17 votes

Why this computer scientist says all cryptocurrency should “die in a fire”

28 comments

  1. vord
    Link
    In addition to all the other great technical, environmental, and political reasons why cryptocurrencies are bad, I'm gonna toss forth my own: a philosophical one that's been brewing in the back of...
    • Exemplary

    In addition to all the other great technical, environmental, and political reasons why cryptocurrencies are bad, I'm gonna toss forth my own: a philosophical one that's been brewing in the back of my head.

    I see the internet as the first resource we have that could be post-scarcity. One that could be run and managed without money. It's not currently, and there is never any shortage of people to say how we need to pay people to build, maintain, and run the internet. And that it must always be this way. But I see a different path.

    Mutual Aid and a Sharing economy. A quick summary on both is: Give what you can, take what you need. At the most basic example would be giving a cup of sugar to a neighbor who ran out. Are you going to weigh it out by the gram and charge them the exact amount it cost you, plus some overhead for the trouble? I'd say all but the most sociopathic would just give it to them. Or my lawnmower just broke and I ask to borrow my neighbors. Any reasonably empathetic person would just loan it, without spelling out a rental contract and fees for returning it in poor condition. And this kind of economy builds trust between the participants. The kind of trust that obviates the need for money.

    So to bring that into an online space, what would that look like? Well, it could look like torrent trackers, where people whom have the resources host the centralized bit which allows the users to exchange data freely. It could be me hosting a website, chat server, and backup solution for myself and friends.

    How does that scale to the actual "providing internet" functions? Fundamentally, the only difference between your local LAN and the entire internet is one of scale. Anybody whom has a Linux server in their home has the ability to learn enough to manage a medium-size community network. Heck, half the stuff is only a few commands and configuration files away if you've got the server already.

    One idea that's been floating in my head for post-COVID activity is to build a community multi-WAN mesh network leveraging average consumer equipment. There's nothing really in the way of technical barriers to doing so, merely personal, legal, and fiscal ones. The fiscal ones primarily just being time needed to meet people and discuss why and how this is beneficial and important.

    To get back to the point: Cryptocurrency stands antithetical to this dream. It's perpetuating this idea of "nobody will do anything unless they're being paid," which is only true insofar that we predicate human survival on getting paid.

    18 votes
  2. [26]
    Macil
    Link
    So, upfront: I'm very interested in cryptocurrency from a software and decentralization perspective, but I agree with a lot of critics that it's irresponsibly promoted as a good investment...
    • Exemplary

    So, upfront: I'm very interested in cryptocurrency from a software and decentralization perspective, but I agree with a lot of critics that it's irresponsibly promoted as a good investment opportunity to many people.

    I feel about this article like how I'd feel about an article titled "Gaming should die in a fire [because it's full of misogyny, racism, in-app purchases, and gambling]". There's issues in the field that people thinking about it should be aware of and the issues should be solved if possible, but ultimately the article doesn't say much about or invalidate the true draw of the field. There's no attempt at all to win over people in the field; they're all just treated as lost causes. Instead a lot of non-universal negatives about parts of the field are loaded onto the reader without much care or nuance to imply that there's absolutely nothing redeemable. ... I thought I was being too harsh on the article for a moment there and then remembered what its literal title is.

    I've seen people I've known harassed within various online communities for being interested in cryptocurrency. There's pushback necessary against the popular portrayals of it as a sure get-rich-quick scheme but a lot of it goes too far without caring for correctness or nuance. I wish critical pieces about it focused more on understanding what they were talking about than making readers feel self-righteous. I wrote a bit about some parts of the article below.

    Thoughts on some excerpts

    The problem is: how do you keep somebody from adding to the ledger and faking stuff? Well, that’s where the notion of the “mining” comes in. What the miners are doing is literally wasting tons of electricity to prove that the record is intact, because anybody who would want to attack it has to waste that similar kind of electricity. This creates a couple of real imbalances. Either they’re insecure or they’re inefficient, meaning that if you don’t waste a lot of energy, someone can rewrite history cheaply. If you don’t want people to rewrite history, you have to be wasting tons and tons of resources 24/7, 365. And that’s why Bitcoin burns as much power as a significant country.

    The top cryptocurrencies Bitcoin and Ethereum work this way and it's bad, but most other cryptocurrencies work through Proof of Stake instead of using energy-intensive mining. Proof of Stake accomplishes security without inefficient use of electricity so painting it as an inherent trade-off is false. Some cryptocurrencies are working on moving away from mining to Proof of Stake.

    It’s not many. And worse, it never could work for payments. So we’ve seen waves come and go of companies saying “We’ll accept payments in Bitcoin.” They’re lying. Because they aren’t actually accepting payments in Bitcoin. They are using a service that allows them to price in dollars, presents Bitcoin to the customer, transfers the Bitcoin, turns it into dollars, and so the merchant is getting actual money.

