What do you think would be an actually good use of blockchain/smart contracts? What kind of problems (big or small) is it a good tool for?
Hello!
When I was creating a CTF for a conference, I've finally got to learn about how blockchain and smart contracts actually works in practice, and the whole concept is simply brilliant. A quick introduction for those unfamiliar with it would be in this summary, but just to summarize how I basically understand it, blockchain is simply a VM that runs code (smart contracts) a both the code, and result of every execution of it is calculated by a bunch of users (so, mining is basically running a VM) and appended into the blockchain based on some kind of consensus and proof of work. This means that you get a single source of truth and history of every execution of a smart contract that is decentralized and you can rely on it.
But, almost every use of blockchain or smart contracts I have seen has pretty large issues either in sustainability in the long term, or in cases where you simply need some form of an authority to prevent and punish misuse. While I'm not really that much familiar with every use of blockchain so far, I will first list what I've already thought about or seen, and the main issues that I think are a deal-breaker for choosing blockchain for that kind of tasks. It's possible that some of the issues are wrong or have already been solved, so please correct me if I'm wrong - my knowledge of blockchain isn't really that in-depth.
First and the most common use is the one you are probably most aware of - cryptocurrencies. If I ignore the biggest and most unfortunate issue of cryptocurrencies turning into an investment-only product, with hugely volatile and inflated price that is not backed by any kind of real value (sure, you can pay with BTC, but it's slow, expensive and super volatile to be useful, so the only real use is to literally sell it to others for a profit - which also basically means you are scamming someone out of their money down the line), I see the following problems with using blockchain for currencies:
Longevity - The ledger size is already getting massive, only after a few year. It's not sustainable, and it will eventually be really hard to keep the whole ledger at a large enough number of places to not run into problems of integrity. It's growing exponentionally, and is at around 500Gb after around 10 years.
Gas cost - It's getting harder and harder to mine and confirm new transactions, which increases the cost while also making less people able to mine new transactions without being at a loss. This will only get worse, and eventually lead to the 50% problem (if someone controls 50%+ of mining nodes, he can confirm fake transactions or do whatever he wants with the blockchain) being a real issue.
Lack of moderation - This may be one of the more controversial issues, because it goes directly against the whole idea of cryptocurrencies, but is one of the biggest problems I see that are in the way of crypto being able to be considered for wider use. We live in a world where some people are dicks that are not afraid to steal and cheat, and something like a currency simply has to be moderatable. You need to be able to punish criminals, and take back what they have stolen. If someone doesn't pay their debts and owns me money, the government should be able to just take the money if they have them. If someone uses an account for scamming and stealing, it should be possible to freeze it.
The last issue will eventually show in most of the other uses of blockchain as well, and while I have included it, I'm still not sure how I feel bout it. In an ideal world, you would not have to deal with something like this. I would also really like to have an option to do my transactions privately, without anyone being able to profile my behavior and data, but such a system would have to allow for some safeguards against missuse to be widely adoptable. (Which is an interresting off-topic question - would it be possible to create a system that is private, but also has the possibility for trusted authorities to freeze accounts and force transactions?) And the more that I think about it, the more I'm certain that I'd rather have a centralized system where you can punish criminals and scammers, than a system where lives of people are regularly ruined by someone stealing all of their savings unpunished. But it is a thin line - I only say that because I live in a country that is all-right and I can trust my government - for now. But I definitely agree that such a private unmoderated option should exist - but can't be considered for widespread use, which I've heard some people say that "crypto will replace cash in a few years". And this is why it never will, IMO. But this discussion shouldn't be about whether this is a good opinion or not - but more about "what blockchain is a good tool for".
Next one are NFTs. I will just quickly gloss over them, because they are even bigger scam than crypto is. Ever heard someone say "Someone has copied and minted my NFT?". Well, it's a shame that there isn't some kind of centralized authority that could, you know, not allow them to do that.
