one of the great things about the digital economy is that it offers the potential to dematerialize aspects of our economy to do more things with less resources of course there's some element of myth about all this but in this context the setup around bitcoin is quite an outlier it doesn't offer to do more with less it doesn't even offer to do the same thing with less built into the bitcoin algorithm is the offer to do the same function each year with an ever growing use of resources in this video i'm going to explain why that's
the case and explore the implications [Music] before we continue let me just say i think bitcoin is fascinating and i believe cryptocurrencies are here to stay and offer interesting new ideas for how we might organize parts of our political economies the kinds of big ideas i'm interested to explore on this channel so while there are clearly a lot of very dodgy activities connected to various aspects of cryptocurrencies i do not think that they're all scams that should be shunned but bitcoin does have a problem a deep tension between its algorithm for security and its ever
rising costs and while this tension does connect to energy usage and climate change that is not the key point of this video as i'll explain in a moment but first i just want to quickly explain a few relevant aspects of how bitcoin works for anyone watching who is not familiar with those details and to be clear none of what i'm saying in this video is investment advice and i'm not making any predictions about the future price of bitcoin okay very briefly bitcoin is a digital currency that works by running on a network of computers all
around the world so there is no single database that holds all the information about this currency rather every node on the network has a copy of the entire ledger of all the bitcoin transactions each transaction in this ledger records a transfer of coins from one person's digital wallet to another and the crucial point is that people trust the information in these distributed ledgers because of the way that the so-called mining algorithm works roughly speaking every 10 minutes a collection of the latest bitcoin transactions is gathered up into a new block the algorithm for mining is
essentially about securely wrapping this new block with a digital signature so that it can never be changed again the first person to do this task gets paid an amount of bitcoin as a reward what's particularly clever about the way that this secure wrapping occurs is that each new block gets linked to all previous blocks so that the whole history of transactions cannot be tampered with and this linking makes the so-called blockchain now it's actually trivially easy to wrap a block with any old digital signature it generally takes less than a millisecond for a computer to
do the maths for creating the so-called hash of the block that is used as the digital signature these hashes are essentially a hexadecimal number that can be used to prove that the block hasn't been altered in any way but there is a catch to make the task harder the bitcoin network sets a difficulty for what will be considered an acceptable hash and roughly speaking this difficulty is expressed as the number of zeros that have to be at the start of the hash a hash with a few zeros at the start is relatively easy to find
a hash with lots of zeros is incredibly hard to find so this transforms the trivial task into an extremely challenging search for a needle in a massive haystack to give you a vague sense of scale the current hash rate for bitcoin is around 140 tera hashes per second which written down looks like this that's how many pieces of hay are being checked each second to find the one needle and the person who finds it first gets a reward that is currently around a quarter of a million dollars that's a valuable needle and the bitcoin difficulty
is set so that on average it will take roughly 10 minutes before someone gets lucky and finds the needle that is they find a hash with the correct number of zeros at the start now a crucial point for this video is that roughly every two weeks this difficulty level is adjusted if more people start looking for the needle the difficulty is made harder by requiring more zeros at the start of the hash if less people are looking it becomes easier again and here's a graph from blockchain.com of how the difficulty has changed over the last
couple of years generally it is getting harder and harder to find that needle all of this effort to find the needle is part of what's called the proof of work approach to securing the trustability of the bitcoin transactions for this security to work it can't be the case that a single individual or organization can perform over 50 percent of the work to find the needle otherwise for reasons i won't go into here they would be able to create fake exchanges of bitcoins they'd be able to steal bitcoins by a so-called double spend attack so we've
got our first two key ingredients for the argument i'm making here firstly bitcoin difficulty adjusts so that it always takes around 10 minutes to mine a block and secondly to keep secure no single mining organization often known as a mining pool can do more than 50 percent of the mining work the final ingredient to look at is how the cost to compute changes over time and this graph comes from researchers interested in the changing costs of computing the maths functions used for ai roughly every four years you can do 10 times the number of calculations
per second for every pound or dollar you spend and note this is not only about computers getting faster but also about the cost of buying computers getting cheaper so this is our third and final piece of the puzzle which we can rephrase as the cost of compute gets 10 times cheaper roughly every four years and bitcoin mining is an algorithm that can take advantage of this particularly well because you can get many different machines to each search for the needle independently it's a highly parallelizable algorithm but here's the problem if we imagine the bitcoin difficulty
stayed the same for the next four years then the cost of compute necessary to find the needle in 10 minutes would drop by roughly 10 times in eight years the cost would drop by roughly a hundred times now some bitcoin advocates might hear this and think great a hundred times cheaper probably much more energy efficient what exactly is the problem well let's look at a simple scenario imagine there are four different mining pools each of whom currently mine 25 of the total hash rate let's imagine they're each mining at a rate of 100 terra hashes
per second so as things stand no one has more than 50 percent of the hash rate so no one can steal bitcoins all good now four years later for the same cost you can do 10 times more compute or you can choose to stay at the same hash rate while spending 10 times less money let's imagine that three of the mining pools decide to go down this second route and take these efficiencies in reduced costs of mining and so they keep their hash rate at a hundred tera hashes but one mining pool mine fast makes
a different choice they keep their costs the same and increase their total hash rate by ten times taking it up to a thousand tera hashes per second this will mean that minefast are now performing 76 of the total hash rate well above the 50 level at which they could start stealing coins another four years of the same choices and mine fast would be up to 10 000 tera hashes per second and 97 of the total hash rate while still operating at the same costs as before and of course the difficulty of finding the needle will
