 Thank you very much. Okay, let's dive in the sustainability of Bitcoin and let's first just take a little step back especially for the central bankers in the room Blockchain technology in general is inefficient by design That's not a bug. It's a feature So I've been seeing some news messages popping up last few days where some parties were testing blockchain technologies And they figured out this hey this technology is slow and expensive. Well, that's not so hard Because if you look at the steps to lock ahead to run a blockchain network from the original bitcoin white paper And you take out what is proof of work then you find that there is still a lot of repeated effort in Running a blockchain network regardless of what kind of consensus algorithm you put into it And Independent validation again Not a bug. It's a feature, but in general it will just slow things down and make it very expensive Now, obviously we're talking about bitcoin here and when we're talking about bitcoin We're diving into the mining the proof of work and that really amplifies bitcoin's energy consumption massively So let's look on let's look at some network statistics first The bitcoin network runs on about 10 000 nodes Which is if they were all single computers would be perfectly fine Only there are millions of mining devices hiding behind these nodes And altogether they are producing 53 quintillion hashes per second Um, we actually saw the 53 ones one time before today and we also saw a number That's at 43 quintillion So what's the correct number? Well, the 43 was based on a single day observation and 53 is based on a full week Of observations It's a statistical error. If you have more data, you get a more reliable number. So this is probably more accurate Um, and where it gets really interesting is if we look at the ratio of Hashing operations to actual use transactions So in 2018 the whole bitcoin network didn't process much more transactions than 81 million That's two and a half per second So overall we have 20 quintillion hash operations every second of the day and we have two and a half transactions No wonder it is very resource intensive Um By the way, the whole banking industry was handling About 500 billion non-cash transactions in total. So that's roughly 16 000 transactions per second Uh, if if bitcoin wants to scale up it has a long way to go in that sense So Let's talk numbers How much energy for the whole network per year? 46 theta watt hours Now I like to do comparisons because these numbers are a little bit meaningless otherwise It is pretty much the same electrical energy consumption as a country like switzerland The footprint per single transaction comes down to 463 kilowatt hours That's the same footprint as your well actually for a german household It's probably a month of electricity the americans are known to be Using a little bit more on average Um, that's a pretty big number And by the way, uh, I just here heard I can't average per transaction. Well, um, I uh I can explain metrics like carbon footprints But the whole idea is that you include direct and indirect costs And without energy consumption there wouldn't be bitcoin transactions So obviously you can attribute these costs um Oh, and this is also an interesting number 82 megawatt hours per bitcoin mine If you mine one bitcoin or sorry, that's actually 8.4 bitcoin worth of gold So even mining gold is relatively more efficient than mining bitcoin um Now those numbers are Big but what I find particularly interesting or Actually, you know shocking is the rate of change Because we went To an energy to the energy consumption of a country like switzerland in less than a year Um, this is a nice little graphic produced by the columnist They don't try to come up with a realistic energy Energy consumption number. They just show the minimum energy consumption for the network Um, and as you can see Clearly in 2018 it completely exploded where it is was Negligible in the years before Um, by the way, they also know that the energy use cat rising after the bitcoin bubble burst That's where my work comes in. So I actually predict and explain why this happened In my last year's paper titled bitcoins growing energy Consumption it was published in may of last year and the data use was from march um But coming back at the changes that happened So we were in 2017 looking at 16 terawatt hours of energy electrical energy per year In 2018 we went up 63 almost 300 increase in one year per transaction 350 percent why? Because the amount of transactions that were proposed actually went down in 2018 More energy use less transactions Um, oh, I should note uh at the start of the day someone mentioned There was a prediction that bitcoin would boil the ocean someone took Mine my work They took these percentages and then kept on extrapolating with it if you actually read my work Um, they should have realized they should have stopped At the uh at the last numbers just a side note So, okay, what does this mean in in terms of environmental impact? In order to know that we first have to figure out, okay, where is mining actually taking place because The carbon intensity depends on where you're mining what power your source you're using um, the global Cryptocurrency benchmarking study shows that