That being said, here's my devils advocate for your 3 complaints:
1.) Yes, you can make your servers redundant, and keep them up 24/7, or you can use the pay as you go model with the Blockchain. The difference is like keeping a car idling vs. using an electric car.
2.) Writing things to the Blockchain can be very beneficial compared to your own database. Ownership of offline things like bikes can be done with a Bluetooth/NFC chip and a bitcoin address. Also for online transactions, it provides a tamper-free source of truth.
3.) You argue that someone would still have to write the logic for the Blockchain apps. That's definitely true but the impact of a Blockchain startup might replace a higher ratio of jobs than you expect. (Like Kodak vs. Instagram numbers)
Give it time, it's still early to call it a bust just yet.
2) My patent is about something like this. Sure, it works, I actually did this with NFCs and an Andriod app. But it's a heck of a lot easier just to ask a REST API who owns this NFC. Tamper proofing can be done using pieces of bitcoin (like ECDSA) without using the blockchain. Then you're just back to using a PKI.
3) The question is whether there's anything actually new here. Computers came to wall street a long, long time ago. You'd think most of the gains have already been made. Back office has been decimated over recent decades. It might continue, sure, I'm open to that suggestion.
I'm not declaring a bust; I think something will come of it. But the things I've seen up to now don't seem to need it. Something like that DNS idea, that wasn't bad. But I do see a lot of business types hyping the blockchain, and I don't think it's entirely thought through.
> using pieces of Bitcoin
ECDSA is great, have you run across any db's using it? Seems like a great component to extract from the blockchain craze
There is no "state", the state gets created and verified by transactions. What this means is that integrity of the system does not depend on a central unit, it hinges on the environment.
hard to spoil a frikking appletree
[1] http://www.risk.net/risk-magazine/opinion/2422606/-smart-der... (not sure if this is paywalled or not)
[2] http://www.ibtimes.co.uk/setls-peter-randall-says-banks-bloc...
The problem isn't fundamentally changed by having transactions on a blockchain. You either ask the blockchain who has what assets, or you set up some other system to do the same. The problem isn't that they need the information distributed, that can be done by any number of existing technologies.
As for value, consider that even for legal businesses in many countries even getting bank accounts set up is a massive pain, and being able to take online payments even more so. It can be a distraction from anything from days to weeks or even months. In two different companies I've worked on billing we ended up having to implement multiple payment processor backends because we needed to be ready ASAP after getting approval, and so started one integration only to get rejected, started another, and so on. Months of aggregate developer time wasted because there was no immediate way of guaranteeing we'd be able to take payment via a specific provider. It's getting better, but slowly.
International wire transfers is another "fun" one, which is usually slow, expensive or both, and with lots of arbitrary limitations (e.g. I once had my old bank refuse to transfer funds directly to a customer of a specific Russian bank - this was not during a period of any kind of sanctions; instead I had to wire to an account held by the Russian bank in a US bank, with instructions to move the funds on to the recipient account). My Norwegian bank at the time had no objections to facilitating that transfer, knowing full well that the funds would still end up the same place, but it did mean I had to waste extra time getting the recipient to talk to his bank to figure out the "workaround" and then going back to my bank with the details.
There is significant amount of friction in traditional banking, and even providing tools for mitigating that (or forcing the banks into fixing that to compete) would provide huge value. Maybe the downsides will continue to outweigh the possible benefits, maybe not.
> Even if you decide to go blockchain, you will have to meet with the other banks to make sure everyone agrees so they'll want to be on the same chain.
A big difference is that you can opt to bypass the bank for a number of services. Many won't want to, but for others it provides more flexibility in e.g. picking software vendors over a bank intermediary, or creating custom solutions.
(And yes, that leaves plenty of opportunities for people to lose money by doing it wrong too - we have years to go before we'll see what the fallout is and which areas this works well enough)
No, that can be done without a blockchain. The only piece you need is digital signature, an established technology. If someone hands you a block of data with a signature, you can hash it and check the signature corresponds to the purported signer.
> As for value, consider that even for legal businesses in many countries even getting bank accounts...
Yes, because there is such a thing as money laundering laws. I have a friend who is a specialist lawyer in this area, and it's actually non trivial to do KYC, especially for people from certain countries. The procedures and various banks' understanding varies as well, hence your issues. And this is an important point. If there were no restrictions (who can have an account, and what do they have to disclose), there wouldn't be much of a point in bitcoin. Everyone would just find some old bank somewhere that would give them a numbered account, and that would be that.
