From the frontend doing what it really says, through intricacies of how the smart contracts work, manual control over wallets, and trusting the business logic.
Transaction security is kind of the last thing to worry about.
But then you only need a mini-blockchain along the lines of what IBM calls blockchain, you may have a few industry players and say a customs agency agree to sign the entries and then there will be no dispute about whether an entry was signed on a particular day. Energy cost of that is negligible. (and you don't need to put the actual transaction info into the ledger... just the signature is enough so everyone on the chain can confirm later that it wasn't modified).
All of this is also pretty ancient tech by now... IBM just basically jumped on the blockchain branding bandwagon I'd say...
Why not just have event payloads signed by a neutral fiduciary?
If I create a leger that requires basic crypto-signing, I would need everyone to agree to use the system, perhaps try and make it the only system so there is an incentive to use it - and that is the difficult part.
Have some blockchain doesn't solve the problem of how you try and get people to adopt it.
How would you express any of this stuff on a blockchain in a useful way? We could have “blockchain yelp” but it’s hard to protect that against fake reviews. And seeing a mix of reviews isn’t enough to know if you should go into business with someone.
Codifying ANY logic is doable, but especially simple logic like that has it's upsides in how many middle men it can help eliminate, and the TONS of human error that comes with it.
In the examples you're giving, you could have codified "here's what happens if the supplier cannot complete the transaction" or "here's what happens if parties cannot agree on the quality of the work" or a zillion other things that already exist.
You did, however, hit on the big one, that when something goes wrong or something is misunderstood or somehow someone needs to talk to a human...well good fucking luck.
The obvious downside of all the decentralization mantra is that of course it means you're liable for your own mistakes, which is just NOT a business model that ever works. That said, the crypto space has gotten massively more centralized (because it turns out market forces still don't give a fuck about your philosophy), and I think the future of the tech lies somewhere else entirely, where end users functionally never interact with it.
The validity of any transaction can only be determined and guaranteed by the state, so honestly just having the state host a central database seems much easier
It could be used for notarising contracts, recording crop yields, declaring bankruptcies, listing property sales - you name it. Pay $1/kb and your data will be stored forever.
Anyone can replicate the changes to their own computer to validate the hashes and prove the government isn’t changing anything.
I think that would capture 90% of the value of a blockchain with almost none of blockchain’s downsides and it would be trivially easy to implement.
This isn't the case for digital only assets though. There are certainly much fewer digital assets people use than physical ones, but they're there. Domain names, forum membership, entry gates, game skins, etc. How many times have you entered a community and thought it would be great to have a tiny bar to contribution, like a small membership fee, enough to weed out spammers and young children?
This is aside from enterprise oriented chains like Ripple that just use chains help facilitate regular business.
You seem to be arguing that blockchain has to magically solve real world physical problems, but paper contracts don't do that either. You still end up in court. Its just a question of how you are substantiating they really signed the contract.
If the problem was about enforcing a signature, you do not need a blockchain, each party to a contract can just store the cryptographic signature themselves or with one or more trusted custodians. That is a fundamental feature of cryptographic signatures not blockchains. And then you need some mapping from legal entities to public keys, but this might as well be a centralized database run by the government (potentially outsourced to a private contractor), after all in many jurisdictions the government already keeps a list of legal entities and metadata about them.
What about blockchains would make them an effective stand-in for building trust over time? One is a ledger, the other is an interpersonal relationship, and the two really are not the same.
If I'm going to take a risk, I'd rather take a risk where there's some human expectation of managing bumps in the road, and which has some means of redress if sufficient malice is involved, instead of one where a smart contract bug means that I've lost a lot of money.
Trust is still required because I need to trust the seller that the car isn't a lemon and he didn't put explosives into the back seat.
The blockchain solves none of that.
We have Yelp, Google reviews, Amazon reviews… how would something like that work on a blockchain/ledger and not be gamed just like the review systems we have now?
If they have to trust me to pay them, do my job, fulfill my contract anyways, what value is there in some sort of digital system that technically decentralizes trust, but where the other side still has to trust me I did nothing malicious with it?
If anything all that the developments in cryptocurrencies show is that a technically trustworthy infrastructure can (and will) be used for nefarious means. That means even if you were to use such an infrastructure you'd still have some interface to the real world, where people cannot pay, not do their job, do their job in a way you did not agree, etc. So you still would have to protect your interest in the traditional way and you'd still have to trust the other side.
So using a blockchain is just like using a database and giving up a certain amount of control over it.
EDIT: replaced "cryptoscammer" with "crypto provider" for generality.
Reminds me of how GWB said that it would be much easier to be president if he were a dictator.
Yeah, I bet! But despite that it's a terrible idea.
Banks don't cause delays because they want to. And it's not a technical problem. Worth repeating: Delays in bank transfers is NOT a technical problem.
All these cryptocurrency "solutions" are trying to tear down a fence they have no idea why it's there. Turns out it was put up over centuries to solve actual societal problems.
People should work on improving the fence instead of thinking it should be torn down to be replaced by nothing.
Then you would need to add some constraints over it for security, but the key difference would be that it's an open market: you can invent one solution, I can invent another and people will be free to choose between them or nothing at all.
When your money comes with some regulatory framework built-in, there is no much room for any choices in that respect and more temptation for all kinds of bureaucrats to abuse their power.
It's not really, though.
Banking is also an open market. You could already start a bank, and some people do. But it comes with a bunch of red tape.
But you can't just say "this is blockchain, so red tape is not required". That's up to the government of the countries you operate in.
And if it looks like a bank (or financial institution), quacks like a monetary transfer or financial product, then why would it not get customers like one, and not be regulated as one?
> people will be free to choose between them or nothing at all.
This was already the case. Wall street invents financial products all the time. Credit default swaps, anyone?
Often they are, but they also already have practical and common solutions
And Europe isn't perfect either. E.g. in Sweden bank transfers are only quick and easy on weekdays. But then again nobody uses bank transfers because Swish is instant and 24/7, and all banks support it.
So yes, it's a technical problem that a transfer is not instant, but the "surprising delays" mentioned by previous commenters are not due to any of that.
Or do you have in mind technical solutions for improving AML/KYC handling that would reduce legit transfers getting delayed?
Sure, systems (legal and technical) can always be improved, but "surprising delay" is a problem from law, not technology. So bypassing it with technology just means (de facto) breaking the law.
The most common complaint I hear is that bank transfers take a long time to settle as a default at a protocol level, not that law enforcement manually interferes with them.
The second most common complaint I hear is that sometimes the transfer is impossible.
For the first there are solutions, newer SEPA transfer can complete in seconds because fraud checking is done faster; this requires some kind of widespread cooperation between banks and governments. For the second complaint there are often no (legal) solutions.