Latest / TFTC: A Bitcoin Podcast / Bitcoin Is Not a Battery or a Power Projection | Rob Warren, Center Of Hash E002

Transcript

Bitcoin Is Not a Battery or a Power Projection | Rob Warren, Center Of Hash E002 · TFTC: A Bitcoin Podcast

• 0:00
  Rob, welcome back to Bitcoin Park.
• 0:02
  It's good to be here again.
• 0:04
  Last time was the Texas Energy Mining Summit.
• 0:08
  It feels like yesterday, but now a couple months in the past.
• 0:12
  Yeah.
• 0:13
  And so we're joined today by Rob Warren, who is the Head of Research and Education at Bitcoin Park.
• 0:21
  Talk a little bit about that role.
• 0:22
  Yeah. So I've worked predominantly in the mining industry and have been really deep in really kind of generally the research and education side of things, just because we've all been figuring stuff out in public as things do in Bitcoin.
• 0:37
  and jump down with Bitcoin Park essentially as the broad head of research and education,
• 0:43
  which gives me the opportunity to do exactly what I'm doing,
• 0:49
  but in the context of now the whole Bitcoin space,
• 0:53
  which is fantastic because it's not just exposure to mining,
• 0:56
  but it's also introductions to folks that are working on the human rights side of Bitcoin,
• 1:00
  who are working on the technical side of Bitcoin, the payments, the custodies.
• 1:03
  So it's really kind of a whole new aspect of work that's opened up, which I love.
• 1:11
  And in addition to work at Bitcoin Park, which we're now members of the organization.
• 1:19
  That's right.
• 1:19
  Here in Austin, happy members.
• 1:22
  You also wrote a book, The Bitcoin Miner's Almanac.
• 1:26
  That's right.
• 1:26
  And in my view, you know, we're friends, but also just in the discussions that we've had, you have a well-rounded perspective of granularly, you know, how a Bitcoin mine works, what the inputs are, Bitcoin at its broadest sense as a money network.
• 1:48
  And the combination of those for these early episodes of this podcast is, you know, you were one of the one, two and three people on my list to record the first episodes to kind of lay some groundwork of not just going into Bitcoin mining 101, but painting a broad structure of the critical aspects of how Bitcoin mining works, the energy components of it, why it's important to the system.
• 2:17
  and today i want to kind of dive deeper into the the energy side of what is actually happening
• 2:24
  you know in terms of why the energy is important to bitcoin also some points around centralization
• 2:30
  but before we get into those kind of deeper points i want to hear you articulate because
• 2:36
  we've had conversations in the past about how there's there's bad analogies of what bitcoin
• 2:42
  mining is or isn't one of which is that bitcoin mining is is a battery or bitcoin's a battery
• 2:49
  whether it's proof of power but just talk a little bit about you know from your perspective
• 2:55
  a function of mining to the bitcoin network and why it's important and what it is sure sure well
• 3:02
  i mean as as most things are it's always good to go back to the source document and kind of
• 3:07
  understand how things were understood at the earliest days of Bitcoin. So functionally,
• 3:13
  Bitcoin mining does two things. It allows for the issuance of new Bitcoin. That to us is called the
• 3:21
  block reward. So approximately every 10 minutes a block is found. You find a block, you are rewarded
• 3:27
  with something. You're rewarded with Bitcoin. That is a way to get Bitcoin into the world.
• 3:33
  You need a way to essentially distribute Bitcoin so that people can then have some utility, whether that's holding or spending or whatever they do with it.
• 3:42
  All Bitcoin that exists has come into existence through mining.
• 3:47
  The next thing that you need to do is you need to be able to transact with the asset.
• 3:51
  So you need to be able to actually send it to people.
• 3:53
  And that's the second part of Bitcoin mining, which is settlement.
• 3:56
  So it's the ability to actually submit a signed transaction to the mempool, have it included in a block, which then the miners are given a small reward for, which is the fee that you pay to transact on the Bitcoin network.
• 4:09
  And fundamentally, that's the two things that Bitcoin mining does from an operational, from a functional perspective in Bitcoin.
• 4:17
  And it's where we stand today is really worlds apart from where we were in the days of the white paper.
• 4:25
  And it's almost unrecognizable to look at Bitcoin as of 2025 and try to compare it to what it was in the earliest days, the 20-aughts, right?
• 4:38
  For one, the biggest difference is that when we talk about what a node is, Satoshi refers to a node as a node that is actually doing mining.
• 4:49
  And where we stand today, functionally, that's extremely rare.
• 4:53
  Now it's coming back and I'm sure we'll hit on this when we talk about the decentralization point, but it's interesting how you can learn a lot about the present by actually going to the original source documents, the white paper, and then understanding, okay, well, what's different and what's the same?
• 5:09
  And how do we hold true to what was actually given to us in the white paper to make sure that we're really appropriately stewarding this thing, this Bitcoin thing as we build on it in the future?
• 5:18
  And so what is it about the mining function and specifically energy that you view, why is it critical to the system?
• 5:34
  Like, why is this function critical to, so if they're, if they're doing, if, if Bitcoin miners are functionally doing two things, validating currency transactions and validating the fixed supply of the currency.
• 5:50
  Why and how is the mining function critical to that, but at a lower level, what is the function of energy and energy consumption within there?
• 5:58
  That's such a big question.
• 6:00
  it's such a big question and it's really interesting because when you when you look
• 6:05
  at the earliest conversations on something like the the bitcoin talk forums where a lot of folks
• 6:14
  who are household names in bitcoin are trying to figure out what is the extension or what is the
• 6:21
  natural growth path of bitcoin there are only very small glimpses into what the future of the mining
• 6:28
  sector would look like. There are some references to when I run my computer, it heats up my room.
• 6:35
  There's references to the noise, you know, how noisy your computer gets if you happen to be
• 6:39
  running your Bitcoin client and it's mining and your computer is now wrapped up, operating at 100%
• 6:44
  all the time. But it was, I'd argue that it's not really something that we have a fully fleshed out
• 6:53
  theory of. It's something that we're actually living through the history of right now.
• 6:57
  and fundamentally it all reduces to this truth,
• 7:02
  which is that to win a block,
• 7:05
  that's what miners are trying to do.
• 7:07
  They're trying to win a block
• 7:08
  and the mechanism that they use to do that
• 7:12
  is inputting information into this algorithmic machine
• 7:16
  we call the SHA-256D hashing algorithm,
• 7:19
  which is what we use to mine.
• 7:22
  Running that algorithm takes energy
• 7:24
  and like running any process,
• 7:27
  that has some kind of expected value, and we could get into what that means, the expected value or the statistical nature of it,
• 7:34
  running any process is going to have some kind of cost associated with it.
• 7:38
  And for Bitcoin miners, fundamentally, that cost, really the largest cost that we incur outside of buying machines,
• 7:45
  is the cost of running that machine.
• 7:47
  And so we find ourselves, just based on the simple fact that we want to run profitably,
• 7:52
  always beholden to the world of energy.
• 7:54
  And it's because we always want to reduce our operational cost as low as humanly possible.
• 8:00
  And what that means in the real world is that we become a very interesting consumer in the
• 8:06
  broad world of energy because we are not stuck in a single jurisdiction.
• 8:11
  We're not stuck operating on a single IP address in a single type of building.
• 8:16
  Our operational requirements are almost nothing like any other business that you could imagine.
• 8:21
  You know, if you want to set up a hot dog stand, you put it outside of the baseball field.
• 8:27
  Well, if you want to set up a Bitcoin mine, you go and you find some energy.
• 8:30
  If you want to set up an AI data center, well, I have some constraints.
• 8:34
  I have demands I have to make of my infrastructure provider.
• 8:38
  Bitcoin miners have an incredible degree of flexibility operationally that is unlike any other consumer in the market.
• 8:45
  And because of that, we are this kind of dung beetle in the market where we are constantly seeking out sources of energy that are abundant, that are waste, that are difficult to access.
• 8:58
  And we're monetizing them in some of the most far reaches of the world, really.
• 9:03
  So two things.
• 9:05
  One, a basic question.
• 9:07
  You referred to what miners are trying to do is win a block.
• 9:10
  connect that to the prior statement like the the concept of a block to
• 9:17
  the fixed supply and validating currency transactions just for for a baseline for
• 9:24
  for people who might the the concept people who the concept of a block might be esoteric okay and
• 9:30
  then second um on the idea of the consumption of energy if you would lay some ground
• 9:40
  work for what is actually like you plug a machine in but talk about in a in a literal sense how that
• 9:50
  power is performing the functions that you talked about enforcing the fixed supply and validating
• 9:56
  how does the bread box what what the heck is a bread box you're going from some natural resources
• 10:00
  to electricity to the bitcoin that's right helping somebody understand like what might be happening
• 10:06
  because they're not just, they're not taking, you know, it's not mining in the sense of pulling a
• 10:10
  resource out of the ground, but first connect the block to the core functions that you laid out.
• 10:15
  Perfect. So I, I like to conceptualize it and I, I did it. I did it in the book as well through
• 10:21
  the concept of the machine itself, right? The Bitcoin mining machine, a lot of us have seen
• 10:28
  images. It's easy to pull up a picture of, you know, an S21 or an S19 or a standard, you know,
• 10:34
  Watts miner in M30. And it's kind of this bread box looking machine. And it has fans on the outside
• 10:40
  of it. There's different versions. There's all sorts of futuristic ways to cool it, but that's
• 10:44
  not relevant. The most important thing is that you have some kind of machine and its sole purpose
• 10:49
  is to consume electricity and do this thing called Bitcoin mining. To run that machine,
• 10:55
  you only need three things. You need something to plug it into, source of energy. You need some
• 11:02
  physical way to cool it. And there's all sorts of creative ways that you can cool these machines,
• 11:08
  whether it's by using some kind of solvent, like a two phase solvent, you can use an oil,
• 11:13
  you can use water cooling plates, you can use air. There's all sorts of reasons to use all
• 11:17
  sorts of different cooling mechanisms. And then you need a connection to the internet.
• 11:20
  And the connection to the internet or a connection to a node is the way that you communicate with
• 11:25
  the network, the peer to peer network that is Bitcoin. So fundamentally what's happening,
• 11:30
  If you imagine, you know, in the simplest sense possible, a blockchain is a chain of blocks.
• 11:36
  And the question is, OK, well, what makes it a chain of blocks?
• 11:40
  What is a block?
• 11:41
  A block is simply a chunk of data included in that are transactions that people want to be settled in a block.
• 11:50
  And that data, just a string of data, is given to the machine as a template.
• 11:57
  and included in that data is essentially a reference. It's a link, it's a hash.
• 12:03
  So what had just happened, the prior block. So I have a chunk of data and in that data is some
• 12:10
  information about the prior block, some information about time, version, difficulty, all sorts of,
• 12:15
  you know, computer stuff. If you're not like a tech literate type person.
• 12:20
  And in that are transactions that people want to send. What do I do with that data? I run it
• 12:25
  through a hashing algorithm. Now, what is a hashing algorithm? You can think of it sort of like a big
• 12:30
  wood chipper. And it's a wood chipper that I put data into, and what it spits out is going to be
• 12:37
  just a cube. Okay, well, for our purposes, it looks a bit just kind of like a random 64-bit string,
• 12:45
  right? Random number. It's called a hash. There's something interesting about this shot256 wood
• 12:51
  shipper, which is that if I put the same piece of data in, I will always get the same output.