in 2017 there was a majority of the network located in china And the year after that that Changed a little but not so much by the way It's really hard to establish where a miners are located ultimately Because even this map Has a few issues you can see the dot on cubaq is really large Under oath the Power company in cubaq hydro cubaq declared for the local energy board that they were only supplying 15 megawatts Worth of power two miners and they were forecasting 110 for 2019 Um, so obviously the real size of the dot should be as indicated much smaller Okay, we can get back to deck after so the point And my only key point is that the majority is closed is in china, right? So that's and the trend is still that miners are moving to china. Why? because There's the miners are being produced in china bit miners in china and it's getting more expensive to move these machines outside of china So they were located in china and they are probably going to be in china for some time more unless China bans bitcoin or specifically bitcoin mining and they were considering that some time ago We'll see if that actually happens, but if that doesn't happen We'll see more miners moving to china and the chinese grid isn't exactly clean Well, if you're paying 25 teras, good luck with that. So anyway, I made a carbon intensity Estimate for the whole network based on the previous mentioned studies and ultimately that comes down to the carbon intensity of 455 grams Of carbon dioxide per kilowatt hour Now as long as the majority of mining is taking place in china, you're going to end up with this carbon intensity I'll explain later why But what does that mean in terms of carbon footprint for the whole network So the whole network has a carbon footprint of about 30 megatons The interesting bit is The carbon footprint per bitcoin transactions 370 kilograms of carbon dioxide The same carbon footprint as 900 000 over 900 000 visa transactions or over 6000 hours of watching tv You can't make this look any prettier. You can come up with the craziest numbers You can say oh the whole network is being run on renewable energy Even still renewable energy does have some Carbon footprint left if you take one percent of my carbon intensity and apply that to the network You're left with a carbon intensity per bitcoin transactions of 3.7 kilograms still ridiculous Because there's still 9 000 visa transactions You can't put lipstick on a pig So What are these solutions getting back to renewable energy? There's a few interesting considerations when You're thinking about applying renewable energy to bitcoin mining The first thing is renewable energies are generally an intermittent energy source If you're looking at shi shuan in china During the wet season production three times higher than in the dry season That's production then you also have fluctuations in demand if you look at kebek During the wintertime they depend on electrical heating The mon goes up 65 percent The surpluses we keep hearing that bitcoin mining is done with surpluses that nobody uses those surpluses aren't around all the time Especially not when your production drops or your demand peaks bitcoin mining Increases your base load demand These machines are running 24 7 non-stop So you have to supply supply them with power at all times It's fundamentally Let's say tough to combine with renewable energy sources This figure gives a little bit more illustration on why that is So if you have a new generation of mining machines The amount of money They will earn will decrease rapidly simply because everyone wants to have the most efficient machines Everyone adds machines to the network the network hash rates goes up And a single machine performs less and less and less So each day or each second of the day that you shut off You'll miss out on a level of income that will never come back Um second problem with renewable energy. It's Ideally green, but technically it's not always green Um she's one again And uh, we just I'm like she's one because the reference that was made just a Few minutes ago was to the coin shares report that said half of the mining is done In she's one which by the way, if that is true, wouldn't be very great in terms of decentralization But we'll keep that on the side um If that is true, then still she's one doesn't really have green green energy Not even from hydropower because they struggle with well first of all the reason there's Excesses on their grid Is because they can't export it, but obviously when the production falls Yeah on the left Three times more production during the wet season When production falls they need coal based power And they can't import it because the same reason they can't export their access So what do they do? They actually construct more coal based power plants locally to act as a backup for when their renewable production falls Ouch So ultimately if you look at this she's she's from province the carbon intensity of the grid Is actually comparing comparable to using natural gas for mining That's not green at all Um last note We're talking a lot about how the hydropower um reservoirs Can be significant sources of methane Now that's still being researched But I just want to say that methane is definitely a more potent greenhouse gas Then carbon