    Is a company lying if they say they accept PayPal, but PayPal actually facilitates some kind of bank transaction behind the scenes? If you pay by a Bank of America check at a store, but the store actually converts that check into a Wells Fargo balance, was someone involved lying? If you go to buy on a site that only supports Bitcoin or a regional payment company that's inconvenient/impossible for you to use, and you pay with Bitcoin, and they convert it to their local currency on receipt, is something bad happening?

    Which means if the system has to balance and you want to buy with Bitcoin and you don’t have Bitcoin, you have to convert dollars to Bitcoin. And this is, by design, a horribly expensive process, because Bitcoin and the cryptocurrencies are fundamentally incompatible with modern finance.

    And if you don't have PayPal and choose the PayPal option on a site ... This is criticizing a nonsense scenario.

    But wait, it’s worse! The problem with the Global Public Square is that it is a single, limited entity, and you have only so much you can add to it at any given time. So Bitcoin burns that much of the world’s electricity to be able to process somewhere between three to seven transactions per second across the entire world.

    This is another spot where Bitcoin is behind other cryptocurrencies which are much faster. So as a criticism of Bitcoin it's right but this shouldn't be extrapolated to all cryptocurrencies.

    Modern finance has this rule that anything electronic needs to be reversible for short periods of time. This allows an undo in case of fraud. Have you had your credit card compromised before? I’ve had my credit card numbers stolen a couple of times. The amount of money I lost is zero. Because we have both good fraud protection and good ability to reverse transactions. That does not exist in the cryptocurrency space. If your cryptocurrency wallet is compromised, all your apes are fudged.

    Cashier's checks and cash aren't reversible. Sometimes those antifraud systems get in the way or have downsides that it's bad that we can't possibly opt out of. It's ridiculous that people receiving donated money / tips like popular streamers and small charities often get penalized because malicious users send tons of transactions and then reverse them, the payment company charges or drops users who have transactions to them reversed despite them not being able to do anything about it. I've had cards frozen while I've been away from home because of false positives from antifraud systems and I'm thankful I happened to be able to opt out from that because I had some cash on me. I've had an account frozen multiple times and then closed with almost no recourse because I tried to send a friend a gift card for his birthday. I keep my computer secure and I wish I could have signed that transaction as something I'm absolutely sure about, don't make this reversible, let me opt out of whatever stupid-ass antifraud system is in control of my money instead of me.

    I can't do anything to make my accounts more secure from being closed for mysterious reasons. I can't do anything to guarantee that support will pay attention to me. I think it's an important freedom that there's alternatives like cash and cryptocurrency that I can opt to control myself. Even if it comes with downsides like theft not being reversible, I'm able to take measures myself to make that less likely. I can't take measures to make it so customer support of my payment accounts takes me more seriously.

    Well, there are classes of payments that the intermediaries don’t allow. The big ones are drug dealing, child sexual abuse material, and ransoms. As a consequence, the cryptocurrency actually used for payments is really only used seriously for: ransomware payments, where companies have to pay $10 million. Drug deals—drug dealers hate it, but it’s the only game in town. And we’ve had cases of websites selling child exploitation material paid with Bitcoin.

    Child sexual abuse material and ransomware are issues I take seriously. I think there's a trade-off against personal liberty where if it was shown that those fields were at a certain large size and specifically wouldn't be if not for cryptocurrency, then my opinions around cryptocurrency would change. But the physical kidnapping ransom industry and plenty of harmful fraudulent industries have existed fine since before cryptocurrency so I have to say I'm skeptical that cryptocurrency uniquely prevents these things from being shut down.

    When it comes to drug dealing ... I think there's a lot wrong with our current drug laws and I'm partial to arguments that privacy tools and escape hatches can be net positive for getting around imperfect laws. (Though it's also an industry that's long predated cryptocurrency.)

    So it doesn’t work for payments. And it doesn’t work economically either. It’s effectively a giant self-assembled Ponzi scheme. You hear about people making money in Bitcoin or cryptocurrency. They only make money because some other sucker lost more. This is very different from the stock market.

    Being an investment vehicle isn't the primary or only point of cryptocurrency.

    Yep. Because it’s a self-created pyramid scheme, you have to keep getting new suckers in. As soon as the number of suckers dries up, it collapses. And because it’s not zero-sum, but deeply negative-sum, there are actually a lot of mechanisms that can cause it to collapse suddenly to zero. We saw this just the other day with the Terra stablecoin and the Luna side token. This was basically another Ponzi scheme implemented in the larger space of Ponzi schemes.

    The collapse of Terra and Luna happened for reasons that had to do with their specific mechanisms rather than anything in common with other cryptocurrencies. The design of Terra was criticized within the cryptocurrency community for being set up in a specifically unstable way and was predicted to fail as it did.