• 12:58
  So it's 100% deterministic that if I put in this, I get this out. But what makes it so powerful is
• 13:07
  that I can never look at the final product and know what I put in, which is a really powerful
• 13:13
  tool because you imagine that it allows you to package and verify data while giving no way to
• 13:22
  an attacker or some outside party of modifying it in transit or doing anything like this. If I want
• 13:28
  to send you a message and I send you a message and the hash of the message, if somebody gets between
• 13:33
  us and changes that message, if you go and you run that message into the hashing algorithm and
• 13:39
  you see that it gives you a different result out of the wood chipper than what I told you the result
• 13:43
  was, you know that somebody is screwed with the data. So it's a wonderful way from a safety and
• 13:48
  security mechanism to ensure data fidelity. But I do that through my Shot256 hashing algorithm,
• 13:56
  and I get an output. Now, the Bitcoin network doesn't just want any output.
• 14:03
  It wants an output that has certain characteristics to it. And the way that we think about this is
• 14:08
  what we call difficulty right so i don't know if how deep down that rabbit hole we want to go
• 14:13
  let's keep it to the um the direct like not let's like not get into difficulty necessarily but
• 14:22
  the direct connection between that the energy you described to how how that relates to you
• 14:30
  mentioned the transactions are included and then the the pointing to a prior block a prior set of
• 14:37
  these transactions, but then maybe articulate how it's connected to the fixed supply of enforcing
• 14:43
  the 21 million. Sure. So, so in the process of, of running this hashing algorithm, um, what you're
• 14:50
  trying to do is you're trying to essentially achieve the difficulty, which allows you to,
• 14:54
  to find a block, right? Um, the only way to do this is to include another little bit of random
• 15:01
  data that we call a nonce into the ingredients going into that wood chipper right The reason I do this is because I can never predict what I get out of the machine So I do this millions upon millions upon trillions of times second after second
• 15:18
  And what I'm fundamentally doing is I'm taking an input source of energy and I'm doing a
• 15:23
  certain kind of work, which is called hashing.
• 15:26
  And I can't know what the output of that work is going to be, but that work has a certain
• 15:30
  value associated with it.
• 15:31
  And this kind of gets into this idea of pooled mining or contributing work to the network, which I think is maybe a later conversation when it comes to the centralization of the network. But this work has a value. And this is essentially kind of a derivative, but still a bit of a commoditized value.
• 15:49
  So the energy that I put in can be essentially the expense side of the equation.
• 15:58
  And the work that I get out can essentially be the profit side of an equation when it comes to running a Bitcoin mine.
• 16:05
  So I have an incentive.
• 16:06
  I have a block reward.
• 16:08
  And the block reward is what eventually issues Bitcoin over time, which then will fulfill the 21 million, which is the hard cap supply.
• 16:17
  That's how we get there.
• 16:18
  We get there by taking input energy, which has a cost associated with it.
• 16:24
  Sometimes, however, we get paid to use the energy because what we perceive as a cost is actually a liability to another type of business.
• 16:32
  We consume something, we convert it into hashes, which have some kind of value associated with them.
• 16:38
  Functionally, what that does in the Bitcoin network is it helps us issue and settle transactions.
• 16:43
  It's kind of the oil in the machine that keeps everything running.
• 16:45
  and that's what brings us on that issuance curve to that eventual implied 21 million bitcoin
• 16:51
  um and then fundamentally over time where that leaves us is um a whole bunch of different ways
• 17:00
  to monetize but through what we call the block or the subsidy is sort of the the the um the the
• 17:06
  inputs the uh the revenues of the business and so without going into necessarily what staking is
• 17:14
  versus proof of work but you could try to explain why this process consumption of energy the
• 17:24
  conversion of energy into these esoteric concepts of hashes and then where it starts to get more
• 17:31
  functional literal which is that allows anybody else in the network to to verify something easily
• 17:37
  even if it was difficult to produce why is this function critical to the permissionless nature of
• 17:45
  bitcoin and and ultimately bitcoin's ability to operate in a way that is resistant to censorship
• 17:53
  like why so it's doing these functions but why does it need the energy to be consumed like why
• 18:00
  couldn't why couldn't we do this without without energy without energy without the mining function
• 18:05
  What the Fudsters say is, why can't we just switch to proof of stake and just make it an internally self-referential system, right?
• 18:14
  It's really indigenous to the design of Bitcoin that it is permissionless.
• 18:20
  You know, it's said in the white paper that nodes can leave and rejoin at will.
• 18:27
  Bitcoin wasn't designed to be the best at something.
• 18:33
  Well, let me rephrase that.
• 18:34
  Bitcoin wasn't designed to compete with a Venmo or to compete with a big settlement layer or a visa right out of the box.
• 18:41
  It was designed to be fundamentally unkillable.
• 18:44
  It was designed to be anti-fragile in its nature.
• 18:47
  And the way that you do that is you have to make really, really smart design choices to ensure that you don't put yourself in a situation where you incur some kind of existential risk.
• 18:57
  Because what you're always trying to hedge against is the existential risk.
• 19:00
  What is the thing that could kill me?
• 19:02
  proof of stake runs into issues without getting too deep into it or it has collusion problems
• 19:10
  if you wind up with too much of whatever your your staking reward is winding up in the hands
• 19:16
  of particular actors or validators you have collusion risk that is just kind of baked into
• 19:21
  the system as as a risk of the function of the system bitcoin does something different it does
• 19:27
  not ground itself in a set of internal game theoretics that are tied strictly to the inside
• 19:34
  of the game. It actually ties itself into the outside world. And it demands that you use real
• 19:42
  energy in the real world if you're going to produce hashes and participate in the network.
• 19:48
  So this fundamentally makes it different. And this has been really the main source of critique that
• 19:53
  we've seen over the last couple of years has been a lot of the energy use of Bitcoin. Now,
• 19:58
  baked into that is the assumption that producing or using energy is a bad thing, which is a bit of
• 20:05
  a silly argument in my perspective. If you look at the development of nations over time and how
• 20:11
  that is directly tied to the way that they produce and consume energy, a bit of a Malthusian take,
• 20:16
  right? The idea that you need to constantly reduce and consume less and do less and do less. It's
• 20:21
  it's the opposite of whatever the E accelerated folks on X are talking about.
• 20:28
  But what Bitcoin does is as part of its actual design, and the reason for its design are really,
• 20:34
  really interesting from a historical perspective that are worthwhile to go into if you're a fan
• 20:39
  of like Nick Zaba or Wei Dai, but not so much relevant to the conversation. It solves a really
• 20:45
  interesting problem. And the problem that it solves is that it needs to be hard to make money.
• 20:51
  I mean by physical creation, not like get a job, make money. There has to be some kind of cost
• 20:57
  associated with producing the money in the same way that there is a cost associated with mining
• 21:03
  gold. How do I do that in the world of computers where everything baked into the world of computers
• 21:11
  is designed around getting faster and better and quicker and cheaper? That is really an
• 21:18
  intractable problem if you try to solve it in the way that kind of a proof of stake does is totally
• 21:23
  in as a totally internal problem what satoshi did that was so interesting in the way that they
• 21:28
  designed bitcoin is that they didn't try to actually solve the problem internally what they
• 21:33
  said instead was i know that people are going to enter this game and leave this game all i can do
• 21:43
  for you is provide the incentive, a stick. I can provide for you the block reward and the fees,
• 21:50
  and I can let you know how many there are going to be. And I can let you know essentially what
• 21:56
  the issue and schedule is going to be like across halvings. Consistent with the fixed supply.
• 22:00
  Consistent with the fixed supply. And I'm going to let you figure everything else out.
• 22:05
  And so if you figure out a really creative way to mine Bitcoin, say you don't want to use a CPU
• 22:12
  anymore and you want to start using a GPU, go for it because you're chasing an incentive.
• 22:18
  If you figure out how to manufacture an ASIC and mine Bitcoin even more effectively,
• 22:25
  go for it. I'm not even going to concern myself with that. All I'm going to do in the network
• 22:30
  itself is build a mechanism that keeps us aligned with that base incentive, which is 21 million
• 22:39
  issued across this rough timeline,
• 22:42
  essentially certain block heights where adjustments are made,
• 22:45
  moderated by the difficulty adjustment.
• 22:47
  So all that Bitcoin knows,
• 22:49
  the only thing Bitcoin knows,
• 22:50
  if Bitcoin can know anything about the world of mining,
• 22:54
  the only thing it knows about the world of mining
• 22:55
  is when block times come in.
• 22:58
  That's it.
• 23:00
  Because all it does when the difficulty adjustment occurs
• 23:04
  is look back and say,
• 23:06
  were blocks coming in faster than I wanted them to?
• 23:09
  or were they coming in slower than i wanted it to okay they were coming in faster well then i need
• 23:14
  to i need to make the bar a little bit harder to find a block coming in slower okay we'll do the
• 23:21
  opposite that's the only thing bitcoin knows about the world of mining it doesn't know who's mining
• 23:27
  it doesn't know where you're mining i think it doesn't know what kind of hardware you're using
• 23:31
  okay so connect that idea doesn't know who's mining where anyone's mining
• 23:37
  it knows that it just saw a new block and it has an objective way to evaluate whether or not it's
• 23:45
  valid without relying on somebody else that's right right connect that to the permissionless
• 23:52
  nature that basically any miner can come in and mine that's right um but then anybody can also
• 24:01
  send transactions that something about that process of no one knowing who's mining other
• 24:10
  than the proof of it that energy is a key component that any anybody can evaluate at
• 24:18
  least probabilistically how much energy would have to be consumed and whether or not it met
• 24:22
  a certain you know in this case difficulty target but ultimately a threshold for rate of block
• 24:28
  creation yeah yeah that is that a fair way to describe it like that something about this
• 24:36
  process is what ensures the integrity of bitcoin being a permissionless system yeah would it would
• 24:41
  another way to say it be something along the lines of like what is bitcoin saying about the world of
• 24:47
  energy because it knows very little so to speak but i do think it's it's saying something about
• 24:55
  the world of energy given the fact that it is and continues to just absolutely explode in hash rate
• 25:02
  is that another way to to kind of address the the question yeah i mean i think it's it's less so
• 25:10
  the because i do want to get to like how bitcoin and what bitcoin's role is in energy systems
• 25:17
  but um wanting to focus of anchoring someone in okay mining is doing these certain functions
• 25:25
  validating the currency transactions validating the fixed supply but in like the world of checks
• 25:34
  and balances of bitcoin if you have bitcoin holders and you have bitcoin miners and you have
• 25:38
  nodes in which you described that in the early days of bitcoin every node was also mining bitcoin
• 25:44
  and now it's more centralized yeah that um software alone doesn't dictate that there will
• 25:51
  only ever be 21 million and software alone doesn't dictate what isn't a valid bitcoin transaction
• 25:58
  that there's this objectiveness that all of these participants can be involved in this system but
• 26:07
  there's something anchored in the physical world that others can verify
• 26:11
  without needing to actually have any direct coordination between them and trying to connect
• 26:18
  that to you know is energy really critical to that system or could it be done another way
• 26:23
  so almost like could we change it if we wanted to or or or to think about it say like if we
• 26:31
  say that the energy use of bitcoin is not a waste and it's actually going to to a utility sure
• 26:38
  trying to test that assumption by saying if it's critical to the system then if it weren't then in
• 26:46
  theory of the system wouldn't work and if bitcoin only works because it's permissionless there must
• 26:51
  be a connection between anybody being able to plug into a power source um that other people can verify
• 26:58
  it's such an interesting way to think about it because the one thing that the one thing that
• 27:03
  really trips people up is that you you sort of spend your 10 000 hours learning bitcoin right
• 27:09
  and then once you start to get into bitcoin mining it's as if you've entered this small
• 27:14
  sub-niche that now, again, is a totally alien landscape. Because it's not really clear on its
• 27:23
  face what the incentives are, because the incentives of a Bitcoin miner don't even in all
• 27:29
  cases necessarily align with the incentives of the adversarial network that is Bitcoin with the
• 27:37
  software itself. And so there is this really interesting tension, I think, is what you're
• 27:42
  getting at that you don't even have to care about Bitcoin to mine Bitcoin. You know, there,
• 27:49
  there are reasons that you would mine Bitcoin that are entirely unrelated to your interest in
• 27:56
  Bitcoin or your, um, your desire to preserve the monetary network or to contribute to the
• 28:03
  monetary network in some way. And in this sense, it's a, I'll give you a concrete example is if
• 28:09
  you talk to a lot of folks who are early oil and gas adopters, oil field mining of Bitcoin,
• 28:18
  they were trying to solve a problem that is extremely hard to solve, which is that I might
• 28:23
  have a well that regulatory-wise is producing integrated gas, so a mix of natural gas and oil,
• 28:31
  but I can only sell the oil to market. That natural gas is essentially, it's a prohibition,
• 28:36
  it's a liability to me, but I can't get the oil out of the ground without getting the gas.