Another use I've heard someone praise as "the future" was lending money. I'm not sure what were they talking about, but the whole point was that you can... Escrow an amount you are borrowing, and then borrow the same amount? It didn't make any sense, so I guess I'm missing something, but then again - we have the same issues as above, while also it being just a bizare idea - why simply not use the amount you already have? The person tried to explain it to me, but it just feels gimmicky. And if you escrow a lesser amount, you then have the same problem with moderation as above - nothing can force you to return the money (unless it is already escrowed, but then, why??)
So far, every use of blockchain I have heard about would be better done in a centralized fashion, especially as far as longevity is concerned. The growing ledger size and increasing gas cost, along with the 50% problem simply makes most of these kind of uses too impractical to work on a larger scale.
But I really like the concept and idea of smart contracts, and I'm sure there has to be some kind of use that is not as "revolutionary" or large scale. I'm just having hard time coming up with any.
I have only one - voting, and maybe transparent randomization (i.e lottery). Smart contracts are an amazing way to collect votes transparently but privately, since you can be sure that no-one can cheat, if you set it up properly. It's also something that doesn't suffer from the longevity problem, because it's more of a one-shot use of blockchain, rather than something ongoing - which also justifies the price.
(tl;dr feel free to start here:)
Which is what I'm interested in - does any of you have similar ideas for use of smart contracts and blockchain, that would be practical in a daily live? Be it one-shot smart contracts for a small task, such as voting or random winner selection, maybe some kind of escrow. It doesn't have to be a "society changing system", or something revolutionary. A common small code snippets or apps that would solve the trust issue inherent to a centralized task is what I'm after - but have hard time coming up with.
And just a disclaimer - I don't plan on building anything and am not fishing for the next blockchain thing, I barely even understand it. I would just like to incorporate blockchain into my programming repertoire as a tool, because the concept feels so clever, but is also misused or misunderstood due to hype, but it has to have it's uses that are overshadowed by people jumping on the blockchain bandwagon without considering whether it's really the best tool for the job.
But is has to be a good tool for some kind of problems, right? And I would like to start a discussion about what would that be, without it being affected by the hype and reputation surrounding blockchain. I feel like that would be an interesting though exercise, and I'm sure we can come up with some interesting little uses here and there, without it being gimmicky but actually the best tool for the job.
Thank you!
EDIT: And I'd like to add that I never got into the blockchain hype, and my opinion on how it's used so far is mostly negative. If a product mentions blockchain, I usually just avoid it as a gimmick. But that's why I'm genuinely interested in this discussion - I don't judge a tool about how people misuse it.
Blockchain and/or smart contracts try to solve problems that were already solved in multiple ways by adding a ton of overhead that makes them unable for large scale deployment and long term usage.
Here's what's stupid about the people who say that blockchain will revolutionize the financial sector: why add a blockchain and all the computing power to store transactions when you can take the obviously efficient route and simply store transactions on a SQL database? Before anyone screams the word "decentralization" do you really think banks will cease to exist? NO. The most likely scenario - if people keep pushing this bullshit - will be to have some kind of closed blockchain that banks use to transact money, so it essentially becomes the same thing we've now with added overhead, environmental impact and technical complexity. We have efficient system in place with safeguards, operations can be tracked, reversed etc.
Frankly it would be a better use of everyone's time, money and effort to simply fix the REAL problems in the banking industry, such as the fact that the US still doesn't have a decently working, standardized digital system to transfer money between account holders in different banks. Europe has this with SWIFT/IBAN and people can transfer money between accounts, banks and countries almost instantly by just providing the amount they want to transfer and an IBAN number (nothing else required). Now tell me, how many people in the US have bank account with IBAN numbers? Most likely only millionaires. The majority of people use a combination of poorly structured system of account and routing numbers that often fail and lead to delays. Oh btw Russia has a similar system to IBAN.