    Its failure says something about the cryptocurrency space in general for letting it get so big, but its failure doesn't say something about each individual cryptocurrency because most have no mechanism as experimental as it.

    I am sure you have heard people say things like “Well, blockchain technology itself has lots of potential applications, it’s really interesting, offers lots of possible solutions to problems.” But one thing you point out in your lecture is that often, they are pretty vague about what these uses are, and usually when you get down to the facts, there’s a much simpler solution to whatever problem it is that wouldn’t use blockchain. You cited the example of someone who touted how blockchain could help with vaccines in India.

    okay okay there's a ton of people who keep saying things like "blockchain can help shipping chains" and I have no idea where that comes from but it is an actual nonsense talking point that keeps coming up somehow out of somewhere from crypto fans. A lot of cryptocurrency fans laugh at this too.

    There are a couple interesting use-cases for blockchains outside of cryptocurrency, but it's mostly stuff like recording who is the current owner of a digital asset like a blockchain-based domain name, or publishing a message in a way that guarantees everyone sees the same message without any ability for you to trick someone into thinking the message was something else (useful for publishing the hash of a software update publicly, so users know your update server isn't secretly giving a subset of users a different malicious update). Most suggestions for real-world non-payment-related uses of a blockchain are nonsense.

    El Salvador

    Yeah everything about what's happening there with cryptocurrency is ridiculous and anyone portraying it as a success of cryptocurrency is getting high on their own supply.

    So even when you have a central database and a central authority, cryptocurrencies don’t work for payments, because they bounce around in price.

    Well yeah, this is a worst-of-both-worlds scenario. The downsides of cryptocurrencies almost entirely come from sacrifices they make in order to be decentralized. The downsides have no reason to exist if you remove the decentralization. This would be like comparing a car seat to a comfortable couch, and finding that the car seat is still less comfortable than the couch and also ugly if you took it out of a car and into a living room.

    One of the things you’ve said, if I recall, is that the cryptocurrency space is “speed-running 500 years of financial history.” By which I take you to mean that all of the financial disasters of centuries past are playing out in short order, and then they have to rediscover the solutions that were put in place for those things not to happen. So you start off thinking, “Oh, wouldn’t it be fantastic if there were no central authority?” and then all of a sudden you realize, “Actually, it really would be nice if we had a central authority to regulate fraud and such” and you rediscover the virtue of banks and government.

    This sounds a lot snappier if you don't realize that cryptocurrencies aren't introducing central authorities to regulate things. In some cases, cryptocurrencies are finding decentralized ways to deal with issues (multi-sig wallets can allow reversible escrowed transactions and are becoming popular in some areas), and in other cases some people are finding that cryptocurrency is not the tool they wanted.

    A smart contract is not a contract. The theory behind smart contracts is “code is law.” So let’s do programs that cannot be updated that handle money. Now, we’ve had programs that handle money for decades now. So I’m a savvy investor, I have an index fund, my index fund is run by a computer that’s running a fairly simple set of programs, trading on my behalf to make sure it matches the index.

    Smart contracts can be made to be updatable.

    It's great that banks get to write code that deals with money, but can anyone else? Could a group write or use a pre-made program to organize a fund of money they could vote on using? ... Okay, you can probably approximate that with regular contracts, but for what amount of legal fees, and would you still end up in a situation where one person has control over the bank account? How exactly do you go from votes to a transaction coming out of the bank account? What amount of legal fees and court process and time would you have to go through if the person with the bank account emptied it out? Is it more complicated legally if some of the participants are in other states or countries? What would it take to make this process more solid? Could one well-educated lawyer in the group hash that out in an afternoon? As someone who knows smart contracts this is an easy task for smart contracts and I think that's interesting.

    So the first smart contract, the DAO, back in 2016, was “Hey, let’s make a voting distributed mutual fund.” So anybody can invest in the DAO and get a say in how we invest the money. Ten percent of all Ethereum got invested in the DAO. And it basically got invested because it’s got a cool name. And it was basically a self-assembling Ponzi scheme.

    What the hell, are we just calling everything a Ponzi scheme now? There's nothing Ponzi about the DAO. It was a pool of money you could deposit in or out of, and people could vote on what to do with the money in it in order to fund projects.

    He showed the whole stupidity of this place by temporarily putting his profile portrait to a collage of apes he didn’t own.

    The apes are a weird status symbol I don't care much for, but for the people who care about it as a status symbol this was just silly. Imagine bragging you snagged a cool domain name, and then someone smirked at you and obviously falsely announced they actually owned that domain name and every other domain name too. The attention is flattering but it's meaningless.

    So there are a lot of securities regulations out there. And the definition of “security” is very broad. It dates back to the Howey Test in the Great Depression era. That happens to be one of the cleanest legal tests ever for “Is this an investment contract?” and therefore a security that should be regulated by securities regulators. It’s very much “if it walks like a duck and quacks like a duck and swims like a duck and flies like a duck, it’s a duck.” [...]

