Alex de Vries – Podcast Notes September 2019


Sam: Alex is the founder of the blog Digi-conomist, and the blog is a platform for in-depth analysis opinions and discussions with regard to bitcoin and other cryptocurrencies. He focuses on sustainability and bitcoin. He created the first bitcoin energy consumption index in 2016 which has played a major role in the global discussion around proof of work blockchains.  Alex welcome to the podcast.

Alex: Thank you.

Sam: So, Alex, can you give a bit of background about yourself? And how did you get into the point of identifying these sustainability issues with bitcoin and other proof of work cryptocurrencies?

Alex: I started writing about bitcoin and blockchain in 2014 where I came across an interesting news headline which stated that a single bitcoin transaction would consume as much electricity as a US household consumes over 1.5 days that was around 2015 so already one and half years after I started writing. For me, that was the first time I personally got in touch with the sustainability side of bitcoin. To be honest when I first read that I was honestly shocked because that is pretty significant and secondly, I was extremely surprised. I wanted to find out more about it, and honestly, I couldn’t find anything up to date other than the article I just read. There were a few studies about it, one form 2014 which I think was the most recent at the time and I really wondered why no one was talking about something which appears to be so significant.

Obviously, as a Bitcoin user, you are never going to be confronted with those costs, for me the reason I did not know about it before was that the energy costs would be paid for by miners. They still are today; so as a user, you don’t see the environmental impact of that transaction, but it is there, and it is real. I wanted to do two things: (a) provide a source of information that is actually providing live information and (b) contribute to the feasibility of these numbers with the help of this information source. That’s how these Bitcoin energy consumption indexes came into being. And I think I have been hosting that one now since late 2016.

Sam: Right, so in addition to the energy consumption index, there is also a E-Waste index as well too, which is the first index I have seen for that. Other studies try to take a stab at how much electricity is being used locally for bitcoin mining, but I had not seen the E-Waste numbers put into any sort of overall chart before. That was one of the more interesting things I saw.

Alex: I launched the first energy index as you said, and nowadays there are a few more of those. This year I added the E-Waste monitor as I call it which displays a real-time estimate of the electronic waste as a part of Bitcoin mining. Because what people forget it that Bitcoin’s environmental impact doesn’t stop at machines consuming electricity. When those machines are done being used for mining they have to go somewhere. The problem with these machines is that they can’t really be repurposed. The machines are specialist hardware that can only do one thing which is mining Bitcoin and if they are no longer profitable through that end, then the only thing you can do with them is keep them on the shelf for some time to see if they become profitable again but with each generation coming that does not really work at all. And the other thing is that you can throw them away and burn them. So that is also another part of the problem.

Sam: How much of the miners, say an S9 when it comes to the end of its life, how much of it is recyclable?

Alex: Well, I think the problem with electronic waste is that it is not recyclable.  If you look globally, (this will probably differ at local levels), but only 20% tops of all electronic waste is getting recycled the rest is ending up in environmentally damaging landfills and/or incinerators. So from a global average point of view, this is not a lot.  Of course it may differ on the location where these mines are.

Sam: So according to the Electronic waste monitor you have on your website, there is 9,411 kilotons of E-Waste being generated every year. That’s down from the peak of September 2018 where there was 12,977 kilotons of E-Waste.  The hash rate has gone up quite a bit, so why hasn’t the E-waste for the year returned to those levels from before? Why has it remained pretty flat or in decline since the October 2018 peak?

Alex: Well one of important things to keep in mind is that hash rate is not a perfect indicator for how many machines there are in the network and what’s been going on the past few months is that we have seen a new generation of bitcoin mining machines entering the market. So what happens is that because of these more efficient machines the hash rate will go up but basically if you have one S9 miner, and instead of that you have an N-Miner17 it might have similar weight, but the hash rate is three times as high. The hash rate can keep on growing while the total equipment stays relatively flat or declines.

Sam: Even though the network hash rate has recovered from where it was even a year ago, at the peak in 2018, the amount of machines has not exceeded that point as it was a year ago correct?

Alex: That is what you can conclude, yes.