• 28:41
  Or maybe I have a tranche of wells that I own that are producing natural gas,
• 28:46
  but there's no pipeline nearby. There's no way to actually get it to market. So I have a stranded
• 28:51
  asset. There's a myriad reasons why an oil and gas operator would be looking at some set of
• 28:57
  minerals that they own and going, what the hell am I going to do with it? It's kind of, it's dead.
• 29:03
  It's a dead asset on my balance sheet.
• 29:05
  Bitcoin mining, in a lot of senses, has entered the equation for these operators as a way for
• 29:10
  them to do something that they know how to do, which is monetize minerals.
• 29:14
  And how do they do that?
• 29:15
  They can take this stranded product, this natural gas, or this waste product, or this
• 29:20
  liability product, this natural gas, and they can run it into a genset, into a generator,
• 29:26
  an engine.
• 29:27
  And when, if I consume it, I can turn it into electricity.
• 29:30
  If I turn into electricity, I can run a machine with it. It's one of the three inputs of the Bitcoin miner. And if I run a machine with it, I can hash and attempt to find a block and contribute in a sense to the Bitcoin network and be monetarily rewarded for it.
• 29:47
  it's really interesting because it's not necessary for a bitcoin miner to understand all of bitcoin
• 29:56
  to understand the utility that bitcoin mining might provide to them as a business operator
• 30:01
  totally unrelated to bitcoin however if you don't understand what your machine is doing
• 30:08
  you are incurring some secondary risks so you actually if you not fully educated in what the heck your machine is doing and if you look at it simply as a load bank
• 30:20
  kind of a dumb asset that just turns electricity into money,
• 30:24
  you actually pose a risk of killing the golden goose in a sense,
• 30:28
  of doing something that might actually be inappropriate
• 30:31
  for the health and function of the network,
• 30:33
  which we've been kind of teasing around this idea of,
• 30:37
  you know, growing centralization risks
• 30:39
  or growing censorship risks or growing collusion risks,
• 30:42
  which are very real in the modern state of Bitcoin.
• 30:45
  Yeah, actually, you touched on something there
• 30:47
  that maybe we could go in that direction for a second,
• 30:52
  which is somebody could mine Bitcoin
• 30:55
  and mine Bitcoin profitably,
• 30:59
  take electricity, convert it to hashes,
• 31:03
  propose blocks, solve blocks,
• 31:05
  validate currency transactions without,
• 31:07
  you know i i consumed x and i got paid y and y was greater than x and i and i made a profit
• 31:13
  in your mind can bitcoin miners really be competitive long-term and profitable long-term
• 31:25
  or effective efficient without understanding the money side of bitcoin the why of what they're doing
• 31:32
  Can they think about everything downstream from that, whether it be their basic strategy, their facility strategy, their energy strategy, if they don't see the field of Bitcoin?
• 31:44
  Maybe they can in the short term.
• 31:45
  Do you think that in how competitive the world of Bitcoin mining is and the ruthless competition to reduce costs or to identify better or worse energy sources?
• 32:01
  how do you like how do you think about that can can a bitcoin miner really get along for the long
• 32:08
  term without understanding it fundamentally no you you can understand the business of bitcoin
• 32:18
  mining from the perspective of the three connections of optimizing your um
• 32:26
  your physical setup, your cooling setup, and that gets you some efficiency gain,
• 32:32
  which increases the profitability of your business. You can optimize for your energy strategy,
• 32:38
  which may be the way that you source your energy or the way that you manage or process it or
• 32:42
  participate in some kind of curtailment or other ancillary service that might assist the operation
• 32:49
  of your business. If you don't fundamentally understand the value of what you're producing
• 32:55
  from the perspective of that third connection, the Ethernet cable, then you have a two-legged
• 33:00
  stool and you can't run a successful business in that way.
• 33:04
  So while you don't have to have an all-inclusive understanding day one of the Bitcoin network
• 33:12
  to immediately grok that this Bitcoin miner as a device is a technology that allows you
• 33:20
  to monetize what would otherwise be a liability if you're an energy producer.
• 33:25
  that's the easy thing to get.
• 33:27
  What do you mean by that liability?
• 33:29
  So energy has to come from somewhere.
• 33:31
  And some of the,
• 33:33
  some of the finest Bitcoin miners that I have met have been able to identify
• 33:37
  other businesses and other locations where energy is a liability.
• 33:44
  So we often think about energy strictly as an asset.
• 33:47
  I want to turn my lights on.
• 33:48
  And when I flip the switch,
• 33:50
  they better come on.
• 33:51
  So it's a thing.
• 33:53
  I have it,
• 33:53
  But we never think that through the connections of that grid, there is a business, there's
• 33:59
  a generator on the other side that has to take some kind of input and then profitably
• 34:04
  sell that energy to the grid.
• 34:05
  So we're not used to thinking of the energy we consume as somebody else's liability, but
• 34:11
  that's what it is.
• 34:12
  And especially sitting here in Texas on ERCOT, I think we have the best picture really in
• 34:18
  the world of what energy as an asset or a liability looks like because you have this
• 34:23
  free-floating mechanism in terms of price in the market. And the best way to see it is to look on
• 34:29
  ERCOT's website. And at any moment, you can see not only the total amount of load, the forward
• 34:34
  curve, what people are paying, what the price is at various nodes, but you can also see the
• 34:39
  distribution in the type of energy that's on the market. You can see how green at any given moment,
• 34:44
  how many megawatts of wind turbines are spinning, or how much photovoltaic is there.
• 34:49
  you realize that the grid is this incredibly complex mechanism that consists of generators
• 34:56
  which are fundamentally businesses trying to sell a product to the market and this is a free-floating
• 35:01
  supply and demand market and when you have a lot of supply that exceeds demand you run into issues
• 35:07
  and so something that you see quite often in particular pockets of the u.s but speaking just
• 35:12
  in reference to ERCOT in west texas you see a lot of photovoltaic and you see a lot of wind which
• 35:17
  kind of run in concurrent production curves often with each other, you know, notwithstanding
• 35:22
  if it's a cloudy day, what you see is that you get an oversupply.
• 35:29
  And so when you get an oversupply, the energy that you're producing at those sites becomes
• 35:33
  a liability and it becomes a liability because everyone is producing it on a hot, sunny day.
• 35:40
  Everyone wants to be dumping their electricity onto the grid.
• 35:43
  However, you have a transmission constraint.
• 35:45
  You can only get so many people lined up and queued up into the party at once.
• 35:50
  And so what does that do?
• 35:51
  That has the effect that it often drives prices negative in many points in time throughout the year.
• 35:58
  You see that quite regularly.
• 36:01
  So energy from the perspective of those producers is a massive liability because they want to get it to market.
• 36:07
  And if everyone else is trying to flood the wires to get it to market, they now have a resource which they may not fully understand.
• 36:13
  They might not understand the Bitcoin network, but they dang well understand that instead of having to pay the network or pay the grid rather to offtake this excess energy that they have.
• 36:24
  And that's when prices actually become negative.
• 36:26
  They have a buyer. And so, yeah, I mean, negative prices are something that occur routinely because of transmission.
• 36:31
  But you're not just talking about a liability in the concept where the energy prices are negative.
• 36:36
  You're talking about I have this resource, I have these costs input to it, and I need to monetize it. Is that fair to say?
• 36:42
  Fundamentally, yeah. A good example of that is what Gridless Compute is doing, which is a fantastic miner out of Africa, where they have found essentially all these producing hydro assets.
• 36:54
  So take a 10 megawatt hydroelectric dam that was a big project that somebody is way underwater on the financing of because they built this, you know, five, $10 million project.
• 37:05
  and now it's sitting in remote Kenya and the village that they thought was going to be receiving
• 37:11
  all this energy can't afford the prices. They have so much capacity, but they can't produce it
• 37:15
  because there's no demand. All of a sudden you have this essentially under monetized asset
• 37:20
  where an operator goes, wait, you're going to show up and on day one, you can buy 50%, 70%
• 37:29
  of my excess productive capacity that's currently not being utilized?
• 37:34
  Yeah. Okay. And then what's going to happen when the town wants to buy the price,
• 37:38
  wants to buy electricity from me? Oh, well, you're not willing to pay as much as them?
• 37:44
  What's a healthy market? Because it's a consumer that is buying somebody's liability,
• 37:49
  which is this excess capacity, this non-productive capacity. It's a 10 megawatt hydroelectric site
• 37:55
  that might be running at 20% capacity,
• 37:58
  they can start to spin it up to 70, 80% operational capacity.
• 38:02
  And they're willing to pay for that
• 38:04
  because Bitcoin miners will pay for their electricity
• 38:06
  or be paid if it's enough of a liability to the producer.
• 38:10
  But then they don't compete with the person
• 38:13
  who's living in the town next door
• 38:14
  because a Bitcoin miner isn't willing to pay as much
• 38:17
  as somebody who is in that village next door
• 38:19
  to that hydroelectric dam.
• 38:22
  And so maybe that's a good opportunity to talk about
• 38:25
  how do you think about the energy sources you just described one that was attractive and why
• 38:30
  but how do you think about the energy sources that are most attractive to mining bitcoin or
• 38:38
  what would be the best candidates and i i don't mean is hydro the best or is natural gas the best
• 38:45
  or is wind the best or is solar the best i mean that could be part of the discussion but
• 38:50
  what are the characteristics or properties of a site like proximity to a point of generation or
• 38:59
  scale and you know maybe in certain cases large-scale works but others that don't that
• 39:07
  or the balancing capacity to be able to balance flows and that might be you know upstream midstream
• 39:15
  or actually on a grid,
• 39:17
  how do you think about the opportunities
• 39:20
  of saying this energy source is better or worse
• 39:25
  than that for mining Bitcoin specifically?
• 39:29
  It's a really hard question to answer
• 39:36
  because structurally,
• 39:39
  what a Bitcoin miner wants to do
• 39:41
  is to drop their operating cost
• 39:43
  as low as humanly possible.
• 39:45
  Because my operational input, fundamentally the largest cost I'm going to incur is my cost of electricity.
• 39:51
  Would you say that the greatest differentiator amongst miners is that the energy strategy?
• 40:00
  It's probably what defines everything downstream from that.
• 40:07
  So here's what I mean by that.
• 40:09
  There are many, many sources of liability or waste energy in the world.
• 40:15
  And that is the direction that all miners who want to survive the next few halvings are pointing themselves in because they understand that the person that they are buying their main input from has to be feeling the pain of having this excess energy or electricity on their hands.
• 40:35
  So people have rapidly understood that.
• 40:38
  Now, what happens downstream of that identification?
• 40:40
  So where, let's think in concrete terms.
• 40:44
  You might think in terms of something like the Atapu Dam on the Brazilian Paraguay border, which has this immense, many, many gigawatts of hydroelectric generation, but total underutilization on the Paraguay side.