    I definitely agree there's been a lot of companies abusing ICOs, but I'm hesitant to agree with the idea that the laws were just fine. Companies are waiting longer and longer before going public because of the legal hurdles required, and so investment opportunities are being increasingly limited to "accredited investors" (it's literally a legal category of people rich enough who are then the only ones allowed to invest in things like pre-IPO companies). I wonder if crypto's use as an escape hatch here might help a new balance to be found.

    It will implode spectacularly. The only question is when. I thought it would have actually imploded a year ago. But basically, what we saw with Terra and Luna, where it collapsed suddenly due to these downward positive feedback loops—situations where basically the system is designed to collapse utterly and quickly—those will happen to the larger cryptocurrency space.

    Terra/Luna collapsed for very specific reasons because it was set up unstably that don't apply to others. It's misleading to present its failure as typical of popular cryptocurrencies.

    That drops the price more. They’re forced to sell more. This creates a feedback loop that drives the price into the ground, catastrophically. The previous times this has happened, we had the bubble at 100, powered by fraud at Mt. Gox. And that imploded down to 10. We had a bubble a 1000 powered by fraud, it imploded and went back down to 100. We had a bubble at 10,000 powered by Tether, it blew up and went back down to 1,000. And now we’re at a bubble where Bitcoin blew up to 60,000, fueled by Tether and falling. But I don’t think there’ll be a fifth bubble. Because basically, they will have broken all the suckers left to break.

    Its booms and busts have repeatedly landed it at a higher price, therefore it's eventually going to crash to nothing? I don't want to argue that pattern is going to necessarily continue or that its price isn't wildly unstable, but I feel like the article just puts sentences next to each other as if they logically follow and it's weird.

    10 votes
    1. [4]
      streblo
      Link Parent
      I think the problem with cryptocurrency's approach to decentralization is that it totally misses the problem domain. The problems with a centralized government currency do not rest with its...

      I think the problem with cryptocurrency's approach to decentralization is that it totally misses the problem domain.

      The problems with a centralized government currency do not rest with its centralization. Instead they rest with the lack of transparency, lack of democratic levers of control, and lack of education in citizens to understand those levers. Cryptocurrency wants to throw the current system away and replace it with math, but that brings with it an entirely new problem domain that led to centralization and regulation in the first place.

      17 votes
      1. [3]
        mtset
        Link Parent
        Yes, so much this! Cryptocurrencies are either pure speculation or a way to further centralize power into the hands of those who already have wealth and clout.

        Yes, so much this! Cryptocurrencies are either pure speculation or a way to further centralize power into the hands of those who already have wealth and clout.

        7 votes
        1. [2]
          skybrian
          Link Parent
          I think the situation is more complicated. To add some nuance: As an example, a bunch of Chinese bitcoin miners managed to make a fair bit of money (I presume) before China banned it. It probably...

          I think the situation is more complicated. To add some nuance:

          As an example, a bunch of Chinese bitcoin miners managed to make a fair bit of money (I presume) before China banned it. It probably did require some money and connections to get started, but I doubt they were connected to the West’s rich and powerful. And this is true not just of China but of many other countries.

          This suggests a kind of business with low(ish) barriers to entry, where entrepreneurs in many parts of the world had a chance to make money. It’s always the case that having money and connections makes it easier to start a business, but this isn’t nearly as much of an oligarchy as established industries, often with formal barriers to entry. It’s been relatively easy to break in, for people with access to cheap power anyway.

          Whether this counts as “centralizing” or “decentralizing” depends on your point of view.

          On the other hand, cryptocurrency’s lottery-like aspects are inherently increasing inequality. The gains did often go to people who weren’t previously rich, though. It’s much like a lottery appeals to poor people who otherwise have no chance to become rich. It’s a false hope for most, though.

          So I guess I’d say it’s not as centralizing as some things but more centralizing than others. At least, it means a different set of people became rich. In this way it’s like previous gold rushes.

          2 votes
          1. mtset
            Link Parent
            They themselves were rich, if not powerful, because they had the money to invest in large, unsafe warehouses and millions of dollars of GPUs. No, it doesn't. People who have money can use PoW or...

            It probably did require some money and connections to get started, but I doubt they were connected to the West’s rich and powerful.

            They themselves were rich, if not powerful, because they had the money to invest in large, unsafe warehouses and millions of dollars of GPUs.

            Whether this counts as “centralizing” or “decentralizing” depends on your point of view.

            No, it doesn't. People who have money can use PoW or PoS to get more money. People who don't have much money can use PoW or PoS to get a small amount of money; people who have a lot of money can use PoW or PoS to get a large amount of additional money. Because mining, whether PoW or PoS, is negative sum - that is, it not only produces nothing but consumes a great deal - the people with more power make the money, and the people with less power lose it, on average. That's centralizing.