dioxide actually A factor 34 times more so Hydropower isn't the ideal source In any case Oh, yeah, and there's one more thing No matter how much green energy you put into the bitcoin network you're always going to be left with electronic waste Um, this is some footage from last year where bitcoin mining machines were being thrown out And this is often forgotten Because everyone talks about energy use, but this is specialized hardware designed with a single purpose mining bitcoin and when they're done doing that They're obsolete and you can Either bring them to the incinerator or to a landfill Um altogether the network has a electronic waste potential of 11 kilotons per year By the way, this is all from my latest paper renewable energy will not solve bitcoin sustainability problems published in march this year in jewel um And this electronic waste problem or electronic waste amount As I was saying is comparable to what a country small country like luxembourg is putting out here But it comes on top Of the electrical energy consumption as a country like switzerland. It just adds up um 135 grams per bitcoin transaction. That's the footprint Um, I made a very optimistic maximum on uh visa Uh, I assumed each of their servers weighed 100 kilos and they were being replaced every year Then still The e-waste footprint Is 30 000 times bigger for a bitcoin transaction And as I just noted most of that just goes to environmentally damaging landfills and incinerators So you'll never have a green bitcoin. No way. Oh This whole story doesn't include the production process And it's just about mining. There's also an ecosystem that we're not considering and there's parties like coinbase, bitpay, whatever, uh, bitcoin adms not considered So how about the lighting network? Well, I have a question for you Let's assume that The lighting network or whatever solution is found for bitcoin scaling problem and we can Do the entire banking industry on top of big time. We don't need the banking industry anymore What's going to happen to the value of bitcoin? What do you think is it going to be the same value? Is it going to be increasing 10 times 100 times 1000 times? Or even more maybe a little show of hands who thinks it's going to be the same value 10 times increase 100 times Okay, now seeing some hands 1000 times More than a thousand times Okay, definitely no less than 100 times Cool, but what does that mean in terms of energy consumption? So my work that was published last year establishes a relationship between value and energy consumption And applying gets very strictly would result in these numbers The funny thing is even if bitcoin Would consume the same amount of energy as it's consuming currently It's still less efficient than the current system Why because the global data center energy uses around 200 terawatt hours Um And obviously visa is a very efficient example But you can estimate the Banking need to be maybe around 20 terawatt hours At a 10 times increase Bitcoin will most likely exceed The total production we're getting from solar The solar PV capacity is 600 400 and 60 terawatt hours per year. We took decades to get to that number Um at 100 times the relationship by the way the bigger this number gets the less linear this relationship will get But let me just summarize it like If it stays the same it's going to be bad if it's going to be a 10 times increase It's going to be a very serious problem At a 100 times increase it's going to be a nightmare and at the 1000 times increase or more. We're probably all dead Sorry yeah, so That's why I was actually going to say this is a bit too much of a simplification But obviously transaction fees are going to be more relevant in the future There's people predicting there will be a 100 000 transaction fee On with the lighting network which would then replace the block reward and if that happens You actually Will we probably be consuming 80 to 90 percent of the global um energy consumption Sorry, what? Well, it would be it would be nice if it was true because then we really need a Dyson sphere to make it happen Um, but by the way, you just predicted a thousand times price increase Um, if I just go back a little Uh, then you also need to Dyson sphere to make it happen because it's Almost 300 percent of a current global energy production And by the way, that's your prediction Yeah, I'm gonna wrap up So what would work? Well, for example, replacing the proof of work with a proof of stake algorithm. I'm not I'm hearing there's a lot of Fans for that idea But what it would achieve if it was done It would reduce the energy need by 99.99 percent And you would do away with the incentive to reduce to produce singular purpose specialized single purpose hardware Obviously, I'm okay with whatever alternative as long as it works And by the way Proof of work is not bitcoin people Somebody on my twitter feed pointed out nicely that Uh Proof of work is designed for a very specific goal Security immutability and censorship resistant. It is not a magic wand That turns air into bitcoins So as long as we keep the eye on the objective Then we might be able to find a better alternative than the current proof of work system So we can really replace these steps into running a blockchain network. That's all