• 40:59
  That's a much smaller country than Brazil.
• 41:01
  They have 50 percent allocation to the production, but they don't use it all.
• 41:04
  Right.
• 41:05
  You might look at West Texas where somebody has an oil field where they can't get their gas to market.
• 41:11
  You might look at a photovoltaic company, a big farm, where they're going to have to wait three years to get a grid interconnect. So they're sitting there with all their PV built, but they have no way to get it to market.
• 41:25
  you might look at a coal plant in Wyoming that has, for whatever reason, been operating at 30%
• 41:33
  of its actual operational capacity. Now, as far as coal plants are concerned with thermal generation,
• 41:40
  coal plants are interesting because they get a lot more inefficient if they're not running within a
• 41:45
  very specific band of production. So it actually becomes marginally more expensive for them to
• 41:50
  produce if they are producing too much or too little relative to what's called the heat rate,
• 41:55
  where they want to be producing, kind of like the optimal speed to be running your engine on the
• 41:59
  highway to get the optimal engine efficiency. So these are all examples of production or sources
• 42:07
  of raw energy where indigenous to that business is energy as a liability. They're trying to produce
• 42:13
  it, but they may have too much. It may be in the wrong place. It may not be operating in the right
• 42:18
  kind of way. What does the Bitcoin mine look like that serves that particular excess? It's very
• 42:26
  different. So if you're connected to something like the Atapu Dam, you can almost always expect
• 42:32
  to be operating at a certain capacity. Why? Because you have pretty good predictive measures around
• 42:39
  how you're going to be able to run your turbines. It's a river and you kind of understand there's a
• 42:43
  seasonal ebb and flow, but there's very likely not going to be some insane, you know,
• 42:48
  drop off one day, you know, notwithstanding some cataclysmic thing happening, you kind of know what
• 42:52
  you're going to get. So it's very steady baseload power. So you can build a warehouse, you could build
• 42:56
  a facility, you could put some fans, you could put some immersion. Okay, I can do that. You know,
• 43:00
  I'm next to the dam. When I go to the oil field, it's a different game. And it's a different game
• 43:05
  because the actual gross volume of energy, the total energy that you can get off a given well
• 43:12
  site is infinitesimal compared to what you would get off of this gargantuan hydroelectric dam.
• 43:20
  So what does it mean? It means that your footprint is now something like a sea can.
• 43:25
  It's a much smaller physical setup. And it looks radically different from what you might have in a
• 43:31
  warehouse, but it's serving a need in its own way for an energy producer. If you go to the
• 43:39
  the photovoltaic side of things and somebody is producing, um, producing a non-dispatchable load
• 43:45
  to the grid or they're not connected to the grid yet. And I'll do this quite non-dispatchable.
• 43:51
  Um, essentially, um, grids don't just consist of wires that kind of passively
• 43:56
  shuttle electrons between producers and consumers, right? Um, they're managed,
• 44:02
  they're managed by a grid operator and a grid operator has a mandate essentially to make sure
• 44:08
  that the lights are kept on at all points in time. There's no blackouts. There's no brownouts. We
• 44:12
  have a robust, reliable energy always being produced and electricity is always going out to
• 44:19
  everybody. And it's always coming at some kind of non-prohibitive price, right? So electricity isn't
• 44:24
  consistently, you know, a thousand dollars a megawatt, which is blowing up everybody's ability
• 44:29
  to actually keep the lights on. That grid operator has a number, a number of resources at their
• 44:35
  disposable at their disposal, one of which is the dispatchability of load, which is, can I call you
• 44:43
  up and ask you to produce more for me? And so in the case of something like a peaker plant or a
• 44:48
  coal plant, or even something like a Bitcoin mine, I can call a Bitcoin mine and I can say, hey,
• 44:54
  remember how I gave you half a cent off your rate? And I told you I would call you every once in a
• 45:00
  while and tell you to shut down because I need to send the power to somebody else. You know,
• 45:04
  it's the middle of August and everyone's running their AC, that's dispatchable.
• 45:08
  I can call you and I can say like, turn off.
• 45:11
  I need you.
• 45:13
  Yes or no.
• 45:14
  I can do that for a certain class of renewables like wind and photovoltaic because they totally passive systems They totally subject to the environment The wind is blowing or it not blowing The sun is shining or it
• 45:28
  not shining. And so from that perspective, these certain renewables are considered non-dispatchable
• 45:33
  in that they're always just trying to produce as much as they can opportunistically at the time
• 45:37
  that they're receiving that energy. They're able to produce electricity. And so I cut you in the
• 45:43
  middle of that, but you were using that to describe how, you know, one of those sites might.
• 45:49
  How the footprint looks very different. How it's not obvious from physically looking at a Bitcoin
• 45:54
  mine that they're all really revolving around this core, solving this core problem, which is
• 46:00
  energy as a liability. If you look at a big, a big solar farm and you say, okay, well, I don't know,
• 46:07
  does that really make sense to have a solar farm? You think about it, the sun's only up half the
• 46:10
  time. I don't know. Production maybe isn't great. When you look at it from that perspective,
• 46:17
  it looks a little bit prohibitive. But if you start to look at it from the perspective of,
• 46:21
  say, a mega miner, and I think one of the best examples is actually Mara who did this.
• 46:28
  They bought a, not a solar farm, but a wind farm. And on its surface, you go, why would they buy a
• 46:35
  wind farm? I thought that they were doing, I thought they were doing this like 24 seven,
• 46:39
  keep it up all the time. It doesn't make any sense to me, but it makes perfect sense when you think
• 46:44
  of it in terms of what a mega miner is doing. A mega miner will always be depreciating and
• 46:51
  eventually disposing of the most recent tranche of machines that they've bought. And they're
• 46:57
  spending hundreds of millions of dollars on these newest allocations of machines because it's an
• 47:02
  easy way from a capital markets perspective, from a publicly traded markets perspective,
• 47:06
  to go and to say, okay, well, we're going to buy the newest efficiency machines. We're going to
• 47:11
  replace the old machines, maybe do some retooling in our facilities. And look at that. It's an
• 47:15
  instant profitability bump because now I'm able to, through this efficiency gain, get more work,
• 47:20
  which has value out of the same amount of electricity. Well, there's something interesting
• 47:25
  that you can do, which is that all those machines that you've bought and have depreciated,
• 47:29
  instead of sending them to e-waste and feeding the trolls who tell you that bitcoin is the worst
• 47:37
  thing to the world you can now send them somewhere else where they are able to monetize something
• 47:43
  but in a different way so if i'm going to be running 24 7 i want to have the most efficient
• 47:49
  machines i want to have the most expensive machines because my incentive is to be just
• 47:55
  grinding and hashing and monetizing the heck and paying off those machines as aggressively as i
• 47:58
  possibly can. But if I think about it from a balance sheet perspective, once I've depreciated
• 48:03
  that machine and it's not worth anything to me anymore, I can keep monetizing it, but I have to
• 48:09
  put it in an environment where I don't care as much about the raw profitability. So if I buy a
• 48:15
  wind farm, I get this intermittence in terms of the production. I get a lot more variability in
• 48:21
  terms of when I can produce or when I can't, which is totally unacceptable if I have the newest
• 48:26
  machine. If I have the, if I have the newest Ferrari, I'm going to be racing it every day on
• 48:30
  the track. But if it's the old one, where can I put it that I can continue to use it for the next,
• 48:35
  the rest of its useful life. It's not worth anything to me anymore. As far as, uh, as far
• 48:40
  as its value on my balance sheet, cause it's been depreciated, but I can still, I can still monetize
• 48:44
  it in a different way in a different setting. Now, does that not have, doesn't that have
• 48:51
  something to do with the cost of power though or maybe say said a different way or it's still an
• 48:58
  asset even if it's depreciated it's still an asset that can run yes it's just not going to
• 49:04
  be as efficient at consuming power correct so why maybe pull you know go a little bit
• 49:15
  deeper on that thread of like why a less efficient machine is acceptable in that environment yeah
• 49:23
  the best analog for this is the way that uh this new class of miners heat punks home heating miners
• 49:31
  have emerged in the market there is a cost premium to buying the newest and the greatest bitcoin
• 49:38
  market, a Bitcoin miner, a massive cost premium or price premium. That's not worth it for me.
• 49:46
  If I have intermittent operation for my site, it's not worth it. It doesn't make sense to buy
• 49:53
  the newest and best thing and only use it 50% of the time. It makes sense for me to buy the newest
• 49:58
  and best and most efficient thing and run it continuously because I'm recapturing what I'm
• 50:02
  paying essentially in that, in that price premium. I'm recapturing in the fact that I'm, I'm able to,
• 50:07
  I'm able to recover that in the additional profitability gains that I get running at 24-7.
• 50:15
  You have this secondary market where, and you shouldn't even really call it a secondary market.
• 50:20
  I don't know maybe the right way to classify it, but you have a whole class of use cases where
• 50:28
  raw monetization is not your primary motivation because if I buy the lesser efficient machine,
• 50:39
  I have a huge cost reduction in terms of what I'm paying to get that machine.
• 50:45
  So effectively, I'm willing to put up with the fact that I'm going to get less efficiency because
• 50:51
  I'm getting some other utility out of it because I don't want to pay. I can't justify it. I can't
• 50:57
  can't justify paying for the newest, greatest thing. So in the world of home heating,
• 51:01
  what a lot of people have found is that we have a sunk cost, a built-in sunk cost in most of our
• 51:09
  homes, which is coming in the result of some kind of electric or gas heating. And you get it every
• 51:16
  month, you pay it. Sometimes it goes up, sometimes it goes down, it changes seasonally. But what you
• 51:22
  can do with Bitcoin mining that's very interesting is that if you take this machine and if it's cheap
• 51:28
  enough as it is with a depreciated or used or last generation machine, I can introduce it to
• 51:37
  my home environment where it might not actually be profitable to run a machine on my residential rate.
• 51:45
  But I get something in addition that makes sense for me, which is that I can generate utility
• 51:51
  because the Bitcoin mine, we said there's three connections.
• 51:53
  We said there's electricity, Ethernet, and cooling.
• 51:56
  There's thermodynamics.
• 51:58
  Well, I can cool it, but what that means is that the machine is producing heat.
• 52:02
  Well, I'm already paying $100 a month heating my water tank or heating my house.
• 52:08
  What if I can now offtake from that liability,
• 52:13
  that sunk cost that I have sitting in my house,
• 52:15
  and I can recapture it through some kind of monetization in Bitcoin mining
• 52:20
  and then generate some utility, which is I get the heat because I know the machine is going to
• 52:24
  produce heat. So there are these very, very, I wouldn't even call them edge cases because I think
• 52:30
  they're growing and I think they're going to continue to grow, but very nascent cases or
• 52:33
  nascent use cases where people are starting to grok this, that if I get an older model machine
• 52:41
  and it is so much cheaper than the newer model machine, I don't have to care as much about the
• 52:45
  raw efficiency of my machine's operation because i'm actually getting some kind of utility from the
• 52:52
  thing that the machine does which is solving some other problem which that's right which is i need
• 52:57
  to be hot anyway so on the on the side of heat it could it could be using a waste byproduct to
• 53:05
  to turn it into some other utility or reduce a cost it could reduce create a redundancy or reduce
• 53:11
  risk right it's like i've got one heating system this could be the second heating system
• 53:15
  but in the in the i don't say traditional mine because all mining is valid in the mine
• 53:22
  in people that are running mining as a business
• 53:24
  you know with other with other businesses traditionally people think of scale
• 53:33
  you know the greater the scale i can perform this operation at the cheaper i'm able to
• 53:39
  produce a marginal widget yeah but based on how you described of bitcoin mining really being
• 53:47
  not necessarily dependent but but whether it fits being um necessarily observant or
• 53:57
  planned around other uses of energy around it um that might be the you know description of grid
• 54:04
  list that you gave of of power not being used it might be um the the natural gas being a byproduct
• 54:12
  of of oil and needing to to do something with that to be able to to increase the amount of oil you can
• 54:17
  get out of the ground that can a a miner that's mining with 150 kilowatt generator on a well site
• 54:30
  versus a 500 megawatt mine,
• 54:33
  can miners be equally profitable
• 54:35
  on a unit basis at different scale?
