers came along and said, "This is going to be the best design and when it goes wrong, we'll just make a better design".  It's more that they made the best design they know how to make, it breaks apart for some reason they didn't expect, they do the investigation and find out that square-edged windows develop cracks, so now aircraft have round-edged windows.  Nobody saw that coming, but now we know that.

What told us that wasn't human ingenuity, it was physics and biology.  So, there's a certain suspension of ego that comes with that, which is these evolved systems are too complex for humans to design ab initio; we had to design by observing, engaging in peer review, taking the best of everybody's ideas and learning from things that went wrong, or nearly went wrong before.  So, when you look at it from that perspective, there's something very wrong about one person saying, "Well, I want square windows, can I have them now?" and nobody found a bug in them.  Well, you might be right, we don't know, but let's continue with the safety process.

In the same way, when Boeing and Airbus make modifications to aircraft, they have to get them flight certified, which involves all kinds of rigorous testing.  So, Bitcoin is not so hierarchical, but it has an enormous amount of testing that goes into it.  So, part of the "Activate it now" is kind of bypassing that process, which is the danger.

Peter McCormack: Well, we've seen what happened to the opposite side with Ethereum, which openly said, "Let's move fast and break things", and they did break a lot of things, including people's money.

Adam Back: Yeah, I mean that kind of platform has tended to lose about $1 billion a year to hacks and rug-pulls and things.  That was last year, so you might have thought it was a one-off, but I think it lost $300 million within the first few months of this year, so it's on track.

Peter McCormack: It's on track!  Yeah, fuck that, I don't want that with my Bitcoin.  Okay, so going back to the approach then that Jeremy's taken, because like I say, I've gone with the "Can't verify, trust".  That's my approach; I should get that as a T-shirt.  I've trusted people and therefore, openly said I'm concerned about what's happening with this.  Is there a different approach you think Jeremy himself should have, or could have taken?

Adam Back: I mean, I think he would say he tried, because it's also a fact in open-source software in general, and that extends to Bitcoin as well, that there are volunteers working on things.  Either they're unpaid, or they're paid a stipend to do independently what they think is good for Bitcoin, so nobody's telling them what to do.  And generally, it's more fun for people to build things, rather than review other people's things.

So, it's just a fact that if you think you have a good idea and you want to get it deployed, that depends on other people agreeing it's a good idea, and being more forceful won't make them agree.  If anything, it will make them pull back.

Peter McCormack: Do the opposite, yeah.

Adam Back: So, I think that's where it went wrong, basically.  So, if you look at Schnorr, the time when I was discussing Schnorr signatures in 2013 on IRC until when Taproot and Schnorr became live, that's a long time, right; that's eight years, or something. 

Peter McCormack: Eight years, yeah.

Adam Back: But there was no force, I was just describing ideas.  I did almost nothing, in fact, apart from describing ideas and other people gradually got more excited and spent a lot of time working out details, and finally you had it.

I think it's something else from the ITEF, and of course the ITEF is a bit more hierarchical, but they have a Chair for each protocol discussion group, and the Chair's role is to find consensus, to see what the group of people are discussing, to declare when they reach consensus.  And it would be unthinkable to be the Chair to be a proposer of a protocol and then bless his own protocol and say, "Well, that one has consensus, we're going to activate it now".

But if you look at what Jeremy's doing, not through any malice, I think, but just because he's enthusiastic and wants to generate action, he's acting like the proposer, the implementer and the Chair, and the person who's sort of trying to organise the meetings and get it activated.  And it looks like a kind of one-man show.

I mean, I think what you just have to do, and I've discussed this with Jeremy in the last year as well, is you have to work with other people.  So, find somebody else who's enthusiastic about it to co-propose it, or co-author it, or champion it.  And then once you've got two people interested in it, maybe you'll find a third, and then that's a step closer to consensus.

Peter McCormack: It's like building political support?

Adam Back: Kind of.  I mean I think it's ultimately, most of the people who are involved in the protocol department have got pretty good at separating ego from technical comparisons.  So, you'll see people having what sound like fierce arguments about details, and at the end of the conversation they'll exclaim, "Oh, you're right.  That's great, now we can move on".  So, they've convinced themselves that they're wrong, that's fine.  They just want to see the best outcome go forward.