            8 votes
    2. lou
      Link Parent
      Most of the time, when something like cryptocurrency is critized, one should interpret it as criticism of the way in which it is currently used, not as criticism against the underlying technology...

      Most of the time, when something like cryptocurrency is critized, one should interpret it as criticism of the way in which it is currently used, not as criticism against the underlying technology in itself.

      5 votes
    3. [20]
      mtset
      Link Parent
      So that I can understand where you're coming from: have you been involved with any non-cryptocurrency decentralization and distribution efforts?

      So that I can understand where you're coming from: have you been involved with any non-cryptocurrency decentralization and distribution efforts?

      2 votes
      1. [12]
        Macil
        (edited )
        Link Parent
        I've read into and actively experimented with torrents, IPFS, BOINC, Tahoe LAFS, Tor, GNUNet, GPG, age, Mastodon, Matrix, Pidgin, CJDNS, Yggdrasil, Linux, LetsEncrypt, running home web and email...

        I've read into and actively experimented with torrents, IPFS, BOINC, Tahoe LAFS, Tor, GNUNet, GPG, age, Mastodon, Matrix, Pidgin, CJDNS, Yggdrasil, Linux, LetsEncrypt, running home web and email servers, and probably a bunch of other things vaguely in that category. I've always thought it was important to have software and distribution systems that individuals who put the effort into understanding could freely use and repurpose themselves instead of being dependent on the whims of businesses or admins. I wonder how much of that came from it being the fashion of Slashdot at the time I used it in my formative years, or how much of that came from me being really into original Xbox modding and being frustrated at how places hosting modding tools based on modified Xbox software would get struck from the internet through copyright claims.

        I had read some cypherpunk stuff mainly when learning about Tor and GPG, and read the original Bitcoin paper as soon as it came out. I remember thinking it was a brilliant solution to the problem of decentralized payments, accomplishing real decentralization in something I had fully believed wasn't possible, but it was a shame it would probably be exactly as popular as Tor and GPG. Didn't occur to me that people would speculate on the price of its unit. Wrote some toy code for doing things with it, mined and day-traded some for a while, sold out of it when I got bored and incorrectly figured its price wasn't going anywhere. Read a lot of dev theorizing in the cryptocurrency space because I was very interested in how various limitations in the system could be lifted and about what else Bitcoin's design could enable. Mostly ignored the general cryptocurrency community because it was like a 1:10:100 ratio between decentralization fans like me, weird gold-bug libertarians who thought Bitcoin would obsolete the dollar/taxes/governments, and people who thought it was purely a get-rich-quick scheme they could get rich from.

        I feel like there's an attitude around open source, decentralization, and making new systems under user control that's not as popular online nowadays. When there's some controversy around a social media company like Twitter, I feel like there's almost purely a reactive attitude of "what are we going to have to put up with next from our big tech overlords" instead of much push toward alternative ideas like Mastodon. Is it just that people mostly ran out of enthusiasm for more decentralized alternatives after seeing most alternatives like that putter out, or have shittier over-advertised vaporware scam cryptocurrency projects poisoned the word "decentralized" through overuse and misuse, or is it a new generation of internet users that don't even have a concept of software that's controlled by users instead of "big tech"? It's strange to me to think that to many internet users today, the internet is exclusively Google, Facebook, Twitter, etc and what they want, instead of it being every enthusiastic amateur programmer making whatever they want.

        10 votes
        1. [8]
          mtset
          Link Parent
          Interesting! I'm really curious, given all this experience, why you're interested in projects that explicitly push towards global distributed consensus rather than P2P or local consensus or...

          Interesting! I'm really curious, given all this experience, why you're interested in projects that explicitly push towards global distributed consensus rather than P2P or local consensus or eventual consistency?

          I remember thinking it was a brilliant solution to the problem of decentralized payments, accomplishing real decentralization in something I had fully believed wasn't possible, but it was a shame it would probably be exactly as popular as Tor and GPG. Didn't occur to me that people would speculate on the price of its unit. Wrote some toy code for doing things with it, mined and day-traded some for a while, sold out of it when I got bored and incorrectly figured its price wasn't going anywhere.

          Yeah, me too. I got burned by Butterfly Labs and everything. It's interesting to me to see so many people who saw all the flaws of Bitcoin in the 2012-2016 era still sticking to their cryptocurrency guns today.

          I feel like there's an attitude around open source, decentralization, and making new systems under user control that's not as popular online nowadays. When there's some controversy around a social media company like Twitter, I feel like there's almost purely a reactive attitude of "what are we going to have to put up with next from our big tech overlords" instead of much push toward alternative ideas like Mastodon.

          That's technocracy, baby! I'm pretty frustrated by this too - though I'm working on it, stand by for that announcement :) - but I guess I don't see how cryptocurrency is going to help us there, since it inherently gives more power to the largest and wealthiest entities in the world.

