Is Bitcoin Dangerous?
I’m crossposting a blog post I wrote here.
It’s not exactly aimed at EAs, but I think EAs would find it interesting.
Two sentence summary: Bitcoin gets a lot of criticism for energy usage, but even removing all cryptocurrencies would have a small impact on global emissions. Nonetheless, the actual novel uses of blockchain technology radically changes funding models and could be extraordinarily dangerous to catastrophic risk coordination.
Critics of Bitcoin often point to its large energy usage as a demonstration of its catastrophic waste. In this view, Bitcoin is bad because it’s utilizing large amounts of energy for something completely pointless since Bitcoin has no real value or benefit, and thus we are destroying the environment and undoing the entirety of our advances in green energy in order to allow Silicon Valley bros to try and make lots of money by scamming each other until it all collapses.
This view gets things exactly backwards; Bitcoin’s energy usage isn’t particularly scary (at least compared to other catastrophic risks), but Bitcoin is pretty valuable, and in fact decentralized blockchains are essentially an unstoppable and wide-reaching force that can never be put back in the bottle regardless of whether their existence is beneficial to the broader world.
Energy Usage
Energy usage differs from emissions. The good news is that Bitcoin users pay for their energy usage. Every 10 minutes, a Bitcoin block reward is auctioned off to Bitcoin miners. Naturally, miners will expend resources until it is no longer profitable to do so, and we reach a natural equilibrium where every 10 minutes, about 1 block reward’s worth of costs is imposed on miners (mostly through electricity) in return for 1 block reward. Pretty straightforward.
Note there is a common misconception that Bitcoin’s energy usage is tied to the number of transactions placed on chain. In reality, energy usage is set by the block reward which is locked by an algorithm (where it declines slowly over time). Transaction throughput is a function of blockchain space, but broader scaling is likely to come through so called “layer 2” scaling, where transactions largely take place off-chain, and only post summaries for final settlement to the blockchain. The number of transactions on a blockchain locked at today’s energy usage could be effectively unlimited under a system like drivechains.
This year, assuming a Bitcoin price of $40,000, we’ll see something in the neighborhood of $13 billion in Bitcoin revenue flow from users to miners. Users pay for this miner revenue either in direct fees or expected inflation, which is known in advance and priced into Bitcoin, reducing value in expectation of future coins being in circulation. This flow of value is about as interesting or concerning as the revenue of a similar sized small company. For example, it’s less than the annual revenue of DraftKings ($22.54 billion) or Dollar Tree ($27 billion). Small companies making this kind of money isn’t particularly concerning or worrying in any way, and neither is the Bitcoin network.
However, externalities are concerning; even if you don’t buy or use CO2 emitting cars, you have to deal with the externalities from climate change. To the extent that Bitcoin network energy is producing greenhouse gases, we should care about its negative effects.
So how much does Bitcoin pollute? According to a 2019 study, in 2018, the Bitcoin network produced 17.29 megatons CO2 equivalent. For reference, in 2018, the United States as a whole produced 5,276 megatons of CO2 equivalent, or 300X more.
But the study might actually overestimate emissions. They basically took the best estimates they had of where Bitcoin mining occurs and used the local energy breakdown applied to Bitcoin. This is extremely unlikely to reflect reality; Bitcoin miners will only use the cheapest available energy in the world because proof-of-work is extremely competitive. Different forms of energy are likely to have different prices; seasonal rains will make hydroelectric power in western China dirt cheap to where coal cannot compete.
Moreover, earlier this year, China kicked out Bitcoin miners from one of its most coal-intensive energy regions, Inner Mongolia. This trend has continued and China is now kicking out all Bitcoin miners. Assuming this goes through, this bodes well for concerns about Bitcoin’s energy usage, as the next best place to place miners is the United States. Here, renewables have been gaining significantly over the last decade. However, the price of Bitcoin has risen significantly, which will cause overall emissions to rise, even if a lot of Bitcoin mining is done with renewables.