You've got to engage in that conversation, and I think it can be frustrating because of the imbalance of energy to build things versus review things.

Peter McCormack: Right, okay.  Yeah, and I guess for Jeremy, it sucks.  He's probably worked on this really hard for two years, feels like it's ready to go, wants to get it implemented and just hasn't been able to get that done.

Adam Back: Yeah, but like I say, it's kind of jumping the queue as well, because I described Christian Decker's experience.  He worked on the Uniform Transaction ID and that didn't go anywhere; SegWit took over.  But it was a better proposal and he recognises that and he had a small part in it.  So, that's just open-source development in general, it's just like that.

Peter McCormack: Coming up against the Bitcoin glacier, which is for a good reason.

Adam Back: Well, it's a competition of ideas, and it's a collaborative competition of ideas.  So, if you propose an organised set of ideas as a proposal, some parts of them might end up being good and used and some parts might not.  And you just have to be open to that.  And I think also, the imposition of a timeframe is risky as well.  So, Jeremy was saying, "I need your comments by here, or I'm going to activate it".  He's backed away from that now, but people really don't like that.

Peter McCormack: It's like a threat.

Adam Back: It's like Boeing, or the Space Shuttle Challenger, "We're going to launch on this date", it became a political decision and then it exploded.

Peter McCormack: Well, it's like S2X.

Adam Back: Yeah.

Peter McCormack: Yeah, okay.  So, let's talk about that, that activation, as well, because it seems then that not only did this idea come out of nowhere, it suddenly felt that there's a way to brute-force get this live, by using speedy activation, Speedy Trial activation, something again that I was told was used for Taproot.  I mean, I wouldn't know anything about this at the time.  Taproot was getting activated, great, the little orange squares were appearing, etc.  Now, I am interested in this, because why was Speedy Trial used for Taproot; what was used previously; what is this?

Adam Back: Yeah, there's a lot of confusion about that.  You see a lot of people, there's a CTV Telegram, and I think IRC and so on, and there's clearly confusion about miners' part in activation.  So, they think that, well you make a proposal and the miners vote on it, and it happens or not, and that's the process; but that's really not the process.  I mean, if we learnt anything from the block size drama and the final SegWit activation, via UASF, it's that miners have almost no influence on the outcome.  85% of miners said they were against that activation; it activated anyway and it continued mining Bitcoin.

So, that's a lesson that everybody should learn, now that there are a lot of new people who don't know about the lesson, and they just don't understand how it works.  But basically, what activates Bitcoin is more economics, the important economic nodes, like the exchanges, wallets, and payment processors and individuals and investors running full nodes.  If they agree and they upgrade, then the miners have to follow, and everybody has to follow to stay on the same network.

Now, of course, you can get forks and that's happened in the past, where you get some kind of minority.  But in particular, all the miners are doing is signalling that they've upgraded and helping users who don't upgrade to be safe, and that's it.  And in particular, if you say with a Speedy Trial, "Well, let's turn the decision over to miners", what does that mean?  I mean, miners use pools, there are very few pool operators, so it basically means delegating the judgement of a consensus decision for the ecosystem and the investors to, I don't know, 20-odd guys who run pools.

Peter McCormack: So, what is the point of miners signalling then?

Adam Back: It's to protect people who don't upgrade.  So, if you don't upgrade through a soft fork, you don't have a full understanding, and miners could steal from you, basically.

Peter McCormack: Okay. 

Adam Back: So, if the miners indicate that they've upgraded, now the cooperation of the majority honest miners will protect you if you don't upgrade, and that's the point of it.

Peter McCormack: So, where are we at now with CTV?  Has it been paused; are there other discussions?

Adam Back: Well, the activation discussion went away and I mean there's a bit more awareness of things that came before, things that were in flight, so I think people realise that.  I mean, I think the way forward is more cooperative comparison, just work between the different designs.

Peter McCormack: Right, so is Jeremy talking to Rusty; are they comparing ideas?

Adam Back: I'm not sure.  I mean, I think Jeremy has kind of spun out and made new discussion channels, which there are no developers in; or, maybe application developers, but no protocol developers, or almost no protocol developers.  So, that's a little echo chamber.

Peter McCormack: So, he needs to engage with protocol developers.