• 54:38
  That's such a good question.
• 54:40
  That's such a good question
• 54:41
  because this is like,
• 54:42
  this is Bitcoin 101,
• 54:44
  something that people misunderstand
• 54:45
  about Bitcoin mining,
• 54:47
  which is that the economies of scale
• 54:50
  in Bitcoin mining
• 54:51
  are only proportional
• 54:54
  when it comes to that Ethernet connection,
• 54:56
  the actual work that the machine is doing.
• 54:58
  where you could start to generate economies of scale are in the way that you operate other
• 55:04
  aspects of your business or the way that you have utility through the heat or the cooling of your
• 55:08
  systems, right? That's such a good question because fundamentally, if you add more hash
• 55:14
  rate to the network and say that you're being rewarded, you're being paid for that hash rate,
• 55:18
  if I add more hash rate, well, that's a linear progression. I don't get an extra five cents
• 55:28
  plus by adding an additional megawatt of load for the network. Because the way that hashing works
• 55:35
  is that it's proportional, you know, proportionally over time, if you are 10% of the network,
• 55:41
  you will win approximately 10% of the blocks. And if you add another, if you grow yourself by 50%
• 55:48
  and you're now, well, you wouldn't be 15% of the network because now the network would also be that
• 55:52
  much bigger. You're actually inadvertently by growing, making it a little bit harder on yourself
• 55:58
  But that's what everybody else is doing as well. Right. So there's there's extreme competition on this side that does not actually give you an economy of scale when it comes to the hashing side of things.
• 56:08
  would you say that on average then just kind of keying in on that point of that example say
• 56:14
  there's 30 gigawatts of power scaring the bitcoin network and say you you have just round numbers
• 56:23
  one gigawatt if you increase you know if you bring another gigawatt online which obviously
• 56:27
  is a massive amount of power i'm just using this description to have numbers be round but
• 56:33
  now there's 31 gigawatts that you're essentially um in order to justify that next unit of power
• 56:42
  that's creating hashes because it now and it's dependent on how much people are valuing bitcoin
• 56:50
  and that has a relationship to your bitcoin denominated power costs but from a relationship
• 56:56
  perspective and over time does it create the incentive that each marginal unit of power that's
• 57:04
  being consumed be cheaper not 100 of them but just directionally because you're you're effectively
• 57:12
  diluting increasing the cost base the cost structure or the hash value because there's
• 57:21
  there's another piece there, which is that you can never draw a direct equivalence between the
• 57:27
  input energy and the hashes that are generated, right? Because it's always moderated by the
• 57:33
  efficiency of the system that is turning that electricity into hashes. And it varies really
• 57:41
  widely. You know, we, we measure it industry-wise in what we call joules per tera hash, right?
• 57:46
  which is um if i put in this number of joules how much work can i squeak out of it what most
• 57:53
  people are familiar with for people that aren't steeped in in energy most people associate energy
• 57:59
  with a kilowatt hour because that's what electricity is filled out at home just connection
• 58:05
  between a joule and a watt rough rough equivalents but but functionally different uh in terms of what
• 58:12
  you're referring to the work being done. The way that you could think about it is essentially the
• 58:19
  sense that every machine has an effective total amount of electricity that it will consume.
• 58:26
  And that is most often labeled just in kilowatts, right? And so it varies really, really widely.
• 58:33
  You have things that are as small as like a bid X, which might be, you know, 15 to 30 watts.
• 58:39
  and then you have machines that go up into the thousands or many thousands of watts of kilowatts
• 58:45
  of energy and the consumption of that is joules of work. So the idea in joules per tera hash or
• 58:53
  work per tera hash is that you fundamentally want to be getting more work for the same or less
• 59:00
  energy that you're using, electricity that you're using. So if you look at an older,
• 59:07
  older model machine. So something like an S9 that would be in like the 80 or 90 joules per
• 59:12
  tera hash. That was considered incredible by the standards of, uh, of 2015, 16, 17,
• 59:20
  where you come today is where now we're in a world where we're in the low teens and folks are,
• 59:26
  are actually estimating now that they're going to bring things to market that are in the single
• 59:30
  digits and like the nine point something or 10 joules per tera hash, which is a ninefold increase
• 59:35
  over the course of, you know, eight to 10 years, which is incredible.
• 59:41
  Essentially, you can have the same energy asset or the same source of electricity.
• 59:49
  And you've now made it nine times as productive because you have the state-of-the-art machine
• 59:54
  that is now consuming that electricity to perform that much more work with the exact same footprint.
• 1:00:00
  Right. So that's another variable.
• 1:00:02
  but say you were running the exact same so it's like there could in theory be a
• 1:00:06
  world where the machines are becoming more efficient you might have access to
• 1:00:12
  them earlier and that allows you to say well I could be actually less efficient
• 1:00:16
  in you know my power cost because it's being offset by greater gains and how
• 1:00:23
  much work can actually be done by each unit of power but say I was running you
• 1:00:28
  know the exact same
• 1:00:29
  same machine but multiplying it by two yeah and i'm consuming twice as much power i'm producing
• 1:00:38
  twice as much work but that doesn't give me an additional marginal uh earning it actually gives
• 1:00:44
  me proportionally a little bit less because i'm growing the total network right and in that world
• 1:00:48
  assuming kind of at least looking at the from the energy perspective and setting aside advancements
• 1:00:56
  and efficiency of work per joule over time i'm incentivized to not dilute myself by bringing on
• 1:01:05
  higher cost energy but reducing my costs yes oh i see what you're saying so effectively
• 1:01:11
  because i know it's going to be more difficult over time i'm strongly incentivized to now
• 1:01:17
  reduce my operational cost because that is part of the equation that i can control and yeah
• 1:01:22
  fundamentally this is um there's very little that bitcoin miners are able to control when it comes
• 1:01:29
  to the bitcoin network um now we have choices that we can make as as to whether you are a miner or a
• 1:01:36
  hasher whether you're willing to take on the responsibility of running a node or making your
• 1:01:40
  own templates um but because you are a big miner does not necessarily mean that you are able to
• 1:01:47
  sway the direction of the protocol. Even if you had five to 10% of the network and you're making
• 1:01:54
  all your own blocks, so you're deciding what's going to go into a block that you're going to
• 1:01:58
  work on and add to the blockchain and be rewarded for, you know, you're incentivized to not kill the
• 1:02:03
  golden goose. So you are assumed to be profit seeking, that you're going to look at the
• 1:02:10
  transactions submitted in your mempool and you're going to pick the most profitable ones because you
• 1:02:14
  want to be as profitable as humanly possible. And you already know what the block subsidy is.
• 1:02:18
  So you're going to combine that into something that's going to give you the most profitable
• 1:02:23
  block that you can generate. And that's how you're going to feed your revenues. You know,
• 1:02:27
  you're not incentivized to be contentious. Now, that being said, there's some really
• 1:02:32
  interesting analysis about an attack vector called selfish mining, which Antoine Poincault,
• 1:02:39
  I don't know if I'm pronouncing his name correctly. He did a really interesting analysis
• 1:02:43
  on breaking down the mathematics behind it.
• 1:02:46
  So there are things that you can do,
• 1:02:48
  mostly in the adversarial sense.
• 1:02:51
  But that being said,
• 1:02:52
  you're always comparing it against the fact
• 1:02:54
  that you don't want to kill the golden goose.
• 1:02:56
  You really have a very strong incentive
• 1:02:58
  to not kill the thing that is your source of revenue.
• 1:03:01
  And that idea of not wanting to kill the golden goose
• 1:03:04
  goes back to a concept that we talked about earlier.
• 1:03:08
  I need to know what the golden goose is
• 1:03:10
  to be able to evaluate your own self-interest.
• 1:03:17
  Like you have to understand power and energy
• 1:03:22
  to have a strategy that will allow you
• 1:03:26
  on a relative basis to be more competitive than your peers.
• 1:03:29
  Yes.
• 1:03:31
  You'll need to understand your function
• 1:03:35
  and how you're consuming power
• 1:03:36
  in relation to how power is being consumed around you.
• 1:03:39
  That's right.
• 1:03:39
  Yeah.
• 1:03:40
  your local local business conditions or maybe a way to say it yeah but then with this concept of
• 1:03:48
  marrying it's like miners can never and should never be assumed to be altruistic correct right
• 1:03:57
  they're always going to be acting in their own self-interest
• 1:04:00
  and and kind of zooming back out so we got you know deep down the energy rabbit hole of bitcoin
• 1:04:09
  and talking about what different uses of energy or sources of energy might be
• 1:04:17
  more effective or less effective and it's a dynamic system based on the specific energy system
• 1:04:23
  kind of re-anchoring people in and all of all of that all that energy regardless of who it is
• 1:04:29
  what scale it is what what fuel source what type of miner they're all doing one thing yeah
• 1:04:38
  securing the fixed supply of bitcoin validating currency supply and at the same time currency
• 1:04:46
  transactions that's the golden goose when you describe is that that fair when you say
• 1:04:53
  you can't be screwing up the golden goose or are you just are you describing something else
• 1:04:58
  yeah it's the it's the fidelity of the of the peer-to-peer network and so one of the things
• 1:05:05
  that puts or potentially puts the fidelity
• 1:05:13
  of the peer-to-peer network at risk is centralization.
• 1:05:18
  And it's not centralization specific to mining.
• 1:05:21
  It could be centralization in any aspect of Bitcoin.
• 1:05:25
  But part of the goal of the podcast is to
• 1:05:28
  go down the Bitcoin rabbit hole on the energy side,
• 1:05:32
  help connect the importance of money to the relationship of both mining and the consumption
• 1:05:38
  of power but another side of it is helping to distribute knowledge with the goal of furthering
• 1:05:46
  decentralization in a part of bitcoin that is either centralized today and has been for a while
• 1:05:58
  or has become more centralized over time.
• 1:06:02
  First really being the manufacturing of mining rigs.
• 1:06:06
  Yeah.
• 1:06:07
  Second being the function of pools.
• 1:06:11
  Describe in your words or from your perspective,
• 1:06:16
  the state of centralization today in either of those two,
• 1:06:21
  to kind of give people,
• 1:06:23
  Marty and I talked about it a little bit on the first episode,
• 1:06:26
  but um how centralized those two functions of bitcoin are what the what the risks are but then
• 1:06:37
  also secondarily someone who's aware of those risks why that doesn't prevent you from saving a
• 1:06:48
  irresponsible amount of your wealth in bitcoin or why it wouldn't or shouldn't dissuade someone
• 1:06:54
  who was thinking about building a new bitcoin mine or investing capital in a mining project
• 1:06:59
  or otherwise just saving in bitcoin of like it's centralized describe the current state
• 1:07:03
  of centralization but then what also makes you more comfortable or still participating
• 1:07:12
  and why yeah the so the fundamental question is is a yes no question
• 1:07:19
  does nakamoto consensus work but most people highly technical people will understand nakamoto
• 1:07:27
  consensus does does the does the design and the incentive structure of bitcoin as an adversarial
• 1:07:34
  network right as this this peer-to-peer network that is permissionless and decentralized does
• 1:07:42
  do the rules of that game do they work and are they self-enforcing yeah do they work
• 1:07:49
  and if they don't work if you don't think that the design of that system of bitcoin
• 1:07:56
  is functional if it really is permissionless decentralized and anti-fragile in this kind of way
• 1:08:05
  then no then you should you should get out but fundamentally you know i haven't made that choice
• 1:08:12
  and if anything i've continuously doubled down on the exact opposite of that thesis which is that i
• 1:08:18
  I perceive Bitcoin as the safest and most anti-fragile, most survivable thing in terms of places that I can store my valuable earnings, my wealth as an individual.