If we try to estimate an upper bound for Bitcoin’s 2021 emissions, we can start with the $13 billion for miners mentioned above. Let’s assume the cost per kilowatt hour to be 5 cents on average for Bitcoin miners. This is much lower than the US average of 13 cents, given the high competition between miners. However, lower electricity costs actually increase energy used at a given revenue level, so this assumption will help us set an upper bound by pricing energy lower. Finally, let’s assume an average CO2 per kilowatt hour of 0.85 pounds from the U.S. EIA, even though the cheapest energy tends to be renewables.
Upper bound =
$13 billion
* 1 kiloWatt hour / $0.05
* 0.85 pounds CO2 / kiloWatt hour
* 1 metric ton / 2204.62 pounds
= 100.2 megatons CO2For reference, the IEA estimates that US emissions will actually be lower than the 2018 figure used in the chart above (4460 megatons), even as Bitcoin emissions will be higher. The Bitcoin network would now have an upperbound estimate of almost 2% of U.S. emissions. This number can be interpreted differently by different people; cryptocurrency enthusiasts will say this is a pretty small price to pay for what blockchains allow, while skeptics will say this is an enormous amount of energy even if the percentage of total emissions remains low.
Regardless, even if you consider carbon emissions to be the number one problem in the world, magically removing all Bitcoin mining at most reduces global emissions by 0.28% (global emissions estimated at 36.4GtCO2) and probably much less.
Moreover, a carbon tax would be a good idea regardless of Bitcoin and indeed have a much larger impact than Bitcoin simply disappearing. Even an extremely moderate carbon tax that reduced US emissions by 5% would massively outweigh any emissions from the entire Bitcoin network. Bitcoin is contributing to this problem, but a world without Bitcoin would face a climate change challenge virtually unchanged from our current plight. Making Bitcoin mining illegal wouldn’t even get us to this scenario, as much of it would go underground or overseas.
For a more in depth discussion from more knowledgeable people, here is a discussion from Coin Center.
Pandora’s Box
I’ve outlined why concerns about Bitcoin wasting energy are misplaced, but the errors I see made by Bitcoin critics go much deeper. It’s common to view cryptocurrencies as largely useless, which is why emissions loom so large as a problem. But Bitcoin is not useless. And because many critics are blind to how Bitcoin can be used, they fail to see the true threat cryptocurrencies could bring.
Before 2009, it was literally impossible to send digital value with no specific third party. Four years ago I wrote an essay discussing potential Bitcoin value scenarios. I would recommend that piece for a more in depth discussion of Bitcoin’s value, but I will summarize several of the points here. First, Bitcoin and other cryptocurrencies already offer an alternative to international wire transfers with lower fees and potentially routing around capital controls that a traditional banking sector cannot do. Cryptocurrencies also offer alternative monetary policy to people who live in nations with poorly governed currencies. Bitcoin has done significantly better than Venezuelan Bolivars and you can obtain it with only an internet connection, whereas buying dollars through official channels is challenging if not impossible. There is also the important quality that if you control your own keys, then it’s hard for a government to seize your cryptocurrency assets (or perhaps more usefully, know you have crypto assets at all). In a developed country, this may not matter as you will need to follow the law in order to participate in regular commerce, but in a state with lax property rights or lackadaisical rule of law, being able to transact without having to trust the courts or state banks may be quite valuable.
Finally, Bitcoin offers uncensorable transactions. There are many applications of this, most of which are criminal in some way or another. Some of these applications could be in areas where laws are actually quite unjust, or in fact it’s not quite laws but restrictive banking policies. In 2011, the FDIC suggested people engaged in several questionable industries like dating apps, pornography, or fireworks could be subject to financial scrutiny (presumably because these could be used as covers for illegal industries like prostitution and explosive weapons).
The use of cryptocurrencies to route around legal restrictions can be seen as a threat by Bitcoin critics and as a feature by Bitcoin proponents. Bitcoin allows WikiLeaks to raise money despite them being kicked off of MasterCard’s and Visa’s networks. Hamas has raised money through cryptocurrencies since their fundraising is largely blocked through traditional means. Are these features of an uncensorable free payment network, or illegal applications of a criminal fundraising system? It depends on your politics.