That's entirely insufficient, as it doesn't prove that the chain you have been given is genuine. It just proves that each transaction is signed by someone holding the right signature. Nothing in your scheme prevents multiple diverging chains. The entire point of the confirmations etc. used by blockchains is that not only each individual transaction can be verified as coming from a specific source, but that the chain as a whole can be verified to meet certain criteria - specifically in the case of e.g. Bitcoin, the criteria that it represents the view of the majority of the mining power of the network.
> Yes, because there is such a thing as money laundering laws.
That's just a small part of it. Providing sufficient information to satisfy that part is the easy part in my experience. I'm sure that would be different in some countries. If it was just about money laundering laws, it wouldn't generally be a problem.
But in my experience the much larger issue is that many banks and payment processors have a lot of policies about what kind of (legal) businesses they want to be involved with, because being an intermediary means they take on risks - whether monery or PR related - and that makes them hard to deal with if you're a small company without tons of cash, and they're not going to tell you quickly in most cases, because they are perfectly happy to bend all kinds of rules and principles if you can convince them there's enough money in it.
In the cases I've dealt with, KYC has never been the issue that has caused delays. The legal hurdles have been cleared in no-time. The issues have been over perceived business risk. And that's fair enough - these companies are businesses too and need to safeguard their own business.
> If there were no restrictions (who can have an account, and what do they have to disclose), there wouldn't be much of a point in bitcoin.
Everyone can have an account. Finding someplace to get an account is trivial if you have nefarious purposes in mind. Heck, with enough money (around $500k last I checked, probably higher now) you can get a fully legally registered bank in a suitably off-shory location, complete with nominee directors and the ability to do your own wire-settlement.
The issue is the hassle involved with getting business banking accounts and associated merchant accounts and a payment processor to service it in a developed country, without lots of waiting or paying truly extortionate rates.
No, this is wrong. I can sign a chain of transactions (Merkle tree is used in BTC as well), and you will know it came from me. If I'm a central authority, I can keep signing it in all of perpetuity, and it will be verifiably from me. As long as you trust the signer, it's very simple.
The reason Bitcoin is different is because you want to be able to have different signers doing each block. The way bitcoin does it is through clever incentives: miners only believe that other miners care about the longest chain, and they want to sign that last block because then the other miners will recognise their mined coins.
But after starting (and quitting) a BTC business with a guy, and getting a BTC-related patent with another guy, I've come to the conclusion that blockchain technologies are a solution looking for a problem, and the only current problem being solved is how to illegally move money.
This is of course a standard thing to say for someone who's just come across BTC (like a savvy old businessman friend of mine) but I've actually looked at the code, talked to people who wanted BTC (a couple of nerds and an army of drug dealers / drug customers), and so on for a long while.
All the blockchain uses I've come across have seemed to be bolt-on hype motivated rather than genuine uses of the technology. I'm even doing a standard web project for a guy who thinks it will be useful to keep track of stuff in containers. I come from the finance industry, so I think I also know a bit about the requirements there, and they don't need it.
So let's look at what blockchain technology lets you do:
1) You get an indelible wall that grows every 10 minutes. You can write whatever you like on it. Well, that's great but you can do that with a normal database. It's well known how to do it, there's backup processes, security, various consistency measures, scaling, and so on that are established.
2) You don't need to run your own database. This is great if you don't want the network to be closed due to malfunction or police intervention. But banks already know how to run a database, they have hordes of people doing that. And they have to comply with information requests from the authorities. And someone, somewhere needs to have all the records, which they will want compensation for. I suggest AWS and the likes (in house), running an old fashioned DB would require orders of magnitude less compensation for similar amounts of traffic.
3) The network automates the ledger system so we don't need as many back office staff. Well, there's no magic here. If you write a blockchain for securities operations, the logic still needs to be written by someone. There's existing protocols for this kind of stuff, and there's existing automation at the banks for processing this stuff. Even if you decide to go blockchain, you will have to meet with the other banks to make sure everyone agrees so they'll want to be on the same chain. You're still going to need back office staff to make sure the inputs and outputs make sense.
Anyway, I've yet to see a blockchain use that both requires and is improved by using the blockchain.