• 1:08:34
  And it can also be true in addition to.
• 1:08:39
  Actually, let me say it in a different way.
• 1:08:40
  So does the incentive system work a Bitcoin?
• 1:08:43
  No, then you shouldn't participate in it.
• 1:08:45
  Yes.
• 1:08:46
  Well, then that's a base level understanding.
• 1:08:47
  If it does, then yes. Can it also be true that the current state of Bitcoin is one that is highly centralized and baked into that centralization are a series of additional very real risks?
• 1:09:02
  Yes, that can be true. What does that look like in reality?
• 1:09:07
  So fundamentally, the argument about two forms of miner centralization, one is physical infrastructure, and then the second one is pooled mining.
• 1:09:19
  So the physical infrastructure, it's an interesting one because where we are in this moment in Bitcoin mining is more competitive than we have ever been in the production of ASICs.
• 1:09:34
  basics so fundamentally the highest or most performant machines that we have really come
• 1:09:40
  out of a single manufacturer and a single manufacturer who is not necessarily the best
• 1:09:45
  actor in the space historically which is who bitmain but they make the most performant machines the
• 1:09:52
  most affordable or the most performant um efficient machines on the market what percentage of the
• 1:09:58
  market would you say they have easily uh 80 plus and then what is the next largest have
• 1:10:04
  i would guess that it's what's minor um 10 15 so the top two at 90 to 95 just like rough
• 1:10:17
  directional yeah and that being said um you know the birth of what's minor came from a former
• 1:10:22
  Bitmain engineer. And then you have additional, you know, seal miners, you have Avalon, Avalon
• 1:10:30
  mining with Canaan. You have a whole series of upstart manufacturers, machine manufacturers
• 1:10:37
  who are now becoming more competitive in the market. So while it is true that it is a highly
• 1:10:42
  centralized space right now in terms of the production, we are also in a moment where the
• 1:10:48
  broader industry recognizes that that is the case. They recognize the associated risks,
• 1:10:55
  not simply the production risks, but the geopolitical risks. The fact that, you know,
• 1:11:00
  like or dislike, it's a foreign nation that is producing the vast majority of these machines
• 1:11:06
  that are running in the US. And if you want to make more robust supply chains, it makes sense
• 1:11:14
  to use your local brain power to develop this technology internal to your country. Now, that
• 1:11:20
  being said, we are also baked into the broader chip manufacturing market, which is extremely
• 1:11:24
  competitive because there's a whole industry of having to get allocations for foundry space and
• 1:11:31
  having to pay and having to have a reputation. And you're not just competing with other ASIC
• 1:11:34
  manufacturers, you're competing with NVIDIA because there are only so many foundries. So
• 1:11:39
  everything is really bottlenecked from accessibility to foundries. And that is a really well-known
• 1:11:46
  geopolitical risk. So it's not the case that Bitcoin mining is uniquely exposed to this. This
• 1:11:51
  is a problem that exists across every aspect of the technology spectrum. There's that aspect, but
• 1:11:59
  I would say that even though it is highly centralized in practice,
• 1:12:04
  because so many people have become so aware so quickly as to the nature of the centralization
• 1:12:10
  and everyone recognizes that it's a problem you've had it in an immense upswing in new businesses
• 1:12:17
  new folks making uh making their own chips doing their own designs which while they may not be as
• 1:12:23
  competitive as bit may not the gate is absolutely acceptable and desirable because what they are doing is it is broadening the brain trust of the industry outside of a single company that exists outside of the borders of the United States right
• 1:12:39
  So that being said, in addition to that, we also have a suite of open source initiatives
• 1:12:44
  that have been effectively reverse engineering a lot of these closed source tools and reintroducing
• 1:12:51
  them to the market now as open source tools through the 256th Foundation.
• 1:12:56
  so that has been a very early initiative that has gotten a ton of traction which is some very very
• 1:13:03
  very intelligent people behind it building a series of open source tools so while yes we are
• 1:13:09
  very centralized in terms of the the production of the physical machines the response to it has
• 1:13:15
  been swift it's been extremely decisive it's been extremely diversifying and it's been oriented
• 1:13:21
  around, in a lot of ways, open source and new technologies that I think is net very positive
• 1:13:27
  for the industry. So that's the physical side of things. While it looks scary on the surface,
• 1:13:33
  I think the moment that we're in right now is extremely optimistic because a lot of people
• 1:13:37
  are realizing very immediately that they do not want their sole provider for the machines that
• 1:13:44
  they have running in their farms to be coming from somebody who is a pen stroke away from having
• 1:13:50
  a 30, 40% tariff put on them. So it's a very strong incentive to repatriate some of that
• 1:13:56
  brain trust or that knowledge and build locally. Now that's one aspect of centralization.
• 1:14:01
  The second aspect of centralization is what you touched on, which is this idea of pooled mining.
• 1:14:07
  And Bob Burnett says it better than anybody else when he says that some people are mining Bitcoin
• 1:14:11
  and some are hashing. And what he's referring to is that, you know, when you're mining Bitcoin,
• 1:14:16
  where we exist today in the world of pools well what is pooled mining to understand what the heck
• 1:14:23
  pooled mining is you have to understand where it came from in 2010 and and slush of slush pool now
• 1:14:30
  brain's pool was the first individual on the bitcoin talk forums to successfully launch a
• 1:14:35
  a pool pooled mining why did he do that he did that and he says in the original post
• 1:14:42
  are you basically sick of having your lunch eaten by gpu miners and this was the very early days of
• 1:14:50
  bitcoin so you either mined on a cpu or you mined on a gpu there there weren't fpgas there weren't
• 1:14:57
  a6 there was nothing the the s9 was some distant future place right he says in his initial post
• 1:15:03
  are you sick enough finding blocks anymore here's what you can do we can all work together and pool
• 1:15:09
  mining. And what that means is that instead of winning a block reward, because it used to be
• 1:15:16
  that you would just run your computer, win a block reward, but then the GPU showed up,
• 1:15:20
  they started eating everybody's lunch and they added proportionally a lot more hash power to
• 1:15:25
  the network. So it became very hard for a single CPU to, um, to win a block reward. He said, if we
• 1:15:31
  work together, we will in aggregate act like a much bigger Bitcoin miner and we will proportionally
• 1:15:38
  hit more blocks. And so instead of getting the full block reward, which was 50 Bitcoin at the time,
• 1:15:44
  how about I just give you a proportion of what you're contributing to the pool? So if there's
• 1:15:50
  going to be 10 of us contributing and each of us has a 10% share, assume we all have the same
• 1:15:55
  computer, we act like a computer that's 10 times our size. And so now we hit blocks more often.
• 1:16:01
  That's great. What that means is that instead of getting the 50 Bitcoin reward, if I contribute 10%,
• 1:16:08
  we hit a block, I get five Bitcoin. Hey, that's a good deal. And it solves a really interesting
• 1:16:14
  problem, which is a cashflow problem. I don't want to run this machine ad infinitum because
• 1:16:21
  it costs something to run it if I'm never going to be rewarded for running it. So pooled mining
• 1:16:26
  solved this really tough problem at the time, which was we're being outgunned and we're not
• 1:16:32
  getting enough revenues to justify us running these computers forever. And what it grew into
• 1:16:37
  was a whole suite of financialization let me can i ask a question sure that makes sense of like
• 1:16:44
  directionally why bitcoin mining has evolved into pools and in a certain way why you know it describes
• 1:17:00
  why there might be a centralizing function around pools but if if more hash rate coming online that's
• 1:17:08
  more efficient and you mentioned going from a cpu that would have been just a standard computer to a
• 1:17:14
  gpu now to an asic that it would seem that nothing about a pool would change your profitability
• 1:17:21
  it might
• 1:17:22
  reduce
• 1:17:24
  variance
• 1:17:26
  right but if
• 1:17:28
  you're mining on a less efficient
• 1:17:31
  machine just joining with others
• 1:17:33
  doesn't
• 1:17:34
  make you more
• 1:17:36
  less profitable
• 1:17:37
  but then where I'm going is that
• 1:17:41
  you go down the line to where we are today
• 1:17:43
  where
• 1:17:44
  everybody is
• 1:17:46
  generally running or the vast majority of hash rate
• 1:17:48
  on the network is running ASICs of some form or fashion
• 1:17:51
  that older races can be replaced by newer races that have better efficiency that
• 1:17:58
  pooling doesn't improve profitability it reduces variance but if we get to this world where
• 1:18:05
  there's a more level playing field why and maybe describe you know as you did you know kind of
• 1:18:12
  on a percentage basis why is the function still seeming to continue to centralize despite
• 1:18:19
  everyone kind of being on the level playing field of running ASICs.
• 1:18:25
  Yeah.
• 1:18:27
  The simplest reason is because pools have realized that
• 1:18:33
  running a pool is not a profitable business.
• 1:18:37
  So,
• 1:18:38
  or rather another way to put it is that it's an extremely competitive
• 1:18:42
  business.
• 1:18:43
  Cause you have to think about it like this.
• 1:18:45
  I'm going to have a bunch of people point their hash rate to me.
• 1:18:49
  Here's the trade-off that happens when you pool mine, notwithstanding the introduction of something like an ocean where you're making your own templates.
• 1:18:58
  But in kind of standard pool mining, there's a trade-off.
• 1:19:01
  So me, as a miner, I have machines.
• 1:19:06
  And those machines I want to monetize.
• 1:19:08
  And I want to reduce my variance, which is that I want more steady cash flows.
• 1:19:12
  so what I will exchange for those steady cash flows is I will point you my hash rate
• 1:19:20
  and instead of me having now the right and responsibility to act as a single node on the
• 1:19:27
  network and to build my own templates and to hash on them I'm now going to give that responsibility
• 1:19:34
  to you. And in exchange for that, you're now going to take the responsibility of aggregating
• 1:19:42
  these transactions in your mempool. And you're going to issue out work to all the participants
• 1:19:49
  and try to optimize a strategy by which we can hit blocks as efficiently as possible,
• 1:19:53
  because we're all aligned in terms of our incentives. We want to make money.
• 1:19:57
  But the trade-off is that the pool now wields an additional sort of power on the network
• 1:20:03
  that the individual who is trying to reduce their variance
• 1:20:08
  has just kind of given up.
• 1:20:10
  They've kind of given the pool their influence of their hash power,
• 1:20:14
  their hash rate by proxy, right?
• 1:20:17
  They've allowed a representative now to serve as kind of their node on the network.
• 1:20:23
  Now, pools are incentivized to want to grow their hash rate, right?
• 1:20:29
  This has happened historically in Bitcoin.
• 1:20:31
  g hash did this where it can grow to a point to being anti-competitive where participants realize
• 1:20:38
  that oh darn if i go and i join this pool it's actually not good for bitcoin because it creates
• 1:20:44
  a centralization risk so it's happened before in the history of bitcoin but it's not just it's not
• 1:20:52
  good for bitcoin it's also not good for the individual might or the individual who was making
• 1:20:58
  this exchange to reduce their variance, but now so many people have done that through a single party
• 1:21:02
  that it's beginning to create a liability for the sake of the network. So it returns to this idea of
• 1:21:07
  you don't kill the golden goose, right? Even if your mining pool is not behaving inappropriately,
• 1:21:15
  you as somebody who is allocating your hash rate to a pool is still conscientious of
• 1:21:22
  the risk that you might be incurring by giving your hash rate to somebody who's already large.