Sam: Do you think this trend with continue? That we have reached some peak and we going to stay around this level of 10,000 kilotons a year? Or as the network or price grows again, the seigniorage costs given to the miners increases their profitability, do you think that will start to rise again?

Alex: Yes, so actually, my first paper called Bitcoins drawing energy problem was very much about the relationship between Bitcoin energy consumption and price.  The relationship would be the same for electronic waste too.  What’s going to happen is that if these miners are generating more revenues, they will be able to spend more money on resources. It’s that simple.

That means they will both consume more electricity and on the other hand, employ more machines. So, both energy consumption and electronic waste will go up if the Bitcoin price goes up.  Obviously, we can’t predict price, so as long as the price stays where it is today, we would not expect much of increase. We wouldn’t expect a decrease either, but that can change in a moment. But if the price doubles tomorrow, we won’t have twice as many machines tomorrow. There is a lot of resource limits there but there would definitely going to be a lot of new machines coming into the network.

Sam: Right, the miners are low down on the totem poles for production. I think one of the biggest suppliers Bitmain goes through TSMC. The amount of machines they are building is very small in comparison to the total semi-conductor market. I know for Bitmain is hard for them to get their chip production done as they need to slot in versus the behemoth’s like Samsung or Apple. The major chip consumers that are using these major semiconductor providers to produce their chips and so the price can get pretty far ahead of where the operational costs are to produce one bitcoin.

Alex: Yes, exactly; it’s a resource limit. In order to have more machines in the network you need to have a higher price to actually support those machines. But if you have a higher price you don’t immediately actually have more machines, because those machines need to be produced. I think that is a very natural limit. It’s just a matter of time in the end before it takes time, but as long as it is profitable to make machines, people will keep on doing that. Eventually, we will use up all the available space…

Sam: It’s essentially a money-printing machine… let’s say the total price of an S15 miner is $4,000 to print one Bitcoin, then that machine is worth $6,000 at this point. You should buy and run as many machines as possible to collect the difference between the operational costs and the actual price of what you can sell one Bitcoin for.

Alex: Yes

Sam: On a lag timescale, you saw this back in 2019, the price actually came down probably within 10 to 20% of the production cost, the marginal cost of producing one bitcoin and from there we’ve gone quite a bit more up to the price level we are at now.  I think that would be something to watch as time goes forward. Ten years is not a big sample size, and the price relationship between hash power and the number of machines online correlates with the price of Bitcoin at any one time and that done on a long term timescale and more importantly this is where your…

Alex: This is also last year, where we saw a pretty extreme situation… where we saw first of all the rally at the end of 2017 where the Bitcoin price reached £20k and then the price got slashed in half then in January 2018 and then the hash rate kept climbing for ten months after that. Exactly what I predicted in last year’s paper as it would simply be because even though the price fell a lot and really used up all the available space in the revenues to spend on resources. Because of production limits, in essence they got a lot of free money in December. The ideal situation if you buy a bitcoin mine is that the Bitcoin price explodes because you can be certain that your competition will not have a very big amount of new machines because they are simply not there. So you are essentially getting a lot of free money.

For a large scale mining facility, it takes 3 to 6 months to come online. That is a lot of time and so let’s say you, and I wanted to build a mining facility it would take a few months to get all the miners, prepare the facility itself and it maybe 3 months down the road from there that we can mine something. And it may be that from there depending on the type of miner it is, let’s say is the new S15 miner, maybe it is an 8-month return on equity to get back to being profitable. Only at that point, a year down the road from where we are now is there any profitability in mining.

Essentially there is a lot of speculation in here I suppose too. I think it makes it hard to estimate what future consumption will be because just like I said in the ideal scenario your miners go online when the price doubles but that’s something you can’t predict, you might expect that upfront but if the entire market expects that upfront then the price is going to go up anyway. The speculation distorts the relationship between revenues and mining, but it essentially means that there will be more mining than revenues would allow due to speculating, and there will be even more revenues in the future.

Sam: Which is the stronger cap on how much more mining can be done? Is it the production side or the power consumption side?