have adjusted over this time period so that it is still taking roughly 10 minutes for someone to find the needle but by now it would be mine fast finding the needle 97 of the time and each of the others finding it just one percent of the time so minefast would then completely dominate the bitcoin network in a way that is totally untenable for the ongoing trust in the integrity of the blockchain the core argument for why bitcoin is secure would be gone so in reality the other mining pools would have no option but to follow
the choice of minefast and roughly keep pace with each other as they increase their hash rate the speed at which they search for the precious needle and any mining pool that instead tries to take the efficiencies as cost savings would soon become largely irrelevant and because there is no governance mechanism to stop a mining pool like minefast from taking this route and huge economic incentives to race fastest to find the needle bitcoin has no mechanism to stop this dynamic from happening and indeed this pattern should hold true for the past as well as into the
future and it does or rather this simple picture i've depicted is a massive underestimate of the rate of growth in the use of computer power by the bitcoin network a fascinating paper from 2020 which i've linked to in the description below looked into the 10-year change in the amount of compute used by the bitcoin network from 2010 to 2020 and they found a 10 billion fold increase in the hash rate over that decade this is a much higher growth rate than i've been talking about so far and that's because of a second dynamic that the
authors describe in their paper in order to keep the blockchain system secure from double spending attacks the proof of work must cost a sizable fraction of the value that can be transferred through the network we estimate that in the bitcoin network this fraction is of the order of one percent in other words as the adoption and use of bitcoin grows there must be a proportionate growth in the overall cost of compute effort used by the proof-of-work algorithm that securely wraps each new block and in the decade from 2010 to 2020 the volume of daily transactions
on the bitcoin network rose by around a million fold from around a thousand dollars per day to nearly a billion dollars a day in 2020 so that will have created a pressure for a million-fold increase in the total costs of compute and assuming that the scale of bitcoin adoption continues to grow then this second pressure for growth in the amount of compute will also continue now bitcoin enthusiasts may well say that even as the total cost of mining rises spending just one percent to secure the network is good value for money but remember that this
rise in the total costs is on top of the rise and efficiency of computers that i talked about earlier the earlier scenario showed how all of the efficiency gains of cheaper faster computers will get absorbed by the competing miners buying new mining computers just so that they can stay relevant even if we assume that the costs remain constant but now we've also identified a separate reason why the mining costs will actually rise so the combination of these two pressures implies an ongoing exponential growth in the amount of actual compute being bought and used by the
miners securing the bitcoin network and of course the bitcoin algorithm will regularly update the difficulty so that it will always still take roughly 10 minutes before someone is likely to find an appropriate hash to securely wrap each new block on the blockchain so this is what i meant at the start even with all of this extra work all of this extra compute just the same function is achieved every 10 minutes it's a fundamental dynamic built into the proof of work part of the bitcoin algorithm and there is no upper limit to this growth than the
computer usage by bitcoin there's no stable point in the future where bitcoin is using the maximum amount of compute that it will ever need yes i know that there's a limit to the number of coins that can be minted but that is not relevant here because the two pressures i've talked about are not related to the number of coins being minted and so in the long run all of this will have to be funded by transaction fees and yes now is the moment when the environment does have to be mentioned because an exponentially rising hunger
for compute power means an ever rising consumption of energy and material resources to make and use newer faster better bitcoin mining computers and this will leave an ever growing pile of discarded e-waste in its path this is exactly the pattern we've seen for the last 10 years and there's no reason to think that this pattern won't continue for the next 10 years and the decade after that and so on the us dollar has been around for well over 200 years and bitcoin must surely aspire to have a similar scale of longevity but on this kind
of time scale such ongoing unlimited growth in the consumption of resources is clearly unsustainable simply from a raw physical resources perspective so this is the key point of this video even if climate change wasn't a worry bitcoin is still fundamentally unsustainable in the long run if it wants its adoption to keep growing while continuing to secure its network with proof of work personally i think it's most likely that in the longer term bitcoin will have to follow ethereum's example and move away from using proof of work somehow i find that longer term outcome more plausible
than bitcoin's resource consumption growing forever and more plausible than it disappearing or completely collapsing but obviously such a change to the algorithm would be a radical and difficult thing for the bitcoin community to navigate and it's certainly not going to happen anytime soon and of course as i've said in other videos it's extremely difficult to make confident long-term predictions today was meant to be a sunny morning but what i've presented here is an example of an argument that gets much of its strength from the diligent use of calculations and so in the context of portfoliosm
it's quite a rational argument and to be clear this is not a new argument many others have raised this kind of concern about the proof of work over the years i've just tried to present the issue as clearly as possible what is surprising is how few people in the bitcoin community seem to worry about this well-known long-term problem but also let me restate i think bitcoin is fascinating and i actually would like it to succeed and become a stable and sustainable part of the political economic landscape indeed i think we need to make some significant
changes to the way our political economies work and bitcoin and its community suggests some interesting ideas for this discussion but for bitcoin to have a confident long-term future its resource footprint has to reach a stable steady state and if it sticks with proof of work i can't see how this can happen i'm very curious to hear what bitcoin enthusiasts think of the analysis i've outlined in this video so please do leave comments below but please be polite is there something i've missed or misunderstood what do you think the next few decades could look like for
bitcoin and how do you think bitcoin can become sustainable over the longer term and if you enjoyed this video then please do click on the like button below or subscribe to my channel and thank you for watching [Music] yeah you