There are tons of other weaknesses in the US banking system around the way credit and debit cards work, for instance why would anyone on their right mind assume that a system where you can provide your credit cards number and CVV/CVC code over a phone to make a transfer wouln't be abused to scam people and steal money? Then, after decades of fraud, to "alleviate" the issue they decided to create a bunch of companies that offer virtual credit cards with limits. Now let this how with works in most European countries: banks will, most likely, refuse any attempt at charging a physical credit card unless its made on a physical payment terminal with the card actually physical inserted on the thing an a 4 digit PIN code typed in. If you want to buy shit over the internet simply open your bank's app or website and they'll have a function to create a single use virtual credit card for the transaction. Way more secure isn't it? :) Either way most European countries also other systems to handle those kinds of payments eg. the online stores provides you with a specific code and you then can go into any ATM or your Bank's App, insert the code and make the payment.
As you can see making the banking system efficient and having fast, secure and usable things isn't about blockchain bullshit, its usually more about common sense and creating standards that companies, such as banks, have to comply with.
You don't need many of the features you listed to make a blockchain. Its basic form is a Merkle tree. For which there are many practical uses, some of which predate Bitcoin, and which you're probably familiar with.
Git is a blockchain. It's one of the most important tools for free and open software, which in turn powers huge parts of internet and technology.
Most of the extra stuff (proof of work etc ) were added specifically for crypto.
Many people don't consider Git to be a blockchain because it lacks more than just proof of work. No argument against merkle tree, but a blockchain is more than a merkle tree, otherwise it would just be called a merkle tree. Part of the issue is I don't think there's any real accepted definition for just a simple blockchain.
For example, you can absolutely undo a commit (it's messy, yes). This is counter to how blockchain operates. Each commit does not rewrite every previous commit.
So git is sort of a precursor to blockchain. Distributed ledger, sure. Blockchain, no.
A blockchain is extremely difficult (virtually impossible) to intentionally tamper with. Git is not. Well, relatively speaking.
I haven’t heard anyone ever refer to git as a blockchain. The main point of “blockchains” is to have a trustless security mechanism, which git doesn’t have. I don’t think blockchain and merkle trees are the same thing at all, even if blockchain uses merkle trees under the hood.
This right here is really the spirit of the post. Yes there's many impractical applications. Much like there are many impractical applications for RDBMSs, but the tech has such a stank on it, it's important to remember it's just a tool that can be useful despite the hype cycle.
There is actually a system in the works called FedNow that banks here can sign up to be a part of to allow national money transfer between any two people. Probably a lot of banks aren't taking part yet since it's barely a couple weeks old, but it's promising.
Yes it is, and I hope it actually fixes the issue. But I'm not sure how it works in detail and it doesn't seem to actually replace the mess that banks made but instead just add a bandage. At least is seem to use ISO 200022 message standards but as I said would it be THAT hard to simply ditch all the crap and model a baking system around what the EU/Russia does? Actually I don't get why the US can't just adopt the internationally used SWIFT system and the IBAN improvements that is used by 77 countries world wide for their national and international transfers. This system won't make the US dependent on anyone as it is mostly a bunch of technical standards and recommendations on how to transfer money between banks.
I live in Europe and have some direct experience with how the banking system works (I was pentesting the system that shares transaction data between banks over their closed intranet), and I had no idea that US doesn't have something like that. That's interesting, that sounds like a lot of inconveniences.
A closed blockchain doesn't need or use expensive hashing nor is expensive hashing required for a public blockchain.
Every rant about how blockchains are bad SQL databases is ignorant of the actual, novel uses of a decentralized blockchain and whatever system it uses as proof to find the current block's validator.
Blockchains allow for the synchronization of many actors up to the agreement of the majority of actors. They eliminate a class of corruption. Which is at least those where the authority over a log of data uses that authority to alter the log outside the will of the majority of those who use said log.
You can make all the arguments you want about the usefulness of that ability, you can make arguments that the cost of adding that ability being too high vs the reward.