Adam Back: Well, he tried, but he ran into the problem that everybody runs into; it's a mutual problem that there are fewer reviewers.  So, I think if more proposals get together and they discuss amongst themselves, I'm sure Jeremy will join in, and that will be a way forward.

Peter McCormack: Is there any amount of, not intentional or not by malicious intent, but a certain amount of gatekeeping that perhaps exists around yourself, Pieter Wuille, Andrew Poelstra, Gregory Maxwell, Matt Corallo, established people within the community who've built that reputation whereby, if you can't get some sort of agreement from you guys that things are slower, and maybe that's a good thing, or maybe that's a bad thing; what happens after you guys?

Adam Back: Well, I mean I think that everybody's nuance, you know, I popped up in 2013 pro-Schnorr signatures and now we have them, and I was certainly not forceful, right.  And there are many new people involved in protocol design, implementation, that were not here three years ago, five years, one year ago.  And for example, Lightning, which is a pretty big thing, there were two versions of that proposed in parallel.  One was duplex micropayment channels by Christian Decker, and the other one was the Lightning proposal by Poon and Dryja.  And no offence to those guys, but nobody had heard of them before, and it was a really ingenious idea, co-invented by two groups of people in parallel without knowledge.

So, that shows that newcomers can propose not only good ideas and get immediate fast traction on them, so for Lightning to work, it needed an opcode; but people saw that idea of Lightning had a lot of potential, so there was enthusiasm to work out the details and get the opcodes done.  So, I don't think there is.  I mean, I think the thing is that --

Peter McCormack: But there may be a certain hierarchy that exists.

Adam Back: There are people with skillsets that you want to somehow enthuse to pay attention to review things, but I think you still have to use some kind of common sense, human interaction skills.  So, if you want somebody to review something, look at how other people are doing things, copy other people.  So, if you're not very good at managing human interactions, find a co-author, find somebody else to help you champion it or navigate it.  I mean, I think ultimately, if you push too hard, it backfires, so I think that's a potential source of mistakes as well.

Peter McCormack: And we've seen that.  Again, it's like a historical lesson, where people have tried to rush changes or push changes into Bitcoin that haven't had consensus.  It's kind of pushed some people away from Bitcoin.  They've had the wrong time preference on upgrades.

Adam Back: Well, I mean there have been some rage quits historically around the forks, for sure, and at the time of the fork drama, you were paying attention at that time and talking about it afterwards, I think investors found the fact that it was hard to forcibly change Bitcoin to be a very bullish signal.

Peter McCormack: Very.

Adam Back: You hear Saylor talking about that; that was the single most important thing.  Another interesting thing you get if you talk to investors, Saylor and people like his sort of investor mindset, about what features they would like for Bitcoin, their primary feature is, "Just don't break it".  So, they're actually not very interested in opt-in micropayments or fancy things, because they're like, "Well, how risky is that?  Status quo is good enough, don't break it".  So I think, covenants, the pitch for the investor is, vaults are good for high-value, long-term storage.

Peter McCormack: And obviously, for Saylor, he's in a different scenario where he holds so much Bitcoin, that something like the spendability if not as interesting to him, because his incentive model is based on the growth of hodlers, which if demand continues to increase, means his billions become more billions, etc.  And I would be cautious of -- I completely agree with him, just don't break it, but I would also be cautious to give too much weight to somebody whose incentive model is slightly different.

Adam Back: Well, yeah.  You can see there are broadly two groups of -- I mean, people use Bitcoin for many reasons.  Some of them would be perfectly happy if it was a new gold ETF, even if it wasn't very bearer, just because of the harder money; I think Saifedean and maybe in that camp.  I think Hal Finney was talking about Layer 2 banks, so they're thinking like a hard base money.

But for a lot of people, what makes it exciting is that it's permissionless, bearer money, with a low barrier to entry; smartphone, it's all on app, you're ready to go with no permission.  So, for that to practically work, you need scalability solutions, you need payment solutions that work for people in emerging markets who don't have that much money.  And so, I think both are important, because ultimately for the investor, they're investing in the adoption rate for something.  And for there to be adoption, there needs to be payment capability.