Alex: I think so far the limit has been on the chip production side because I don’t see the miners having trouble getting electricity last year, but chip production took a very long time.

Sam: On your website, it says the Bitcoin’s current estimated annual energy consumption is 73.12 Terawatts per year. So the closest country in terms of energy consumption is Austria which is a big country. That is a big European country. Do you find this energy consumption concerning? What is your long term views on how policy should be created around Bitcoin mining?

Alex: Well I mean the total number itself represents 1/3 of a per cent of global energy consumption which is not insignificant, but at the same time it is not going to be the end of the world. But what to me is the most concerning part is how disproportionate the energy consumption is but as to the actual use of the bitcoin network. If you look at the Bitcoin network, it processes around 80 million transactions per year. 3 to 4 transactions per second. The whole network is running on about 80 quintillion calculations per second relative to those 4 transactions that are being processed in that second. So even if you don’t know the energy cost associated with it, you can feel without having any information about this topic, that it is somewhat disproportionate. And if you start to attach the energy consumption footprint to each transaction, it comes really apparent how disproportionate the energy consumption is.  73.12 Terawatt hours for the whole system for the whole year, on a per-transaction limit, you are talking about 600kwh of electricity which is an insane amount, it’s like 5 Tesla batteries worth.  It is extreme, and you can compare it to many things, and people can say comparisons are far from perfect. Nevertheless, suppose you compare it to a visa transaction, Visa is doing about 1 to 2 kWh per transaction. That is a very big factor difference.

Sam: My counter-argument to that would be the size of the transaction, the monetary value of the bitcoin transaction is vastly bigger than a visa or PayPal transaction. That your average visa transaction is going to be $10-$20, but your average Bitcoin transaction is going to be in the tens of thousands of dollars range.

Alex: Yes,  but my counter-counter argument would be that Visa Transactions are typically economically meaningful transactions. Where Bitcoin transactions are mostly people moving money around to themselves. In that sense, you can approach this from multiple perspectives I suppose.  And for sure no comparison is really perfect. What you can do is in the end, whether or not these costs are extreme depends on the way you intend to use the system. If you are just going to use it to send money to another country then use a centralised system for this which runs much more efficiently for that, then you can argue that you don’t want to use a centralised system for that because I don’t trust banks then Bitcoin isn’t your only option, you have other cryptocurrencies that run on proof of stake that are a lot greener than Bitcoin.

Sam: Let’s say I wanted to an international wire from the US to here in Russia for over a $100,000 dollars, the point becomes a significant amount for the bank. I may need that cash in a couple of days, for the bank that higher amount transfer is going to take three to four days to finalise all the paperwork for Swift and after that even though the payment may clear the bank has to do compliance measures to make sure releasing the funds to the correct person and that it is not being used for any illegal activity. That in itself could take another two to three weeks depending on the amounts involved.  Once you get into these large transfers, there are many factors that can slow down or stop these transactions from happening and from having access to the funds. Bitcoin may be slower, but I can send them $5 million dollars with a click of a button, and it will arrive in 5 minutes.

Alex: there are several things we can touch on here. First, there is speed. I don’t really think you need bitcoin or blockchain for speed in general. Anything you can do really fast on a blockchain you can do faster on a centralised system except for adhering to, for example, sending money to Russia or Iran, you might have to take sanctions into account. That’s where things start to slow down because the bank will not allow it or take a long time to approve the transactions. But if you are using Bitcoin for those, you are basically protecting, or avoiding sanctions with the help of Bitcoin which is possible I guess.

Sam: This is why countries like Russia and Iran, like Bitcoin because it is outside the US payment system.  You see this in Europe too. Europe is now creating an alternative to SWIFT so it can go around the US-controlled SWIFT system and conduct payment for Oil to Iran and basically evade US Sanctions because they don’t agree with the political decisions that have been taken against the Iranian government.

Alex: This is what I see as the real unique value of Bitcoin, the lag of control by a governing authority like the US or Europe or whatever. Being able to do peer-to-peer money without an intermediary, that for me is the essence of Bitcoin, not speed. The trustless part is very powerful, and Bitcoin is actually used in that way to send money to places where it might not go.  That is a use case. I am not sure we should be advocating for that use case.