You vague argued that its complicated and it wont end banks so its pointless. Neither of which is much of an argument. Also, the blockchain replaces central banks, not member banks.
Is that worth doing? Likely not at the current state of the technology.
Blockchains are only useful in cases where non-repudiability (the ability to prevent users from denying that an event happened) is more important than any other factor. And there are preciously few cases where this is the case, the vast majority being related to audit - tracking receipts, votes, certificates, or similar attestations in an environment where no single party can be trusted. Disclaimer, I've worked in the past in projects related to the aforementioned - fortunately all of them related to the field of audit.
The only useful use case I've seen is for when you absolutely MUST be able to track historic data and ensure edits don't destroy the original. Blockchain "solves" this by never allowing saved data to be edited.
The only place I've seen it actually being used properly for that was within Brazil's medical vaccine tracking (ptbr article), which is what allowed them to confirm that Bolsonaro falsified his vaccination card. It doesn't offer details on what kind of protocol it uses, but it could just be a "decentralized, distributed" database, for all intents and purposes.
We’ve had that for literally 40+ years though with CQRS, which is used in every banking system (it’s why your bank can show you every transaction that has happened on your account). It’s also used by airports and airlines. It’s incredibly common tech that doesn’t require a blockchain or decentralization.
I think you are saying the same thing, but it's less that history can't be changed and more than all changes are collectively acknowledged. The Brazilian vaccine record shows why that can be a highly desirable attribute; it prevents a class of corruption while also automating the tasks that could have been corrupted.
These three requirements are tricky and often conflicting.
Block-chain might be an inefficient way to achieve these, but the glove does fit quite neatly.
Logistical paperwork
Purchase Orders/Invoices and packing slips
Waybills/Bills of lading and CMR's
These kinds of documents require multiple stages of matching and approval by untrusted 3rd parties.
There are dozens of ecosystems of interacting systems that support processing these documents, but most people still use paper. Paper is more reliable when you need to deliver a container full of diapers from Poland to North Sudan. It's more reliable but incredibly prone to fraud and forgery.
Having all of these approvals and transactions tracked on a blockchain and letting different systems interact with the same chain, would make it possible without each ERP having a rest API to each other ERP.
Yeah the problem isn't the veracity of the logs, it's providing a mechanism for third parties of proving that the sequence of events in your log hasn't been tampered with after the fact
It's more about tampering with the audit log. "Company A provided their audit logs to prove their innocence." Did they? Well. Maybe. How do we know it's the full log. How do we know it wasn't altered? Sure, the company can digitally sign it, but what does that prove?
I've given it some though and wouldn't the fact that the blockchain is public by design be a problem in regard to forward secrecy (I'm not sure I'm using the term right here, but I suppose you get the idea)? If your keys would leak, you are then stuck with a lot of private data leaked without any way how to pull them back.
Not every log needs that kind of security and a chain does not need to be public. You download blocks from peers and do your own accounting.
Nothing is preventing you from only giving access to your chain to a trusted circle of peers.
Something you could do is encrypt your logs and push them to a chain shared by a number of peers who do they same with their own keys. Now you have a pool of accountability buddies, because if someone tries to tamper with the logs, you all hang together.
If you're doing some spooky stuff and need to prove a high degree of integrity is you could push encrypted logs to a chain. The auditor then can appoint several independent parties whose only job it is to continuously prove the integrity of your logs. After that is proven you can release your keys to the auditor who can inspect your logs knowing that they have been complete and untampered during the audit period.
Again I understand it's not the most efficient system, but there are less efficient and less flexible systems out there in enterprise land haha
The audit logging sounds interesting. If you combine it with some kind of encryption, then I can imagine it working pretty well. Aside from the logistical problems/gas cost, that is.
There is no incentive for adding the friction of gas or PoW for these types of systems.