Peter McCormack: Well, my actual use for Bitcoin is pretty much the same as it was day one.  I buy it and I transfer it to my wallet, and sometimes I transfer a bit of that back out and I sell it.  That's pretty much it.  My only advancements have been transferring some into a Lightning wallet, which I've had on my phone, I've used when I've been away, and moving to a multisig cold storage solution.  So, even if Bitcoin ossified right now, for me it would still serve a great purpose for the rest of my life.

Adam Back: Yeah, I think that's true, that it's good enough in a way.  But with careful changes, I think it could be better, and it could satisfy different use cases.  And of course, privacy is another one, fungibility, privacy.

Peter McCormack: I feel like we've covered CTV now.  Where are we at with privacy and fungibility; any big changes being proposed; anything you're interested in?

Adam Back: Well, I mean I'm interested in privacy and fungibility; it's kind of where I came from, pre-Bitcoin, was working on enormous electronic hash protocols and things, and that's where confidential transactions came from, and I would try to improve Bitcoin's privacy, and realised that this is too complicated a change to expect to get into Bitcoin as a BIP in 2013, so I changed focus to work on modularity, so that you could have sidechains where you could do things like that.

So, maybe eventually, that will reach a state of safety and compactness and characteristics that it could actually get into Bitcoin, but there are incremental things happening for privacy, I think.  The Taproot and Schnorr help a bit as well, and I think there are, at times, network-level things, which are less consensus, more network policy privacy things.  Unfortunately, there's not really any major silver bullet technology that's reached maturity.  I mean, certainly confidential transactions are interesting, candidate tech, the SNARK technology and bulletproofs are interesting, and the layer versions of confidential transactions use bulletproofs. 

But those are still improving, so it's probably some more years.  So, it's sort of a patience game, I guess.  But there is a certain sort of safety and efficiency bar for things to go into Bitcoin.

Peter McCormack: Just move it slowly.  Shall we talk about Luna before we finish up, just because it's a bit of fun?

Adam Back: If you like.

Peter McCormack: Well, I mean look, it's not fun that a lot of people have lost a lot of money, that's fucking terrible.  And also, it's not fun to see Bitcoin being dragged about by shitcoin activity elsewhere; that's the most annoying thing for me.  And where it becomes really annoying, Adam, is you have these big crashes in the markets, suddenly it's on Sky and the BBC and the Daily Mail, and then your friends speak to you and they say, "Oh, I hear Bitcoin's crashing again" and people want to ask you about the negative side of it.

You're like, "Yeah, but this was because of some stupid Ponzi stablecoin idea that had nothing to do with Bitcoin.  It was that somebody was then using the money they printed to buy a bunch of Bitcoin, and then they had to resell".  We're affected by these decisions, and you were chatting to me beforehand about this perhaps affecting leverage traders, in some ways unfairly.

Adam Back: Right.  I mean, people are investing in Bitcoin because they believe in it, because they look at the macroeconomic background, you know, money printing, inflation, interest rates and looking for a hedge.  If you just buy and cold store and the price drops a bit and then it halves, then it doubles across the period of a year, it doesn't make that much difference, once you get accustomed to the volatility. 

But some people are more risk on, they've got leverage on.  So, if somebody creates a bit of a flash crash drama, through market misunderstanding, that can hurt them economically, through no fault of their own.  So, you get this sort of contagion effect.  And of course, people are leveraged.  They're either forced to manually deleverage to protect themselves, or the trading platforms will force delever them.  So, that amplifies the flash crash basically, and you see stats posted, so many hundred liquidations; that's what it's about. 

So, yeah.  I mean, Luna basically was sort of like a fractional-reserve bank, operating as a so-called decentralised thing, and their idea was basically sort of shares in the bank, which were the Luna tokens, and there were some VC investors that put in the original money.  It's a bit unclear whether it's somewhere between $1.5 billion and $3 billion.  I'm not sure actually how much VC money got committed.  I presume they got Luna tokens; bit hazy on the details. 

Like yourself, I only really paid attention when people started getting enthusiastic about Luna Foundation buying $1 billion-plus in Bitcoin.  I was like, "Wait a minute, these guys are going to get themselves flash-crashed and sell all the Bitcoin in one go, and that won't be helpful", and that was in late March.  So, scroll forward about six weeks and it happened.