Sam: Sure, if it is something that needs to be done. It is a valuable point… for the last 50 years, the US has weaponised the dollar, and I know that Europe has taken certain things with the Euro to punish South America and Asia and Russia and Iran to cut them out of the payment system when they feel they are not acting in the political way they want them to. If they are not bowing to American or European power, they can be shut out of the payments system, which if you control the payments system and it's your currency then great but from a greater global standpoint I don’t think it’s a good system to have. It’s not going to work in the long term.  You have china coming up, wanting to buy Iranian Oil, wanting to build ties with Russia and other countries which may be looked down upon by the US. It’s going to create rifts when the two largest superpowers are not aligned in how they deal with smaller countries or dealing with sanctions or at a geopolitical level.

Alex: I think that what we can say is that if you want to use Bitcoin to this end then there is no alternative in the regular financial system to use in a decentralised way, so what you got to do then is look within the cryptocurrency space there are definitely greener alternatives to bitcoin from a sustainability point of view and you could use them in the same way. Even if you are using Bitcoin in that way, you could do it in a much greener way by not using Bitcoin but a proof of stake cryptocurrency instead.

Sam: If that is the case, they would have switched already. There has to be something in the proof of work energy burning that lends value to the network as a whole.

Alex: I think it is two-fold. I think it is awareness. As soon as I start showing people my numbers and then ask them if I should demonstrate a Bitcoin transaction, they say no thanks. Whereas if it is the other way around, they are like sure show me the demo.  Oh wait was it that bad? I would not have done it. That is one side of the hand. Raising awareness is part of my mission with my blog and my energy index simply because as a user you don’t get confronted with those environmental costs.

I am sending a bitcoin transaction anywhere I am not paying for electricity that is expended minus the process of my transaction are subsidised by the protocol and they pay for the energy that is expended. So that is one part of it and then the discussion around how to do a cryptocurrency consensus algorithm properly, and there are people who believe that proof of work is the only way to actually do it, but I think we can be certain that this is not the case because there are other cryptocurrencies running on proof of stake and they have been doing so for several years. But then it becomes a theoretical debate about does it offer the same long term security that is offered by proof of work? Because in the end, proof of work is just one part of Bitcoin, it is an important part of Bitcoin. It is not a part, that if you take out proof of work and put proof of Stake in its place, you don’t change Bitcoin. What you might change is the security provided by the algorithm. Obviously, part of the security is people are spending a whole lot of electricity and machines to do this work. So if you want to attack the network, you need a whole lot of electricity and a whole lot of machines as well to do an attack. If I wanted to do that I would need a majority and you would need a few billion dollars of resources to do this.

Sam: The sunk costs  the miners already have in producing the Bitcoin which they own at this point is enough to prevent them from switching over to this model.

Well if you are a miner, you are not going to advocate for an algorithm chase. The miners will be the last person to do so as it will cost them their revenues as they have invested in expensive machines. The longer they are online, the more they can earn back.  In Ethereum they manage this a bit differently. When this was launched proof of work, they were planning to move to Proof of Stake, and they put in a difficulty bomb which would make miners that stayed on proof of work unviable because the difficulty would be exploding at a set point in time. By now they moved back the bomb explosion date a couple times, so it is not a credible threat anymore. But at least that is what they were trying to do to incentivize miners to go along with the changes.

And an additional thing in Bitcoin is that even the developers to some extent are invested in Bitcoin mining. There was just a news article that came out that states that Blockstream now owns a very large mining facility that consumes 300 MWh of electricity meaning that the developers who are invested in this way are very unlikely to start working on a proof of stake alternative.  Those are definitely things that play a role in how likely Bitcoin is to switch. It is definitely going to be very hard. The miners will not go along, and the developers are incentivised in part not to go along with it. On top of that, this has nothing to do with security properties; it is really about personal interests, not about what offers the better security because even proof of stake offers some level of security depending on how you set up the algorithm. The question is how much is the difference? You can even wonder is proof of work really secure in the long term because the block rewards go to zero which is great news for the environment, but it makes it wonder about all the resources we are now blowing on the system. Will it be in vain in the long run because it cannot sustain itself in the long run?