The parties involved can have a shared log and private keys for signing entries. Party A provides a thing and Party B signs an entry that says they were provided with the thing. Party A can wait for that signed entry before releasing the goods, etc. The problem with block chain to track physical stuff is that that handoffs are not instantaneous, so there’s always lag between the real state of the world and what the log says. In practice, this may be a few seconds, and a human might wait for confirmation before physically granting access to a recipient.
To put it another way, the party that is signing is not incentivized to forge that they have received an object from someone else, as that is effectively the fulfillment of the obligation. They’re only going to sign an entry if they get the object.
Yeah it's not ideal, but you only need to pay the gas cost when you need to prove integrity and that's alot cheaper than having to constantly be in sync with the world.
Git is Blockchain and it's pretty much the only use of the tech I actually see make sense. Most other uses add too much expense where we could just used a trusted ledger. I know people are all about zero trust but the cost benefit of Blockchain doesn't pan out for almost anything. It's not hard to develop cheaper ways to trust an actor, such as laws. Which is how we create trust today. When's the last time your westernized bank stole money from you?
Git is not blockchain. Git is a distributed ledger. You can rewind a git commit if you know what you're doing. If fit were a blockchain, removing the last commit would corrupt the entire chain. Every transaction is part of building that trust.
Your conclusion of uses of blockchain is spot on though. I also feel it's extremely expensive and complicated when there are much cheaper and more efficient ways outside of it.
When did this become a thing where people started calling git a blockchain? Git is much older than blockchain. I don't see anyone giving Torvalds credit for inventing blockchain. Because he didn't. And git isn't blockchain. It's sort of useful as an illustration of how blockchain almost works, except it's much more efficient because it doesn't do all the things that blockchain requires.
Well, if you remove a commit in the chain it will disrupt the chain. It's a DAG where the next commit requires the previous commit to be unchanged. So Git does use essentially a cryptographic chain.
The thing is we don't care about guessing the next part of the chain, which is where the mining aspect of Blockchain comes into play. So Git does not include validation in it's process. But if someone does modify something in Git I assure you all of the people who have cloned that repository are able to validate a change happened.
Nobody credits Linus with it because signing things was never a novel idea. The part where you're mining the next hash is the interesting part of Blockchain that's not in Git. And that's also the wasteful silly part.
The paradox is that it must solve money first. The original intended usecase is not only the best, but it is a prerequisite to higher order use cases.
Think of the money supply layers, “M0, M1, M2” and so on, and what they mean.
Smart contracts are an “M1-M2+” problem solver. But M0 still isn’t solved.
In other words, having stock markets or NFT housing chain of title will always fail if the money usecase upon which it stands is still fallible. Ethereum put the cart before the horse. Bitcoin or Bitcoin Cash, “dumb” M0 solutions, have to be widely useful for the rest to work.
When NFTs were invented, people imagined them being used for things like titles/deeds. Instant transfer and verifiability would be a huge thing. especially in places that have real estate scams due to the slow/corrupt bureaucracy.
A good while back I read a paper, blog post....I read something somewhere a while back that laid out an interesting use case involving vehicular service records for fleet vehicles. And I know exactly about as much about blockchain then, as I do now, but I did spend some time in fleet logistics for a large scale service company with about 20+ field vans and at the time, the notion seemed compelling and interesting on the face of it.
After a very brief google, it seems the topic has been widely written about but nothing in depth compared to the piece I read all those years ago (which felt more like a full on white-paper). Looking around and will edit the comment if I find it so the people in the room who are smarter than I am can weigh in.
I don't know much about the topic in depth, but I can tell you the greatest problem with using a blockchain for such record keeping: there is no way to ensure that the service that was recorded in the blockchain actually matches the service that was performed. And this is the same problem that every single record keeping system has, so it's not unique, but simply because of this all the greater reliability of the blockchain is meaningless.
Its backing store is an (immutable) merkle tree, which is a chain of crypographically signed object (commits, trees and blob), aka a chain of block, aka a blockchain.