Peter McCormack: Yeah, I mean I smelled a different rat, in that it reminded me of Block.One, this idea that you can create a token, print and sell a bunch of it, and then take that to buy a bunch of Bitcoin.  One is, I feel like you're openly telling people they're making a bad decision, because you're taking their money and buying Bitcoin; that's what you're doing.  You're saying, "Here, give me your money, I'll give you this token, I'm going to take that, I'm going to buy Bitcoin".  You're giving a market signal to that person.

That idea, because it came out of nowhere for me.  I don't know how long Luna had been around, because I don't pay attention to this shit.  I've never actually used a stablecoin ever, never used one.  I know I'm meant to use Liquid!  But to me, I was just like, "This smells fishy, I don't like this".  I hadn't even got into the mechanics of how it worked, but the 20% interest, I was like, "When anyone's offering 20% interest, that smells fishy as well".  There were just so many things that smelled fishy, and suddenly here we are, it's a complete market crash.  But you'd spotted it early on.

Adam Back: Yeah, well I said some of the same things early, which is they were buying -- people were getting enthusiastic about how many Bitcoin they said they were buying, so I just asked on Twitter some pointed questions like, "Where is the money coming from that you're buying the Bitcoin with?  Is it ICO money?  Is the ICO money your collateral?  If it is, the whole thing could flash crash, because the value of your ICO can get stressed at the same time, then the whole thing can collapse".

Of course, they protested and said that some of it was real money, that they'd actually got wire transfers from investors.  But evidently, a lot of it wasn't real money, because their total supply of Luna at one point reached, I think, 17 billion or something, and so they had, I don't know, maybe 10% or 15% of that in real money.  So, yeah.

Peter McCormack: Fractional-reserve shitcoining.

Adam Back: I mean, about stablecoins, I think it also is quite unhelpful, because it's brought disrepute to conservative types of stablecoins, like Tether and Circle and Gemini, all of which have one-to-one, or more than one-to-one backing of actual US dollar short-term deposits, cash equivalents, and various dated interest-bearing things.  So, they can redeem one-to-one all day long basically, because I think for example, during the last few days, Tether redeemed $3 billion out of, I don't know what they're at, 80-something, in the space of a few days; so, clearly they can redeem.  But UST versus USDT, it's only one letter difference.  I think some people get confused.

Peter McCormack: Well, it does get confusing.  There are so many of them, they get confusing, a lot of them appear on different chains, you don't know what each one stands for.  I was on Pomp's show and Pomp was asking me about it, and I was like, "It sounds a bit shady", and this guy dropped me an email and said, "You don't really understand Luna and Terra", and he's sent me some articles.  And I was like, this article was, I don't know, 20 pages long if it was printed on A4, and it was all about the explanation about how it works, how it maintains its peg, how the Luna token interacts with the stablecoin.  I was like, "I don't have time to understand all this shit.  I don't have time".

Again, it's a bit like trusting almost.  If I was going to use a stablecoin, I think I'd trust, like you say, Gemini, Tether, Circle; I trust those, I'm going to use it.  A bunch of people have trusted this and have got completely and utterly rekt, because they have no idea what they're dealing with.

Adam Back: Yeah, I mean some of the transparency went out the window too, because I think you could observe the Luna token versus the Terra UST, because it was on-chain.  And because it inflated the Luna supply by orders of magnitude, it created more and more of them to try and keep the thing afloat; but the Bitcoin reserves, they transferred them all, and you could see that transfer to a professional market maker to try and defend.  So, that's lost visibility and they're not answering questions.  Did they sell all of it; did they sell some of it; did they sell none of it?  We don't know. 

I think another curious aspect of this, and really should cause some introspection for the VCs and professional investors, who put up the $1.5 billion to $3 billion in the beginning, is their reputations and money helped bootstrap this thing.  They may, I presume, get Luna tokens, or some kind of discounted tokens or stock in it; and because it's tokenised, one of the attractions for an investor is they can take the money out very quickly.

There was a period where this thing was going up in value very rapidly, right.  So, some of the VCs may have actually made a large profit long before it imploded.  And of course, taking a profit out actually leaves the thing high leveraged and more fragile.  So, there's a question there of whether professional investors are moving away from funding innovation and products that provide value to users, and there's an exit and people make money when they succeed in making a popular product versus renting reputation to Ponzi-like things, taking the money out early and letting it collapse, and then doing it again.