          3 votes
          1. [7]
            Macil
            Link Parent
            There's certain things like payments, domain names, and permanent storage of small data that definitely benefit from global consensus in a blockchain. But I think a lot about personal clusters of...

            Interesting! I'm really curious, given all this experience, why you're interested in projects that explicitly push towards global distributed consensus rather than P2P or local consensus or eventual consistency?

            There's certain things like payments, domain names, and permanent storage of small data that definitely benefit from global consensus in a blockchain. But I think a lot about personal clusters of a user's own devices, in which case synchronizing personal data directly between them is useful. It's ridiculous that the most convenient thing to back up my photos from my phone is to have them be uploaded in the clear to Google; it should just happen that they get synced to a configurable number of my personal devices and optionally to a cloud VPS or object store of mine, fully encrypted. All my personal files should be like this. I should be able to wipe any one of my devices and not have lost a thing when I connect back to my personal cluster.

            I should be able to share one of my files from my cluster with someone as easily as sharing a link, and I should be able to effortlessly give them some storage and compute power in my cluster without worrying about exposing anything of mine or interrupting my own use of it. I should be able to set my cluster to automatically accept micropayments for offering spare storage or compute to anyone online. Many people offering their spare resources means research computation that doesn't need datacenter-tier connectivity can be done for dirt cheap. Accepting these payments shouldn't be an ordeal requiring an account with a specific payment company that only works in countries where that company operates; trading resources like this should be a standard element of computers that can be used in many open protocols. This being limited to centralized services should be seen absurd the same way software crippled with "export grade" cryptography was.

            There are businesses that make money from operating cryptocurrency-connected services, but that goes for regular finance too while cryptocurrency instead gets openness and privacy benefits (obviously bitcoin is limited in that part but others do it well). Proof of work has a strongly centralizing effect because of the monopoly that forms on mining hardware, but proof of stake does not have that barrier to entry, allowing more people to participate on level ground. Sometimes PoS is criticized as a rich-get-richer system, but the more people participating in it lowers the profitability, so its profitability should fall until it's on par with commonly available investment opportunities.

            I don't think of cryptocurrency as the end-all panacea of decentralization or a necessary ingredient in everything decentralized. I think the popularity of the cryptocurrency industry has distorted the field of decentralized projects a lot. But I do think monetary sustainability of projects is a big concern -- the classic web approach to avoiding money as much as possible means good contributors and hosts don't stick around, except for ones that find alternate revenue streams like ads, which is what got us into "big tech" centralization to begin with -- and specifically where payment is applicable, decentralized payment systems are a good fit for decentralized projects.

            4 votes
            1. [2]
              mtset
              Link Parent
              We have this! Syncthing is one of the best pieces of open-source peer to peer software ever written, second maybe to qBittorrent. Should, perhaps, but does not in practice. I was an early Peercoin...

              It's ridiculous that the most convenient thing to back up my photos from my phone is to have them be uploaded in the clear to Google; it should just happen that they get synced to a configurable number of my personal devices and optionally to a cloud VPS or object store of mine, fully encrypted.

              We have this! Syncthing is one of the best pieces of open-source peer to peer software ever written, second maybe to qBittorrent.

              Sometimes PoS is criticized as a rich-get-richer system, but the more people participating in it lowers the profitability, so its profitability should fall until it's on par with commonly available investment opportunities.

              Should, perhaps, but does not in practice. I was an early Peercoin investor - as you may recall, PPC died because it became hypercentralized.

              I think the popularity of the cryptocurrency industry has distorted the field of decentralized projects a lot. But I do think monetary sustainability of projects is a big concern -- the classic web approach to avoiding money as much as possible means good contributors and hosts don't stick around, except for ones that find alternate revenue streams like ads, which is what got us into "big tech" centralization to begin with -- and specifically where payment is applicable, decentralized payment systems are a good fit for decentralized projects.

              This is the best pitch for cryptocurrency I've ever heard; "what we want is decentralized infrastructure, and to do that we need decentralized payments." I don't agree that decentralized payments need blockchain, though; a standardized and fast protocol by which banks can send micropayments to each other, which we actually have in many countries, would be enough, without the massive, absolutely unforgivable environmental cost of proof of work.

              5 votes
              1. Macil
                (edited )
                Link Parent
                Oh yeah I definitely like Syncthing and recommend it. It's been my starting point for imagining what a personal cluster should look like. It's still pretty limited compared to what I'd like, still...

                Oh yeah I definitely like Syncthing and recommend it. It's been my starting point for imagining what a personal cluster should look like. It's still pretty limited compared to what I'd like, still too much manual control over what files go on what hosts. There's no way to do something like making sure files in a folder are present on any 2 hosts; instead you pick which directories with everything in them go on what hosts. I'm glad it recently got support for keeping files on encrypted on specific hosts which I was very eager for, though it doesn't support plain external file hosts like S3.