• 1:21:28
  So this is where the pool story gets interesting, is that pools figured out that they could, if they built up a large reserve of Bitcoin, find creative ways to reduce variance even more for participants in the network.
• 1:21:45
  So in exchange for me giving you my hash rate and you telling me what to work on, I've given up my responsibility and right to build templates that contribute to the tip of the chain, right?
• 1:21:57
  In exchange for that, we have payout strategies that have come to the point where we have reduced, we've reduced variance so much that FPPS, full pay per share, has now essentially removed the actual day-to-day block finds of a pool from the equation.
• 1:22:19
  So I can sign up for certain pools and know what I'm going to make day in, day out, where even five years prior to that, I would have to actually wait for that pool to hit blocks and I would be rewarded some proportion of those block finds that we had.
• 1:22:36
  So FPPS has been described by some people as basically crack for miners, because once you get it, it becomes extremely hard to get rid of, which is this kind of constant payment, constant ability to monetize your hash rate.
• 1:22:51
  But it creates a massive centralization risk and it fragilizes the network in ways that don't break the incentive structure, but create a risk for censorship and create a risk for centralization in the network.
• 1:23:09
  So fundamentally, what pools look like today, these large FPPS pools, is that they have giant, giant treasuries of Bitcoin where they are now running this equation of I have this much hash rate pointed at me at any given moment.
• 1:23:25
  And I am going to pay those people continuously for contributing that hash rate, whether or not we find a block.
• 1:23:34
  which is interesting because the pool can only monetize by finding blocks so statistically they
• 1:23:41
  have some rough guess about based on how big they are relative to the network when they should be
• 1:23:46
  hitting blocks but there's still going to be an element of luck there's still going to be variants
• 1:23:50
  from day to day so by them having this huge bitcoin treasury this huge bitcoin reserve
• 1:23:55
  they're essentially now playing this kind of actuarial game where they're always trying to
• 1:24:01
  to avoid the risk of ruin where they're having to pay out liabilities to people that are hashing to
• 1:24:07
  them continuously against the probability of never hitting a block within a point in time that I
• 1:24:12
  might run down my treasury and not be able to monetize up. So that has actually, while it is
• 1:24:18
  essentially kind of like crack for Bitcoin miners, because they're more than willing to, in many
• 1:24:23
  cases, give up their right and responsibility to build templates for the quick and immediate
• 1:24:28
  that they can get from FPPS,
• 1:24:30
  it creates an additional expense that actually becomes in the long run uh less profitable for miners So it is a centralizing force It is very very difficult to get rid of long run
• 1:24:49
  It, it cannot survive. And the fundamental reason it cannot survive is because as we transition
• 1:24:56
  from a world of block subsidy to more fees being what pays out the miners,
• 1:25:02
  you can almost not keep a treasury that's large enough to account for the variance in fees on the
• 1:25:11
  network. It becomes almost too hard to predict, at least by the way that we have an understanding
• 1:25:16
  of it today, right? So actually, it becomes very, very, very hard to manage. And the fee that you
• 1:25:22
  have to take as a pool to be able to actually offer that service becomes prohibitive against
• 1:25:29
  just taking the variance risk and saying, listen, I'm willing to mine through somebody where
• 1:25:34
  I can maybe have a little more variance, but I'm willing to put up with the variance because I get
• 1:25:40
  more profitability on the back end. So I do think that that day is coming in, in, um, sometime in
• 1:25:46
  the next, you know, four to six years. And just in your own, from like, when you look at the
• 1:25:52
  landscape put numbers on the the the points of centralization like if it made in what's minor
• 1:26:01
  to mining rig manufacturers account for 90 to 95 percent of the network describe
• 1:26:06
  state of centralization and mining yeah so that's that's massive as well and the centralization has
• 1:26:14
  really uh been a function of a couple of things so we have a couple of large predominating pools
• 1:26:19
  that Foundry and Bitmain, their pool, AMP pool,
• 1:26:27
  which account for 40, 50% of,
• 1:26:31
  probably closer to 50, 55% of total network cash rate
• 1:26:36
  at this point in time.
• 1:26:37
  However, there are also what are called proxy pools,
• 1:26:41
  which are smaller pools that want to offer FPPS,
• 1:26:46
  this continuous payout scheme as a service,
• 1:26:48
  but they don't have the treasury reserve to be able to do that. So they actually,
• 1:26:52
  there are a number of, it's a guy named Borst who has a tool called stratum.work,
• 1:26:59
  which is a really beautiful, more technical tool that you can look at to get a sense of
• 1:27:02
  what the centralization looks like. He has been doing some really deep research into trying to
• 1:27:09
  understand how a number of pools have actually been white labeling ant pool because they want
• 1:27:17
  access to the fpps as a product so they have essentially been taking what amp pool is doing
• 1:27:24
  submitting the exact same templates to all the people that are hashing to them and basically
• 1:27:31
  serving as a proxy for this larger pool so it could even be larger in in excess of like 60
• 1:27:36
  plus of the hash rate coming through only two uh two mining pools which is fundamentally the same
• 1:27:43
  thing as two miners on the network because a node in the traditional white paper sense a node is is
• 1:27:50
  one cpu one vote that you're you're constructing templates and doing work so two pools controlling
• 1:27:56
  you know 60 plus of the network is essentially two nodes running 60 of the network and
• 1:28:03
  the assumption is often well if i need to switch i can switch but if there's
• 1:28:11
  too few participants
• 1:28:13
  and they collude
• 1:28:15
  to either collude on pricing
• 1:28:17
  or collude on
• 1:28:18
  the rules that their nodes are running
• 1:28:21
  it seems to me that
• 1:28:27
  there's all this work
• 1:28:30
  in bitcoin miners
• 1:28:31
  the ones that are actually hashing
• 1:28:33
  are naturally distributed
• 1:28:35
  as a function of
• 1:28:38
  the high cost of capital
• 1:28:41
  to actually build a Bitcoin mine.
• 1:28:43
  Then also just the natural distribution
• 1:28:45
  of combination of energy resources
• 1:28:49
  and how those resources are being developed
• 1:28:51
  and where underutilization or waste
• 1:28:56
  or stranded energy might exist.
• 1:28:59
  And so there's this natural distributing function
• 1:29:02
  on the hash rate, but then it's centralized
• 1:29:07
  almost because of software.
• 1:29:11
  not in the interest of like there might be a reason to to reduce variance of payouts
• 1:29:20
  but there's not an interest of individual miners to
• 1:29:25
  participate in this in a centralization function that centralizes it not in their own hands but
• 1:29:34
  somebody else's what is the individual miner's role yeah in furthering decentralization
• 1:29:44
  and in a way that is both consistent with their own self-interest without you know with that idea
• 1:29:51
  of part of their self-interest being ensuring the integrity of bitcoin working in a permissionless
• 1:30:00
  way in a censorship resistant way which only is viable long term because of decentralization
• 1:30:09
  real decentralization we're we're in the first ending of bitcoin mining you know if you
• 1:30:16
  if you look at as i mentioned earlier the way that the early bitcoin talk forums would discuss
• 1:30:26
  mining, there was, there's nothing there that would let you even remotely map the world that
• 1:30:36
  we live in today. Nothing there. You, you can get almost zero sense or almost zero guess of,
• 1:30:47
  of where we are today from having read those very early commentaries.
• 1:30:51
  so much to say
• 1:30:55
  this is happening universally across the bitcoin mining industry which is that we're
• 1:31:04
  figuring out all of this real time at the speed of public markets which is insane
• 1:31:16
  because it's a boom and we're still in the first inning.
• 1:31:20
  So you have to imagine all of the major manufacturers
• 1:31:24
  for the buildings that Megamines are in,
• 1:31:28
  all of the home mining setups,
• 1:31:30
  all of the heat recapture setups,
• 1:31:33
  all of the creative oil immersion setups,
• 1:31:37
  all of the fancy fans and specialized filters.
• 1:31:41
  almost none of that existed 10 years ago
• 1:31:46
  and what's more in the last five five and a half years just in that time we've seen a radical
• 1:31:56
  shift in terms of the machines themselves they've become more efficient but operationally they have
• 1:32:02
  they have different characteristics and require different things so you have an industry that
• 1:32:08
  has had this insane cash injection in the case of mega miners and a massive incentive to go and mine
• 1:32:17
  at this present moment because we know that in the future we have to figure it out now because
• 1:32:21
  we're only going to get squeezed down the road on the the hashrate side of things
• 1:32:25
  the result of that is that we do have pretty large centralization in the hashing side of things in in
• 1:32:36
  the pool side of things, the protocol side of things. Is this existential, right? Or why haven't
• 1:32:44
  people figured this out? My take is kind of generous, which is that miners, you know, we're
• 1:32:51
  kind of the crow magnans of the Bitcoin space. You know, we're, we're kind of just bashing our heads
• 1:32:57
  into things and trying to not let anything blow up and eventually being profitable and just trying
• 1:33:02
  to figure out the tools and the pieces and how to build the rest of the industry around us as we're
• 1:33:05
  going. I think that as you look at us over time, the professionalization in the industry
• 1:33:13
  has been massive, even only in the last 18 to 24 months. The quality of folks that have come in,
• 1:33:19
  the quality of tools, the quality of resources that people have, they're now folks that have
• 1:33:24
  been building things through two, three different versions of machines and kind of understand the
• 1:33:29
  basic principles of how to run mines. It is slowly but surely becoming more and more obvious
• 1:33:34
  to a Bitcoin miner, that their job is not simply to master the energy side, the electricity side,
• 1:33:41
  and the cooling side of their mine, but to actually take on responsibility for the way in which they
• 1:33:48
  are monetizing their hash rate. And the trend thus far has been towards centralization. It's
• 1:33:54
  been towards convenience. It's been towards fast money. But with every passing day, with every
• 1:33:58
  passing block, you are only going to see a fragilization of FPPS because the model really
• 1:34:05
  doesn't work once you get a couple more havings down the road. And as a result, as you start to
• 1:34:10
  master those other two sides of the stool, which is that as I start to really master this energy
• 1:34:13
  side of the equation and I start to really master this thermodynamic, how do I cool it and do it
• 1:34:18
  efficiently and get my providers and get all my, um, the folks who are managing all of my dry coolers
• 1:34:24
  and whatnot, uh, naturally people will understand that they need to take more responsibility for
• 1:34:29
  the hashing side of the equation. And it doesn't have to be super hard. That's the things like we,
• 1:34:34
  we have the tools today, but they're just, uh, they're just now returning us to where we were,
• 1:34:39
  you know, 10 plus years ago. Do you expect that happens marginally? Or do you think that
• 1:34:43
  something will have to, um, not to say break in the Bitcoin sense, but something like a Mount
• 1:34:50
  gox in the in the mining world like some manifestation of the risk either by um a mining
• 1:35:00
  pool supporting a um fork or a mining pool becoming insolvent that large say you know
• 1:35:10
  no one point at foundry but any any you know well there's only two large pools but um where
• 1:35:16
  they have to switch and the alternatives aren't viable and that spawns more people joining smaller
• 1:35:25
  pools or the creation of new pools greater capitalization everything hits the fan scenario
• 1:35:29
  shift of miners figuring out that they have a risk that is either costing them
• 1:35:38
  versus something more system level like a shock yeah i think from the perspective
• 1:35:46
  of miners or hashers, folks that are allocating hash rate from the perspective of the pool
• 1:35:55
  and then from the perspective of the health of the network. So from the perspective of the health
• 1:35:59
  of the network, I don't think that there is a growing risk that leads to an existential threat
• 1:36:08
  to the network. I think the risk factors lie firmly within the hashers or miners themselves
• 1:36:15
  and primarily through the pools.