That is a risk for the end user.  It diverts capital from productive uses to casino behaviour, and it risks the ire of regulators.  I mean, immediately regulators are talking about Terra, and apparently the SEC was already investigating Terra and Do Kwon, one of the founders.

Peter McCormack: I mean, he could end up in jail, right?

Adam Back: I presume so.  I mean, he's in Singapore, but Singapore is a regulated financial centre in its own right.

Peter McCormack: Yeah.  It is frustrating, because every one of these rug-pulls ends up bringing Bitcoin into the same ecosystem of bullshit, and we end up having to not only defend, like I said earlier, defend it to my friends who are like, "What's happened to Bitcoin?" and you have to defend it to regulators.  And it's just bloody frustrating that people don't do a better job with this, especially with all these crypto VCs raising $1 billion, $3 billion, $5 billion to invest in Web3, and I feel like we're going to see this over and over again.  I also feel like over the next week or two, we're going to hear about some funds who have maybe blown up.

Adam Back: Yeah, I mean somebody's got to have been hurt.  And you know, the other thing was the explosion was considered to be a speculative attack; that was a funded, coordinated thing by somebody with a lot of market intelligence, trading capability and money to play with and some leverage.  So, it's kind of like the George Soros attack against the British pound peg in pre-euro days.

If you have money and you see something that's illogical and people are suspending logic and pumping up the value of something, you can short it and burst the bubble.  I think, some people feel that was a bad act; but actually, it's good, because it flushes out silly stuff, so that what's left makes sense.  And the fact that they succeeded at that will give them funding and embolden other people with that skillset to work their way down the list, and do that to anything else that doesn't make sense.  So ultimately, you end up with the survivors being less stupid ideas, more robust things.

But I'm not sure that the mechanism is there really to discourage VCs and big investors from lending reputation and putting discounted early money into these things, because some of them probably got their money out and made a profit on it.

Peter McCormack: Oh yeah, almost certainly; I don't think it changes that.  Hopefully, what it changes is the mindset of those who are potentially investing, but perhaps not.  I mean, we've seen a bunch of new Bitcoin maxis minted this week, people declaring they're done, and that's great; but we've also seen, I don't know if you've been on the Reddit, but all the terrible comments from people saying they've lost everything.

Adam Back: No, I mean it is bad.  I saw somebody commenting that one particularly bad thing about it is that, because advertised as a stablecoin as a dollar equivalent plus an interest rate, that's actually marketing itself to conservative investors who are saying, "I don't want the price volatility, I just want an interest rate".  As you said, it's naïve to think there's a risk-free 20% interest; that doesn't exist.  The free market interest rate, the risk-free interest rate is super-low at the moment, so that's a red flag.  But a lot of people aren't professional investors, they wouldn't spot that.  So, if they were looking for stability, they got the exact opposite.

Peter McCormack: Yeah, and also Alex Gladstein talks about how important stablecoins are for the developing world; we've had a long conversation about that, and it's one of the reasons I've become a little bit less harsh on other protocols, as much as I don't own any Ethereum, I wouldn't use it, or Tron.  At the same time, I know there's people in, like, Palestine or parts of Africa who are using stablecoins on these protocols, and they're using it for survival.  I mean, that's the money they trade, it's the way they protect their wealth.  Bitcoin's too risky for them.  But have these people got -- are they sophisticated enough to understand the difference between a Tether or a UST?  All these coins tend to sit near each other.

Adam Back: Yeah.  I think the network that's used to transport the stablecoin, the centralised ones like Circle and Tether and Gemini, doesn't really matter, because ultimately it's a claim on a central party and it's reserved, and the reserve is set aside for the client funds, right.  And for example, with Tether, you can use that on a Bitcoin Layer 2, on sidechains, and there are proposals to do similar things for Lightning, a couple of competing proposals for that. 

I mean, of course, I think US dollars fall rapidly in value at the moment with the amount of money printing, so it's short-term stable, but it's just predictably going down.  Now, Bitcoin has the opposite effect, which is it's a bit unstable, but it tends to go up, and so has a positive price trend over the longer term. 