Sam: So the argument that I have heard many Bitcoiners make is that Bitcoin is efficient because it will force energy producers to meet their needs. They will, therefore, build more sustainable energy sources such as hydropower, nuclear, etc. other means that are not as damaging to the environment. What would you say to that argument?

Alex: Well I suppose if we get something that incentivises us to use up all the energy on this planet, then we will have a big reason to build a Dyson sphere. I can’t really follow the line of argumentation very well. More renewable energy is not going to make up for Bitcoin mining that happens with machines that run 24/7. You don’t shut down your miner because more and more machines are being added to the network, meaning your slice of the pie is always decreasing so any time you take out is going to cost you money that you will never get back. The machines are running 24/7 and basically increase baseload on the grid. Now that is a very important concept to keep in mind. Especially if you are looking at renewable production. Renewables are an intermittent energy source. It is not always raining; the sun is not always shining etc.

Alex: What about nuclear? Nuclear is great, it is very consistent, but nuclear waste isn’t really great for the environment. Nuclear doesn’t really have any fans these days, but at least it is clean and consistent.

Sam: I don’t know. I would say that there has been great advancement in nuclear technology and over the past fifty years, we have learned a lot from our mistakes. There are less nuclear power accidents than other forms of energy production.

Alex: Miners need two things. They need consistent energy. It does not come from renewables; it comes from fossil and from Nuclear energy. So having nuclear power is a good source of power for Bitcoin miners. Nuclear power is not necessarily in remote locations as is the case with renewables. Let’s talk about that in a little bit. If it is not in a remote location, you probably going to have some level of displacement going on because you get relatively clean energy from a nuclear power plant that you are going to use for bitcoin mining rather than cleaning up your grid nearby.  The reason renewables have been used for bitcoin mining simply because during certain times of the year there exist excess in certain areas. One of these is in Sichuan, China which became a very big mining centre because during the summer wet season they have an abundance of hydropower that they can’t get rid of. The Chinese Grid is extremely poor, so they are unable to properly export the excess capacity to other provinces in China. During this time, Bitcoin miners can take part of this excess. These companies are using the renewable energy that would otherwise go to waste, but then the problem comes when it is only part of the year. It was very late this year in China due to climate change, and that is just getting worse. When it is the dry season in the winter months, production drops significantly. I think the hydropower is three times as much as in the summer as in the wintertime. Now we come back to the baseload demand these bitcoin miners have which is the same in summer. During the wintertime, they have the opposite problem they have in the summer in . Being that during the summer they can’t export their extra capacity, in the winter they can’t import capacity so what they need to do is actually build additional coal power plants for providing power during the winter months. If you have something that increases baseload demand on the grid, you actually have incentive for building more of these coal power plants. Unless of course you can get special arrangements with them. I am not aware of this existing in Sichuan.

I do know in Quebec in Canada they have a similar situation, during the winter they have a demand peak. They are very dependent on electrical heating so the demand peaks at 65% during the winter months. Although their hydropower production is very stable year-round, they have to supplement capacity during the cold winter months. What they do is, what I saw, Hydro Quebec made an agreement with several miners that they reserve the right to cut power to them for a period of 300 hours during peak demand which sounds okay.

It is definitely a solution to the volatility problem but now think back on your return on investment period on these bitcoin mining machines; we were talking about 8 months earlier on any mining facility. Most of the machines earn most of their money during the first half-year of their existence. 300 hours is quite a lot in that relatively short period. Meaning that if you decide to go along with such a scheme like in Quebec, you need be able to afford it. There needs to be some access to get lucky, and the Bitcoin price is very high, so you have longer to earn back your investment. But essentially you are putting yourself in a very disadvantageous position compared to someone who has access to consistent energy.