If you have a trusted Oracle on whatever chain you're using, then your smart contracts can start working with real world data. And that opens up lots of possibilities. Of course it puts a lot of faith in oracles. But for instance you could have an inheritance scheme that triggers on the death of a relative. You could do life insurance. You could affect any sort of contract was transaction money or currency.
To your points about central control, I think it's anthema to the idea of a distributed system. If your resistant to malicious actors you have to be resistant to censorship. Because it's only a matter of time until the central authority becomes a malicious actor. At least for some of the population.
Trust is a fundamental thing in human life an society. It get around the abuses by having checks designed into the system and smart ways to stop corruption. If you can't trust a company to deliver a service after payment why would you even work with them? This is the "issue" that smart contracts allegedly fix.
You can make the same argument against escrow services. But they're absolutely fundamental to housing transactions. Where the risk of a downside is unacceptable.
Blockchain gives a protocol that allows new data sources to join and leave without any coordination while also ensuring data consistency in the data added up to the will of the majority of validators. If, for example, the validators are people, the data sources are doctors and hospitals, and the data is medical history. You can have all your medical history collected together and only accessible to you.
Land ownership (in the US) already is public. What problem would this solve? What problems would it create? How would the solutions outweigh the problems?
There's only a few recorders who actually check title, called Torrens, when you record something in the US. 99% of them work under abstract where literally anyone can record anything as long as they pay the county recorder and meet the basic requirements like have a notary stamp.
There's a good chance if you're in the US that I could just record a deed giving me ownership of your house or apartment complex. I'd have to fraudulently sign your name as grantor, but the county isn't going to stop me. You'll have to stop me.
There's a whole huge industry around recording and verifying deeds for sales to deal with that type of nonsense. First, they won't want to get dismantled. Fidelity is huge for example. Sort of like how TurboTax inserts themselves between you and paying taxes. Fidelity inserts themselves between you buying and selling a house.
A verified Blockchain would essentially turn everything into torrens instead of abstract title. I think that's a good thing and I'd rather pay the government to verify the transaction than done for profit company that's going to review title as quickly and cheaply as possible.
Nevermind having to deal with a title insurance industry -- like all insurance -- that's inherently be incentivised to reduce costs by not paying claims.
How do you prove someone in the central government didn't take a bribe and tamper with the records? What if you're from a country where the central government is less than stable? How would you prove ownership if something were to happen to that database? How do you prove that someone is who they say they are? How do citizens and businesses access that database? Is there a standardized format for it? Does it use some proprietary software built by the lowest bidder?
Not saying that blockchain has all the answers or that it is the right tool in all cases, but these are some of the problems it is trying to solve.
A central database would be just a list of all the land and who owns it.
Right now, the deed system is a bunch of deeds that say "remember when I got this land, on page 302 of book 75 in the county recorder's office? Well now Jimantha owns it actually, since they bought it from me for ten dollars and a peppercorn.". This is great for accountability: it lets you trace ownership history and provides a piece of evidence to substantiate every transfer, and so helps you answer inconvenient questions like "why should you own that house when it was my grandmother's house and I want to own it?". It also lets you roll transfers back if they are found to be fraudulent, and neatly captures how all current ownership is contingent on the theft of the whole place from any disposessed original inhabitants.
This is also basically how ownership works in many current blockchain systems: you select something you own based on the transaction that gave you ownership, and then you say who should own it now in a signed message.
But the blockchain systems verify signatures cryptographically, whereas the county recorder verifies the authority to transfer stuff on the "you think someone would just tell lies? On the Internet?" principle. And the centralized database doesn't even keep the transfers around for review, it just has the database operator in charge of who owns any given thing at the moment.
Would you rather walk up to a grumpy person with a shotgun and demand that they move out while brandishing a printout of an SQLite database recently recovered after the ransomware attack at the county administrative building? Or with a deed with their spouse's signature on it?
Then the problem is to make the deeds more machine-readable, and to get better at not putting in deeds from people who have no business writing to that part of the ledger, for which pieces of blockchain technology might be useful.