The most remarkable implication that cryptocurrency critics have yet to reckon with is that there isn’t much that can be done. For example, there have been several high profile ransomware attacks this year. There will be more. There is nothing we can do to stop this. There is no returning to a world without cryptocurrencies. You could try to make all on-ramps to cryptocurrencies register with the government, but you’ll miss hand-to-hand exchanges like LocalBitcoins, and you’ll miss foreign exchanges entirely. In some markets, automated exchanges running on other blockchains are already available. You could try to ban cryptocurrencies entirely, but cryptocurrencies just run on the internet. You could try to do intense blockchain analysis, but completely private cryptocurrencies already exist like Monero and Zcash, with new technologies like mimblewimble already on the way.
There is also the possibility of threats to monetary policy. Some Bitcoin advocates envision a “hyper-bitcoinization” where Bitcoin elbows out the dollar as the world’s primary transactional unit. Tyler Cowen argues that cryptocurrencies are unlikely to supplant developed nations’ currencies, but they could easily do such things to smaller countries. In places like Venezuela, we’re already seeing significant cryptocurrency usage as the government destroys their own currency’s value, and it’s much easier to import Bitcoin than dollar bills (stablecoins, a sort of digital private banknote, might be better still). However, not all small countries will mismanage their currency like Venezuela did. And having a central bank capable of expanding the money supply can be valuable in some situations (just ask Greece). But if many citizens of a developing country simply decide it would be better to own USDC or Bitcoin or Ethereum instead of their local currency, even mild monetary policy could be ineffective.
Tax evasion is a further challenge for low capacity states. Presumably if cryptocurrencies are restricting the maneuverability of developing countries’ monetary policy, they could also make tax collection more difficult as well. If citizens are holding a lot of their liquidity in cryptocurrencies, the default for most transactions is pseudonymous or perhaps even more broadly private in the future as the Bitcoin protocol is updated (e.g. Taproot). That could make it easier to avoid taxes if people were so inclined. Cryptocurrency advocates might be philosophically opposed to government taxation in some respects, but from a policy perspective, it would be challenging for a country to improve its state capacity if the state is unable to collect revenue.
Another development is the rise of financial interest in Bitcoin. A few months back, the Senate accidentally ran right into a huge kerfuffle with the cryptocurrency industry which delayed and sidetracked the Biden infrastructure bill. The industry’s influence is already enough to make headlines during a major (unrelated) piece of legislation. This influence is only likely to continue growing. That puts a cap on the political will to clamp down on this technology.
Existential Risk
Finally we arrive at the largest problems. In The Precipice, Toby Ord discusses the most dangerous challenges humanity faces in the next century. Advancing technology means year over year, the financial cost of creating more dangerous products of human ingenuity continues to drop. The invention of the atomic bomb is the most salient example. Over the years, more and more countries have added themselves to the list of nuclear powers. For the most part, humanity has taken steps to avoid the usual effect of technological advances on new inventions; nuclear weapons are not as cheap and ubiquitous as smartphones and TVs.
But weaponized biology has not gotten the same attention as nuclear arms while DNA sequencing and editing has become ever cheaper. The global experience of the COVID-19 Pandemic is also a poor reflection of our resilience to biological challenges. Consider if a terrorist group deliberately created a more deadly or contagious virus.
Any independent terrorist group that attempts such an attack will require specialized knowledge, equipment, and funding. It’s unclear to what extent knowledge of dangerous pathogens can be contained. Regulating equipment that could be used to “print” the DNA of dangerous pathogens may be plausible. But unfortunately cryptocurrencies makes stopping the funding of terrorists basically impossible.
It’s not that Bitcoin will directly cause a frightening biological weapons attack; rather, Bitcoin fundamentally changes the nature of what’s legally and financially possible. Indeed, the upshot of this new technology may ultimately be net positive, but the risks brought on by cryptocurrency are not relegated to a few additional tons of carbon; they are radical changes in a world where illegal funding is unstoppable and perhaps ubiquitous.
Solving existential risks requires coordination. Cryptocurrencies in one sense allow for a decentralized form of coordination; perhaps in the future, EA organizations will be able to incentivize important risk-reduction work through novel smart contracts or DAOs. But it’s just as possible that this technology will make coordination more difficult through decentralization.