• 1:36:17
  And it's really the pool's game to lose
• 1:36:20
  as far as I'm concerned,
• 1:36:22
  because they're now offering a product
• 1:36:25
  that folks are becoming well aware
• 1:36:28
  is not sustainable in the long run.
• 1:36:30
  And they're also carrying all of this additional power,
• 1:36:35
  which is the hash rate of folks that are contributing hash rate to them And should they take a strong position on some kind of protocol change Because if I have all this hash rate
• 1:36:48
  and I'm going to be adding blocks to the blockchain,
• 1:36:50
  I'm in a situation where I now kind of wield
• 1:36:52
  this immense power by proxy on the network.
• 1:36:56
  If I can determine proportionally
• 1:36:57
  a large percentage of blocks,
• 1:36:59
  does that create a risk for me?
• 1:37:02
  It does.
• 1:37:02
  And I think pools recognize this.
• 1:37:04
  I think they recognize that it's their game
• 1:37:06
  to lose. That being said, does it come to a head? Yeah, I think it eventually will, because
• 1:37:14
  it's not reasonable to assume that pools and hashers have perfectly aligned incentives.
• 1:37:21
  It's not the case. There are different types of businesses. How it will come to a head,
• 1:37:26
  I don't necessarily know, but how it resolves will likely be some kind of large PR crisis,
• 1:37:34
  but it won't be kind of the knockout, drag out, should we fork or should we not fork situation
• 1:37:40
  when it comes to something like a protocol upgrade. I think it's more likely to take the
• 1:37:46
  form of new technology that resolves the problems that are incumbent in other operators.
• 1:37:53
  And with the emergence of that, you're simply going to see people slowly migrating themselves
• 1:37:58
  away because folks who are hashing also understand that there's a risk to only having a single pool
• 1:38:05
  that I operate with. Folks tend to have backup pools or other sort of ancillary ways to connect
• 1:38:09
  because they don't really want to put all their eggs in one basket in terms of having a provider.
• 1:38:15
  The worst case scenario is that you wind up with a pool that begins to do something like,
• 1:38:21
  you know, overt censorship of transactions, or it takes some really strong,
• 1:38:25
  particular ideological stance that people might or might not agree with.
• 1:38:30
  The second they do that, they run the risk of alienating folks that have been hashing
• 1:38:35
  with them for years.
• 1:38:37
  So it's a very tenuous position to be in as far as a pool is concerned.
• 1:38:42
  I think maybe this is a bit provocative to say, but maybe the golden age of the pool
• 1:38:48
  is kind of behind us because I think enough people understand that they should be taking
• 1:38:53
  on certain responsibilities.
• 1:38:54
  They should be taking on more variants as an operator.
• 1:38:57
  And there are now tools to hedge that risk in other markets as far as a business operator
• 1:39:05
  are concerned.
• 1:39:07
  I think maybe this golden age of the pool may be kind of winding down and a pool, you
• 1:39:14
  know, five, 10 years from now may become this kind of very benign or even as decentralized
• 1:39:19
  as possible thing that is now extremely open source, extremely low cost, and is kind of
• 1:39:26
  just boring by design.
• 1:39:27
  And we put up with the variance because doing anything other than that could incur some kind
• 1:39:33
  of risk on the network.
• 1:39:34
  Yeah, I think that that's a key concept of putting up with the variance of miners, maybe
• 1:39:38
  recognizing their, I mean, I think they all certainly understand mining more than I do,
• 1:39:44
  But they are ultimately in the business of variance, the business of ensuring the viability of the Bitcoin network, and that in the pursuit of their own self-interest, every business tries to evaluate key supplier risks,
• 1:40:04
  functional or key centralization risk where if a provider was to go out of business or to
• 1:40:10
  prove to not be a great service provider that they would they would shift so um it'll be interesting
• 1:40:17
  to see you know still certainly feels like there's a centralizing force today but maybe
• 1:40:22
  maybe that starts to subside um i want to thank you for flying down we got to get you back out
• 1:40:30
  on a on a flight this evening so i think we're running close to yeah almost two hours um
• 1:40:36
  last question which i asked them already goal this podcast is take people down the energy rabbit hole
• 1:40:43
  yeah of bitcoin to increase energy literacy for bitcoiners the mining rabbit hole for
• 1:40:50
  bitcoiners and people who might not yet be involved in bitcoin but then also the mining side of
• 1:40:56
  you know understanding these risks that are give me five names of people that you would like to see
• 1:41:05
  that you think highly of that could go down various different rabbit holes
• 1:41:10
  in the convergence of bitcoin energy and mining oh man well there's a ton and there's so many
• 1:41:19
  little niches because the thing about mining is that
• 1:41:21
  I don't think there are ever going to be industry experts in mining because it so quickly devolves
• 1:41:29
  into somebody who's an expert in chip manufacturing and somebody who's an expert in grids and somebody
• 1:41:34
  who's an expert in thermodynamics. So the best that we can do is just try to get these experts
• 1:41:38
  in a room and figure out what the hell's going on. If you're looking into the heat reuse world,
• 1:41:44
  which is a great world to work in, Tyler Stevens, author of the Heat Punk Manifesto,
• 1:41:49
  So, um, fantastic resource to have there.
• 1:41:52
  Who's a very strong advocate of, uh, mining for utility, mining for heat.
• 1:41:58
  If you want to understand some of the interesting centralization risks from the perspective of
• 1:42:06
  a more technical analyst, uh, Borst who created stratum.work is absolutely brilliant.
• 1:42:14
  and has what I love about him is that some of the finest people in Bitcoin are the quietest people
• 1:42:23
  producing the best research. It's very easy to get sucked up in kind of the drama of the day.
• 1:42:28
  And it's very hard to find people who are really just very, very deep lovers of understanding
• 1:42:35
  technology and Borst has that in spades. If you want a really good take into a lot of the
• 1:42:41
  renewable side of things and the policy side of things. Troy Cross is a fantastic person to talk
• 1:42:48
  to on that front. So obviously a professor at Reed and produced some really interesting
• 1:42:54
  demographic research, which is outside of mining, but has a very deep understanding of
• 1:42:59
  the mining side of things and renewable side of things. If you want to understand
• 1:43:04
  some of the unique ways to monetize i would talk to actually
• 1:43:11
  eric at gridless is a great resource eric who uh eric herstman his um his cto philip
• 1:43:24
  If you could get Phillip in, Phillip is brilliant at modeling the monetization of underutilized assets. And Eric is obviously an incredible entrepreneur. He could give you just a masterclass in startup in a way that would be unbelievable.
• 1:43:48
  the funny side note, I have a good buddy who spent a couple of years, uh, after doing a master's in
• 1:43:55
  London, living in West Africa. And I was actually at the Nashville energy and mining summit a number
• 1:44:01
  of years ago. And I was talking to him, but I'm at this Nashville energy mining summit. I'm going
• 1:44:05
  to meet a bunch of people. Oh, this is cool guy from Africa here, Eric Hurstman. And he goes,
• 1:44:10
  Eric Hurstman, you know, Eric Hurstman. I go like, yeah, the Bitcoin mining guy. He goes,
• 1:44:17
  no do you understand he's the reason why like half of africa has like internet connectivity
• 1:44:22
  his first startup was like this incredibly ruggedized like internet connected box
• 1:44:28
  uh which is this really beautiful story that kind of dovetails into what they do now with
• 1:44:33
  remote minds but yeah eric herstman brilliant dude legacy entrepreneur uh and then fifth one
• 1:44:39
  somebody who i want to see develop more of a thesis who has kind of danced around it a guy
• 1:44:48
  named rev hodl i think he's only a noster he has referred to and it's a concept that i really like
• 1:44:54
  but i don't know if he's fully fleshed it out he has referred to this idea of the permaculture of
• 1:45:01
  bitcoin mining permaculture this idea of like how do you robustly cultivate a space
• 1:45:07
  how do you kind of maximally utilize it there's something there's something in there that i think
• 1:45:13
  is really really powerful when we look at bitcoin mining how do you maximally utilize all of the
• 1:45:19
  various aspects of bitcoin mining that i think he would be an interesting person to talk to
• 1:45:23
  but i i'm not sure he has the thesis fully fleshed out i'd have to double check because
• 1:45:30
  he's only on nostr but really interesting guy all right well i'll give you a count that as six
• 1:45:35
  because you gave me the cto of oh you got eric and philip yeah and rob you also with bitcoin park
• 1:45:42
  have a weekly show that you do is that correct just to share with the audience who uh um not
• 1:45:49
  just where you can find you but but the work that you're doing at bitcoin park yeah so the the big
• 1:45:54
  thing for us with bitcoin park is that and you know this as well is that if if you're going to
• 1:45:58
  have an exponentially growing network just of people that are beginning to get interested in
• 1:46:04
  Bitcoin, over a certain amount of time, the new entrants are always going to outnumber
• 1:46:09
  the people who have been in the space for a little while.
• 1:46:13
  And what becomes really important by our estimation is, is how do you curate, create and distribute
• 1:46:19
  the highest quality information so that people coming in can immediately know that that's
• 1:46:24
  where you go as your source of your source of truth, your source of knowledge?
• 1:46:28
  Where do I connect to talk to people that are actually doing like the building and the
• 1:46:33
  the innovating in the space. So that's really our whole design thesis about what we're building.
• 1:46:39
  So thus far, there's two really easy ways for folks to kind of keep eyes on what is happening
• 1:46:44
  at the park. One is our substack. So bitcoinpark.substack.com. We do a daily newsletter
• 1:46:51
  called Op Daily. So the idea is highly digestible, four bytes in two minutes. You can get it every
• 1:46:58
  single day and you can know that this is going to be the highest quality Bitcoin only freedom
• 1:47:02
  technology focused summary of what's happening in the space. And then more long form, we have
• 1:47:09
  streamed up at Bitcoin Park dot com, which is essentially an aggregator of what's happening.
• 1:47:15
  And it's everything from here videos from recent summits to hear all the recent news stories to
• 1:47:20
  hear what's happening with the friends of the park to here's the next meetup that's coming in
• 1:47:24
  Austin or Nashville. So two resources. But the idea is where do you go if you want to find the
• 1:47:30
  highest quality, very, very focused information and try to avoid all the fluff and all the
• 1:47:36
  all the nonsense that inevitably grows with the growing space.
• 1:47:40
  Well, appreciate you coming down to the comms, Bitcoin Park Austin.
• 1:47:43
  Appreciate everything you guys do, but I really do appreciate you coming and joining me today,
• 1:47:49
  lying down for it and doing that with a young one just a few weeks away.
• 1:47:54
  It's OK.
• 1:47:54
  I told her last night that I'd be back.
• 1:47:56
  It's actually two months.
• 1:47:57
  So it's not, you know, it's a little bit more than a few weeks, but it does mean, it means a lot to me that you came.
• 1:48:01
  So it's a pleasure to be here.
• 1:48:02
  And I love that you're doing this and I love that you're, you're growing it because fundamentally, and I said this to Pierre when, to me, Pierre is an interesting, Pierre Richard, what's so fantastic about him.
• 1:48:14
  And I think it's fantastic about you is that you just incessantly double down on what you're doing and there's no frills.
• 1:48:22
  There's no side quests.
• 1:48:23
  You're so mission focused.
• 1:48:25
  And I think that there's nothing that the space needs more than folks who have put in the work and are so well versed in in what's happening in the space, continuing to dedicate their efforts to education in the space.
• 1:48:37
  Like, I can't tell you how excited I am for for the growth of this.
• 1:48:42
  Well, well, appreciate that and appreciate you.