                I didn't follow Peercoin closely enough, but my impression was that it or early Bitcoin-alternatives failed to get enough attention, or that PoS wasn't trusted as much back then (possibly because PPC's early attempt was flawed? I remember reading about the nothing-at-stake issue and multiple PoS systems' increased reliance on checkpointing that newer PoS schemes avoided), or that staking was less normalized then, which are issues I don't expect to affect all newer popular PoS systems including Ethereum's PoS chain. The developers in the Ethereum ecosystem have done a lot of effort to ensure that there are multiple popular fully interoperable implementations of the PoS client, which is a step further than I previously considered necessary for keeping decentralization, so it makes me expect that the Ethereum community would be able to prioritize solving other kinds of PoS centralization if they came up.

                I wouldn't call a payment system decentralized if arbitrary new users can't act as banks in the system, if it doesn't work in some countries, or if authorities could ban some users from it. I think the term distributed or federated would be better there. A more standardized payment protocol that people could make direct use of would still be great of course, probably replacing some cryptocurrency uses. Smart contracts would still be exciting even on a less-decentralized system.

                2 votes
            2. [3]
              skybrian
              Link Parent
              Nitpick: “not end-to-end encrypted” isn’t the same as “uploaded in the clear.” Most traffic is encrypted with TLS these days, and that’s a significant advance over the early Internet.

              Nitpick: “not end-to-end encrypted” isn’t the same as “uploaded in the clear.” Most traffic is encrypted with TLS these days, and that’s a significant advance over the early Internet.

              2 votes
              1. [2]
                Macil
                Link Parent
                Yeah I debated in my head how to word that, and had settled it by writing it from the perspective that an external file host might be considered just as untrusted as the rest of the internet. It's...

                Yeah I debated in my head how to word that, and had settled it by writing it from the perspective that an external file host might be considered just as untrusted as the rest of the internet. It's definitely good though that TLS is so universal now.

                It's actually interesting to think about how much the web has improved over time around HTTPS. I used to use the HTTPS Everywhere extension purely just to enable HTTPS on sites that offered it but didn't default to it, but they just recently announced that they're retiring the extension partly because it's so unnecessary to do now because most sites handle that correctly by always enabling HTTPS. It's so good that this kind of security is just standard now instead of being something only tech-enthusiast users in the know could opt into.

                3 votes
                1. Eric_the_Cerise
                  Link Parent
                  Quick shout out to EFF's certbot and the LetsEncrypt CA, for making this conversion dog-simple. Before those tools, purchasing and applying SSL/TLS certs to websites was both expensive and a big...

                  It's actually interesting to think about how much the web has improved over time around HTTPS

                  Quick shout out to EFF's certbot and the LetsEncrypt CA, for making this conversion dog-simple. Before those tools, purchasing and applying SSL/TLS certs to websites was both expensive and a big hairy PITA.

                  3 votes
            3. Eric_the_Cerise
              Link Parent
              To a very large degree, you are describing Nextcloud here. I've been using it for years, doing exactly all of these things. No micropayment solution that I'm aware of (though I never looked for...

              It's ridiculous that the most convenient thing to back up my photos from my phone is to have them be uploaded in the clear to Google; it should just happen that they get synced to a configurable number of my personal devices and optionally to a cloud VPS or object store of mine, fully encrypted. All my personal files should be like this. I should be able to wipe any one of my devices and not have lost a thing when I connect back to my personal cluster.

              I should be able to share one of my files from my cluster with someone as easily as sharing a link, and I should be able to effortlessly give them some storage and compute power in my cluster without worrying about exposing anything of mine or interrupting my own use of it.

              To a very large degree, you are describing Nextcloud here. I've been using it for years, doing exactly all of these things. No micropayment solution that I'm aware of (though I never looked for one, either), but apart from that ...

              ETA: Just noticed the Syncthing recommendation below, and seconding it. It's different from Nextcloud, better/worse depending on your use case.

              1 vote
        2. [3]
          wcerfgba
          Link Parent
          I've heard of all of these and played with most of them except 'age'. Please could you tell us what this is or drop a link?

          I've read into and actively experimented with torrents, IPFS, BOINC, Tahoe LAFS, Tor, GNUNet, GPG, age, Mastodon, Matrix, Pidgin, CJDNS, Yggdrasil, Linux, LetsEncrypt, running home web and email servers, and probably a bunch of other things vaguely in that category.

          I've heard of all of these and played with most of them except 'age'. Please could you tell us what this is or drop a link?

          1 vote
          1. Macil
            (edited )
            Link Parent
            That's the recent age encryption tool. I felt obligated to mention it since I mentioned GPG. It's a modern and minimal GPG alternative for encryption but not signing use-cases. I've been working...