And to reiterate, you can’t put cryptocurrency back in the box. There is no organization or person to arrest; the code is out there and the Bitcoin network’s very nature means every node is equal on the network and consensus can’t be changed by force. It’s my opinion then that cryptocurrencies’ existence must be taken as a given, as efforts to stamp it out may only work to push it underground. The radical benefits of cryptocurrencies would then be held out of reach for most law-abiding citizens, leaving only those the law deems criminal to reap the benefits.
One observation though: things can be good or bad for the world independent of whether they are a good investment. In other words, cryptocurrencies could still have massive returns even if you think it’s bad that they were invented. From an effective altruist perspective, it may be best to invest in crypto so that you can control more of the windfall of the results of this technology, and then use those resources to focus harder on existential risk.
Regardless, the best way to think about Bitcoin is to understand the entirety of how it works and why it’s so transformative. Focusing solely on energy usage mistakes Bitcoin as a minor skirmish in a larger climate discussion, when in reality Bitcoin is a figurative gravitational superweapon. Whether it will be used to divert deadly Earth-destroying asteroids or accelerate them is still up for debate!
“Bitcoin fundamentally changes the nature of what’s legally and financially possible.”
Not really, no. (Ethereum? Maybe—but even that’s unclear, and it certainly doesn’t change what’s legal on it’s own.)
”...cryptocurrencies makes stopping the funding of terrorists basically impossible.”
No. Really, really, no. I could talk a lot more about this, but if you think terrorist groups can manage infosec well enough to overcome concerted attacks by the NSA, or Mossad, or FSB, etc., you’re fooling yourself.
Other than the misunderstanding and conflating nodes for hash power, this is also not true. Hash power is concentrated, so you’d need to somehow convince the biggest mining groups that they don’t care about countries keeping their operations legal, and as we’ve seen, they do. That means they (ETA: And other key cryptocurrency players, such as exchanges,) will continue to run to embrace KYC/AMF regulation, and will do whatever else makes their investments go well—including cooperating with nation-states in almost any way you can imagine.
“Impossible” might be an exaggeration, but it does seem to make it much easier. That’s also what the article you link to suggests. Edit: Are you skeptical because of the on/off ramps, the security of terrorist’s computer infrastructure or something else?
So far, the only serious KYC/AMF happens at the level of centralized exchanges. Nation-states cannot enforce KYC/AMF at the level of decentralized exchanges. They can also use chain-analysis and put pressure on mining groups within their countries to do KYC/AMF or to create address “black lists” but so far there hasn’t been much political will for this, and probably would lead to a big backlash from the crypto community. And this becomes impossible for privacy coins such as Zcash and Monero.
It’s not harder to stop, it’s easier—blockchain is far easier to trace than cash, which is what is used now. And I’m skeptical for both of those reasons.
NatSec people already have large groups working on this, and my understanding is that even the best BTC wallets and mixers aren’t actually much help. And despite limited political will, OFAC has added people and addresses, and as far as I’m aware, no one has successfully moved money out of any of the (very few) OFAC-blocked addresses. Maybe that’s not pools enforcing the rules, and obviously blocking individual addresses is incredibly naïve as a strategy, but I will note that pools and other key players have every reason not to want to piss off the US Government.
I don’t follow your point about infosec. The RAND link summary seems to argue that the case against crypto is that it may not be widespread enough to be a good money laundering source, not that laundering won’t work. Or maybe that’s what you’re saying? So sure, I agree the NSA will continue to target terrorist groups...but they can’t do it through monero or zcash! There just isn’t enough information leakage. But maybe your point is just that there are enough other attack surfaces that intelligence services will continue to target them?
(edit: just saw you are the author, ha! So I took a few more minutes to read, and tbh I’m not sure we disagree. I understand Bitcoin can be somewhat deanonymized, but there are lots of easy ways to make that harder, and monero and zcash seem very private. My point is more that this contributes additional risk overall and we even see examples of this already even w/o e.g. monero being widely accepted.)