So, I think a lot of probably the biggest amount of volume in stablecoins, bearing in mind there are competing ones by market cap, but by transferred volume, Tether is like 10X ahead of the rest, and that is because it's embedded in most exchanges as the de facto way to move dollars.  And so, it's mostly used for trading, people to move money between one exchange and another, as collateral for some of the derivative products and as Bitcoin secured lending, like on Hodl Hodl, you're going to borrow in a stablecoin basically, and then you can convert that into a bank account, or something.

The other annoying thing is that the media doesn't get it often, so it amplifies confusion.  So, they tried to say that Tether had broken the peg, but it's absolutely not the case; it was being redeemed at one-to-one the entire period.  But the point is that banks suck, so if you have a short time preference, you need some dollars in your bank now to cover a share loan, or you need Tethers to trade now, you'll pay a premium to get it fast.  And the only reason you're paying that premium is because banks suck.

So, if the market is being dramatic, the premium could be high or low, but this happens all the time.  And it's fine, because if you're not in a hurry, just wait; and if you're not in a hurry, just wait.  And if you are in a hurry, you're probably in a hurry for a reason, which is you see something that's a cheap price and you want to buy it, and you need to convert to Tether or to dollar to get it, so you'll pay the premium.  And there will be other people incentivised by that premium saying, "Well, I'm not in a hurry, I have some Tethers.  You can have them and I'll go convert them, pick up the few percent difference next week".

Peter McCormack: It's wild how often we have to deal with this shit.  I'm feeling like the NFTs are starting to crash as well now, and I think a few other projects will collapse, and I wonder if we're just going to do exactly the same in four years' time, three years' time now.

Adam Back: I notice there's various dominance indexes, but the Bitfinex one, which is BTCDOM, was up 20% in the last two days.

Peter McCormack: Wow.

Adam Back: So, when you get a global financial drama, the dollar tends to get stronger, flight to safety.  It might have its own problems, but people know it, and so it's kind of perceived flight to safety, and you see that in Bitcoin; flight out of the long tail of altcoins to Bitcoin.

Peter McCormack: How's everything in Blockstream?

Adam Back: Getting good, yeah, lots of mining infrastructure builds, and the solar Tesla project.

Peter McCormack: Yeah, fuck, I forgot about that.  Tell me about that; that's fascinating.

Adam Back: We did a jointly-funded project with Block and engaged Tesla to build the solar infrastructure and provide the megapack batteries.  So, it's off-grid, and the idea behind it is -- our thesis is that the economics of Bitcoin mining can help fund green power infrastructure, to increase power infrastructure.  The reason is that most power is actually overbuilt and underused, because it has to deal with peaks in demand.

Peter McCormack: Yeah, peak load.

Adam Back: So, if you were to build some power infrastructure, and somebody would come up and say, "Well, I will pay X cents per kWh, 24/7.  Anything you've got spare, I'll take it", that makes your economics much better.  So, you've got 100% due cycle on the supply, so that should make it more profitable and easier to finance.  So, that's what we're doing, starting with this project, and then we're interested to grow it hundredfold.

Peter McCormack: And Tesla were totally into it?

Adam Back: They're supplying the megapacks.  They actually do the solar install as well, but they're not an investor in it.

Peter McCormack: Okay, and where are you building that out?

Adam Back: This is in Texas.  So, we have a grid-connected site, and we didn't connect this part of it, but it's on the same site, and that site is primarily a windfarm.  So, if you look in the background, there's loads of big wind towers.

Peter McCormack: And, is this a test, or are you guys pretty happy with the economics of it?

Adam Back: I mean, we're happy with the economics of it.  But what we intend to do is make it transparent, so publish the financial model, profitability, real-time mining metrics, partly because there's a lot of confusion and people are writing reports and nobody quite believes the report, because one person's report says the opposite to another person's report.  But if we say, "Well, we'll actually just build it and show you and you can look at it for yourself, and then there's no debate, the numbers can speak for themselves", so that's what we're doing.

Peter McCormack: Amazing.  Well listen, always good to talk to you.

Adam Back: Likewise.

Peter McCormack: Not sure when we'll speak next.  We'll see you next probably at some event, but appreciate you going through all that technical stuff for me.  And yeah, good luck, Adam, and guess I'll see you some time later in the year.  Speak to you next week actually, why not!

Adam Back: All right.