There are people now who are moving to Iran because they can get access to cheap energy but also consistent energy because they are using natural gas rather than using hydropower. If you are in Quebec and the price gets a bit tight which actually happened last year, the miners that are situated in Iran getting both cheap and consistent power are going to be better off in terms of being able to survive than the ones that are located in Quebec. So I don’t think even though such a scheme in Quebec sounds nice, it is not where the long term winners are going to be. And I think they are more likely to be in Iran where they can get both cheap and consistent power which is a lot dirtier than obviously hydropower in Quebec.

Sam: Right and there is less regulatory oversight and possibly bribes being paid out to be able to access the energy in less than straightforward ways. There is probably a huge process you go through if you are going to set up as a bitcoin miner in Quebec you have to get a lot of permits you have to get approved by the government. But if you are doing something in China or Iran, maybe it’s only knowing the local government official who can slot you into a local warehouse next to the local powerplant.

Alex: We saw some raids on Bitcoin mining facilities in Iran in June/July this year.  So very recently still.  If you have a big facility, it is very hard to stay under the radar regardless of where you operate. They will observe peaks in the electricity demand and will ask what is this? What is going on?  Electricity consumption in Iran actually increased year on year by 7%.

Sam: Wow.

Alex: And primarily, they expected Bitcoin mining to be responsible for that. I did the math on that, and if that 7% is fully attributable to Bitcoin miners, it would mean that a quarter of the total bitcoin mining is located in Iran right now.  I am a bit sceptical it is the case. The numbers would imply that if that 7% is fully attributable to bitcoin miners, nevertheless Iran is a very popular destination for Bitcoin Miners again because they have very cheap electricity. We have even saw mosques being filled with bitcoin machines because they get free electricity. Here is no limit on that you can mine until the end of days. Someone might observe this eventually like why is this mosque using so much electricity? But they are now working on legalising bitcoin mining in Iran at least they made it into an official industry. They have not decided what the tariff will be on mining, but they will not get the local rates because they were planning on charging the export rates to bitcoin miners. We will have to see how that goes. I am not really sure how that will go down.

It is for sure that Iran is getting a little bit more friendly towards Bitcoin mining which is very significant because at the same time. China is getting more unfriendly towards Bitcoin miners. And this year they said they are considering banning all bitcoin mining because they consider it a waste of energy.  And Iran is not that far from China. We are already have stories of people who made the move and talk about it anonymously. It is a sensitive topic mining in Iran, and you do not want to get caught on that. At least there are stories about people moving from China to Iran. There is no way to know how many people have already made the move. I think it will be quite a few and there will be more in the future as China becomes more unfriendly towards mining.

Sam: So if you were in charge of policy, what would you do? What type of policy directives would you put forward to regulate bitcoin mining? What direction would you encourage European policymakers to take?

Alex: Well, one thing you can do to cut the power consumption of this bitcoin mining network is you can prevent Bitcoin miners from getting access to the cheapest energy. And because it is very hard to stay under the rader if you are running a big facility. If you want to take advantage of industrial rates you are going to have to apply somewhere. Governments can say no you are not getting the cheap rates which doesn’t really work well locally so they would need to coordinate this across the EU.  This something you need to coordinate on a global scale as bitcoin is a global money and bitcoin mining is a global activity. So the only way policy is going to work is if you have a global taskforce and that is going to be very hard. We have already seen in the United Nations that collaboration is extremely difficult.   But otherwise you are going to have this waterbed effect where in Europe they say we don’t want you, in China they say we don’t want this here, but Iran says come over here, and Russian says that as well. Then everyone is going to move to those areas, and then it’s not necessarily going to be any prettier.

Sam: Right then you lose oversight over it internationally because then it’s just actors who do not want to be part of the international system, taking in the miners, which puts them out of reach of international regulatory bodies.

Alex: I am curious what will happen if Bitcoin mining becomes very big in places like Iran, and there are lots of coins are you eventually going to come from Iran. The US is very strict with Iran in the regular financial system, Iran is on the blacklist. Eventually people convert their bitcoins into goods or into other money. Those places might be less inclined to do business in Bitcoin if that money originates in Iran and because it might get them in trouble more.