I suppose that is somewhat of a rebuff to my thesis as written, perhaps too provocatively. I guess what I mean to say is that *some funding will leak through* and stopping all illegal funding through cryptocurrencies is an impossible task (but perhaps a silly point as maybe you could say similar things about cash). And by proof, I’d present that Hamas already uses cryptocurrencies to route around sanctions and regulations, as does North Korea. My concern is that you only need one crazy terrorist group to make a superflu, and I think cryptocurrencies could make that more likely.
Hmm, this doesn’t make sense to me. Bitcoin miners don’t have any KYC requirements, and don’t track that information. That’s why people were annoyed with the Biden Infrastructure Bill, as it appeared to change the rules, effectively making mining illegal in the US (obviously lots of regulatory guidance still to be issued, etc.). Coin Center covered a lot of that, here and here probably good places to start. But yeah I’ve done a quick 5 minute check and I’ve found many Bitcoin mining pools, none of which have KYC. Maybe you could elaborate what you mean here?
Also, this isn’t particularly relevant but I find it interesting (and controversial!) so I’ll dive in a bit: I maintain that nodes (that is, users) determine consensus in Bitcoin, not hashpower. If you take 90% of Bitcoin’s hashpower and start breaking consensus rules, you can mine blocks, but no one will accept them. It would be chaotic and crazy for sure, and of course would result in a hard fork. But the mainchain would continue at 10% of mining hashpower, all the exchanges and users would just ignore the other chain because those blocks break consensus rules. It’d be slow for a bit, but it wouldn’t be a huge problem. The 90% hashpower would lose all the resources they sunk into their forked chain that no one uses. We even saw this play out in 2017 when Bitcoin Cash hard forked with a bunch of hashpower. The main Bitcoin chain kept going, because most nodes stayed with the main chain regardless of where hashpower went.
Right. So terrorism will likely neither be materially helped nor significantly harmed.
Yeah, I didn’t mean that miners themselves would be subject to KYC, I meant that the key stakeholders in Bitcoin will want to ensure that the government won’t ban it, which would hurt its value. That means that the vast majority will support things that require AML/KYC for most bitcoin usage—and if miners were faced with a choice between regulation and rules which hurt prices or KYC, they would pick KYC in a heartbeat. And we see that it’s happened everwhere else—all the exchanges, all the consumer front ends, and a vast majority of transaction go through places which have KYC requirements—meaning that not only is most of it tracked, but the remaining fraction of the blockchain is far easier to deanonymize.
Thank you for writing this piece! I think that there should be a serious discussion if crypto is net positive or negative for the world.
In my opinion, there are a few more ways how crypto could contribute to existential risk. Since you can accept donations in monero, it is much easier to make a living by spreading dangerous ideologies (human extinction is a worthy goal, political measures against existential risk are totalitarian, etc.) Of course, you can also support an atheist blogger in Iran or a whistleblower in China by crypto, but it is very hard to tell if the advantages or disadvantages weigh more.
Moreover, crypto can be used to fund more dangerous stuff than “just” artificial pathogens. Think of AGIs that are build for a criminal purpose, imperfect mind uploads that suffer, or perfect mind uploads that are build with the intention to torture them, underground companies that avoid AI regulation in order to cut costs etc.
These scenarios do not prove that cryptocurrencies are net negative; especially since it may be possible to build DAOs that will solve some coordination problems. Nevertheless, I would be happy if more smart people were thinking hard about these issues.
One downside of decentralization you missed is the idea that protocols are slower to update than any other software, which in some scenarios leads to a lock-in risk.
To be more specific, suppose mechanism designer A encodes beliefs/values/aesthetics X into a mechanism M, which gets deployed in a robustly decentralized fashion. Then, upon philosophical breakthroughs totally updating X into X’, A encodes X’ into a new mechanism M’. The troubling idea I’m pointing to is that coordinating the pivot from M to M’ seems exceedingly difficult, likely much more difficult than coordination in the absence of a robustly decentralized fashion. And this is only in a world of one mechanism designer: things get much more troubling in the real world of many competing mechanism designers A and many competing beliefs/values/aesthetics X.