            That's the recent age encryption tool. I felt obligated to mention it since I mentioned GPG. It's a modern and minimal GPG alternative for encryption but not signing use-cases. I've been working recently on a small project of a reverse HTTP proxy that transparently decrypts age-encrypted files. It's mainly intended for making files usable while they're kept encrypted stored on public/untrusted file stores like IPFS or S3.

            2 votes
      2. [7]
        FlippantGod
        Link Parent
        Mostly similarly, I am very interested in the blockchain, and do retain a little interest in cryptocurrencies, for more or less Macil's reasons. I have history with irl mesh and p2p networks,...

        Mostly similarly, I am very interested in the blockchain, and do retain a little interest in cryptocurrencies, for more or less Macil's reasons.

        I have history with irl mesh and p2p networks, decentralized micro power and water treatment in third world countries, as well as various local nontechnical decentralization and resilience projects.

        3 votes
        1. [6]
          mtset
          Link Parent
          Interesting! So, I'm curious then as to what benefit you see from the global-distributed-consensus model over pure P2P and local-consensus or eventually-consistent models?

          Interesting! So, I'm curious then as to what benefit you see from the global-distributed-consensus model over pure P2P and local-consensus or eventually-consistent models?

          3 votes
          1. [5]
            FlippantGod
            Link Parent
            I am particularly interested in the Ethereum Name Service, which you might note is also leveraging IPFS. DAOs are also conceptually interesting and I am not aware of anything else quite like it....

            I am particularly interested in the Ethereum Name Service, which you might note is also leveraging IPFS.

            DAOs are also conceptually interesting and I am not aware of anything else quite like it. Though frankly, for collective fund management, open collective seems to be a functional solution.

            1. [4]
              mtset
              Link Parent
              Unfortunately, the people who run ENS are fundamentally bad people: Yeah, I'm not sure what a DAO really does that you can't do with any other democratic organization, like a co-op.

              Unfortunately, the people who run ENS are fundamentally bad people:

              The latest example of the trend involves Ethereum Name Service (ENS), the blockchain-based domain name protocol that sells .eth names. Last week, an old tweet resurfaced from ENS Director of Operations Brantly Millegan in which he declared: “Homosexual acts are evil. Transgenderism doesn’t exist. Abortion is murder. Contraception is a perversion. So is masturbation and porn.”

              DAOs are also conceptually interesting and I am not aware of anything else quite like it. Though frankly, for collective fund management, open collective seems to be a functional solution.

              Yeah, I'm not sure what a DAO really does that you can't do with any other democratic organization, like a co-op.

              6 votes
              1. FlippantGod
                Link Parent
                Aww for f*cks sake, this is why we can't have nice things; we have too many bad people. Yes, conceptually the DAO makes strong guarantees (due to autonomous operation). In practice... well,...

                Aww for f*cks sake, this is why we can't have nice things; we have too many bad people.

                Yes, conceptually the DAO makes strong guarantees (due to autonomous operation). In practice... well, sometimes it pays to be a pragmatist. Perhaps literally.

                3 votes
              2. [2]
                FlippantGod
                Link Parent
                Edit: an advantage over co-ops is that they are subject to state and federal regulations in the US. I would imagine trying to apply for a coop with people around the world to formally jointly own...

                Edit: an advantage over co-ops is that they are subject to state and federal regulations in the US. I would imagine trying to apply for a coop with people around the world to formally jointly own and operate a community web services would be challenging. This doesn't mean a DAO is actually the best solution.

                2 votes
                1. mtset
                  Link Parent
                  Agreed! I do think that it's only a matter of time, however, until we see that DAOs are subject to those regulations; it's just that right now nobody's enforcing them.

                  Agreed! I do think that it's only a matter of time, however, until we see that DAOs are subject to those regulations; it's just that right now nobody's enforcing them.

                  4 votes
  3. stu2b50
    Link
    Wow, it's a small world - I took my computer architecture class with him a handful of years ago! I distinctly remember a project explicitly labeled as a "kobayashi maru"... I remember getting that...

    Wow, it's a small world - I took my computer architecture class with him a handful of years ago! I distinctly remember a project explicitly labeled as a "kobayashi maru"...

    I remember getting that bitcoin lecture in person, in fact. There's was a big blockchain club on campus, so I do think he enjoyed getting to force all of them to listen to his thoughts since that class was mandatory.

    At this point I think my thoughts on blockchain as a technology (the merkle tree + consensus mechanism) is that it indeed is a clever way to prevent the double spend problem in a decentralized system as per the original Satoshi paper... but that it's a very expensive way to do so. Not in raw cost (electricity, for example) per se, but in complexity - it becomes so unwieldy to work with, and imposes such significant technical limitations, and there's still so many things to build because double spend is just one issue that a transaction system needs to solve, that I find it hard to believe the tradeoffs are worth it.

    7 votes