Sam: Coming back to the power question, do you think the competitive demand for Bitcoin mining could be used to increase the construction of nuclear power plants to take some of the baseload used at peak times to put it into the miners.

Alex: To summarise my very long answer, I think we have the opposite incentive going on because miners need cheap and consistent energy and I think it is more of an incentive to build or keep open coal-based power plants.

Sam: Interesting. Pretty much everybody in the Bitcoin industry says we are moving toward renewables, environmental impact in the long term everything is going to be hydropower or some other renewables, but you don’t think that is the case?

Alex: No, not at all. I think you need really need to study grids a bit more; you are trying to argue that line. It is not really about renewables, but it is about grid management, and it is a lot more complicated than a lot of people in the Bitcoin community pretend it to be.

Sam: Can you talk about that for a bit? I don’t know much about grid management myself.

Alex: Well, you need to manage both the demand and the supply side. With renewables you have problems on both ends. You have problems on the production side, the actual production is unstable, sometimes to the extreme. And at the same time you are dealing with volatile demand while you have uncertainty. Bitcoin miners are just consuming energy at the same level throughout. So have you ever heard of the Duck curve?

Sam: No, I haven’t.

Alex: It is something you see in renewable energy generation, especially in solar during the day, there is an obvious decline in the night.  And again in the morning, the whole curve becomes like a duck. If you have something that increases your baseload demand you are going to make your duck curve a lot steeper, that is specifically for solar but you have the same problem with wind power or hydropower (less over longer timescales) but you are just going to worsen your demand at peak times which is where the real stress comes from because you need to be able to supply power during those peak times or you can cut out Bitcoin miners during that time, or you are going to need more backups. It is one or the other, or cut off residents. No one is going to like that.

If you are going to increase the baseload at peak times, you are going to need more coal-based powerplants or something like that to have that supply during the peak times. I think this is much more challenging than people present this as facilities have a constant surplus of renewables that these bitcoin miners are now taking in – that is obviously not the case.  The surpluses are temporary at best, especially in renewables. Then it becomes a whole different story.

Sam: Interesting. So what we are really doing with the Bitcoin mining network is spurring on greater fossil fuel (energy) production?

Alex: That is the most likely the outcome, and it is more likes than incentivising renewable energy production.  Someone on Twitter said it very nicely; I am going to leave my diesel car engine running to incentivise more green energy sources. It doesn’t really make a lot of sense from that perspective. No matter how much green energy because we were talking about electronic waste early, no matter how much green energy you put into the network the big pile of electronic waste you have at the end still has a large carbon footprint associated with it and 50% of that is going to end up in Incinerators. Altogether it is not a perfect solution. As I stated on Twitter, it is like choosing between vaping and smoking. Both are pretty bad for your health. One is a bit less damaging than the other, but both are bad for you.

Sam: Is vaping bad for you? I thought it was just water vapour and nicotine.

Alex: The jury is still out on the long term effects of it, but it is not looking good. But I think that it is better to not do either of those.

Sam: With Digi-conomist, is this helping you spur on any policy objectives? How do you use your website? You have put together all this information. Who is coming to ask you for your opinion about energy usage and electronic waste?

Alex: The information is used by all kinds of people, it is used by miners themselves which is quite interesting I found. Because they were using it in Quebec to argue in front of a judge that the amount of…what happened in Quebec is at some point they decided to not allow any new miners on the grid because they had an extreme amount of requests and they used my information to show that these requests don’t make any sense because the money to support that huge amount of consumption is just not there. Cutting us off based on this nonsense is just not fair. Even miners are using this information.

From a policymaking perspective, I saw someone from the European Commission working on the bigger topic of digital transformation. One thing they found that was lacking from the Paris agreement was the result of having all new technologies entering the market at this point, not just Bitcoin, but they were looking at the impact of AI, IoT devices, Machining learning entering the market… there is very little of this impact being discussed in the climate agreement at all.  I think they might be using this to at least consider some guidelines on Bitcoin because it is a global system, but with AI how to deal with this stuff.

Sam: Where are you taking it? What is your next steps? How will you grow your models?

For me, it is seeing if the principles I have applied to Bitcoin can be applied to other areas in a broader sense. Bitcoin is not an insignificant issue, but it is not the only one. It is the same thing, if you look at electric cars, they sound like a good idea but when you look it, they are only as green as the energy you put into them, it’s fossil-based and it’s not that great. You can actually look at the sustainability issue in many areas. I definitely want to apply these principles from the Bitcoin market to another market. For example you create a new medicine for a new market; what does that mean? You will have a certain profit potential. There is a benefit to that, but it is harder to measure. You could measure what is going to be the environmental impact of this. I think it is going to be a lot more complicated if you look at things this way. But I think you can use the same principles. If I have a new technology for machine learning, what is going to be the environmental impact of that new technology? I think it is very important that people are aware… it is really cool we have all these new technologies these days, but then again we should be very careful in doing something because we can without considering our planet. Because we only have one planet, we really need to start acting now to make sure we can still live on this planet two decades from now.

And because of that, we need to wonder I might be able to do something but if it has this impact compared to whatever benefit I expect is it worth it? It is the same consideration with Bitcoin. If you can’t do Bitcoin in another way, I am just going to assume we can’t do it in a proof of stake way, is the cost worth the result? I think that is a very subjective decision in the end. I think it is about getting the proportion right.

Sam: It is a very difficult question. Balancing the cost versus the potential environmental impact, it will have especially if it grows to a significant size like the gold market.

Alex: Yes, this is my real point. Even if Bitcoin is not at the end of road right now, it is really about happens when we start doing this on a global scale. I was recently at a bitcoin conference. Supposedly we solved scaling, right now the limits is about 3 to 5 transactions per second, lightning network layer 2 solutions are development which potentially will allow bitcoin to handle many more transactions, so I asked the audience… let’s assume transaction limits are no longer there, let’s assume bitcoin replaces the current financial system, what do you think would be the value of bitcoin at this point? And I gave a few options do you expect the price to say the same, nobody thought that would be the case. Do you expect the price to go up 10xs, I still saw no hands until I got to 100x. And then the people said 1000x, or more …

Sam: I think that is what a lot of Bitcoiners expectations are.

Sam: But then it is really hard to predict the networks’ environmental impact at this point.  If the bitcoin price was more than $10 million per coin, then bitcoin miners would have more money than the entire US federal budget to spend on resources. Now I don’t know what they is going to be for the environmental impact, but I do know it is not going to be great. And because if that is the case it would be more than 10% of global energy consumption would be used by mining.. and I think we are going to need a Dyson sphere. For those who don’t know what that is, it is a way to harvest the energy of ourselves.

Sam: That is a good point on which to wrap up.  Alex is there anything else you want to cover or talk about?

Alex: I was going to share some fun statistics at the end of the podcast.

Sam: Sure, that would be great.

Alex: Apart from the energy footprint of a bitcoin transaction; we also have a carbon footprint of a bitcoin transaction of 300 kilograms of carbon, that is per transaction and the fun stuff, I don’t know if it is really fun, people often ask me, okay so 300 kilograms of carbon seems like a lot what does that really mean?  I have a few examples of what 300 kilograms of carbon emitted by each bitcoin transaction is equivalent to, for example:

-        The carbon footprint of 72,000 emails, or

-        1.44 million google searches, or

-        Also if you watch videos on youtube a lot, you could watch 120,000 hours (I think that covers a lifetime of watching cat videos).

-        8.5 km of flying a Boeing 747

I think this illustrates how big the carbon footprint of a bitcoin transaction really is. It is a performance indicator. But the trends over the last few years is that it is going up. You could even say if you managed to run the bitcoin network on pure renewable energy, hydropower has a lifetime emission factor of carbon of 20 kilograms – even it is not zero.

Sam: You are never going to get to zero.

Alex: No matter what you do, you always have construction costs and carbon emissions even on the most opportunistic numbers, based on natural gas. There is still going to be several kilograms of carbon created by the bitcoin network even if it was run only on hydropower.