> It's really not hard to think of useful applications for blockchain.

Then please list some useful, concrete applications of blockchain that would not be solved better and cheaper by a normal database.

Better/cheaper don't enter the equation for things that can't be implemented on a centralized data structure:

- Financial transactions without a trusted authority

- Elections without a trusted authority

- Anonymous messaging (this is admittedly solved by federation if the federation is distributed enough, i.e. Tor).

The first sentence of the Bitcoin white paper describes a problem that can't be solved by a normal DB.

"A purely peer-to-peer version of electronic cash would allow online payments to be sent directly from one party to another without going through a financial institution."

> "A purely peer-to-peer version of electronic cash would allow online payments to be sent directly from one party to another without going through a financial institution."

Except it failed completely at that. The blockchain has proven through real world use that it is unusable as a currency. By its very design, it is fundamentally impossible to scale it to even serve even a moderate sized city worth of transactions.

> By its very design, it is fundamentally impossible to scale it to even serve even a moderate sized city worth of transactions.

1. This is far from proven. Scaling is definitely a problem, but I don't think it's at all been proven to be impossible.

2. You still have to admit that bitcoin has solved this problem at a small scale, which isn't nothing. Transactions without a trusted central authority are literally impossible with previous technologies.

The initial problem was solved. It couldn't be solved using a mundane DB.

Scaling is a separate issue. I don't believe scaling is even mentioned in the original white paper.

But that problem isn't solved by Bitcoin (not can it be solved by any digital currency); Bitcoin doesn't allow online payments to be sent from one party to another without the need for third party clearance, it just increases the number of parties required for clearance.

That's a different point altogether. The original question was what does a blockchain solve that a normal DB can't.

This is a great point, I've been trying to wrap my head around why so much of HN is almost irrationally anti-blockchain.

But it is quite hard to think of a useful application for blockchain that isn't done better and cheaper by existing methods.

No, not really.

Anything where decentralized trust is useful: money, anonymous messaging, voting, etc.

Cheaper/better isn't really relevant, because these things can't be done by a centralized method without trusting a central authority, period.

> Anything where decentralized trust is useful: money, anonymous messaging, voting, etc.

I completely disagree with your examples: trust in money is centralized because I trust a central government will ensure it is legal tender, and that a central bank will act in the best interest of stability. If there's no centralized trusted party in messaging - how can I trust the messages? Trust has to be rooted somewhere - when a newspaper quotes anonymous sources, I'm implicitly trusting the newspaper and its source-vetting process. When it comes to voting - confidence in the vote counter is essential.

I suspect our difference is more fundamental - I assume that just like any other system - decentralized systems can be subverted, and there will be parties who will subvert it. The only difference is that for centralized system, I usually will trust the central authority to fix/mitigate problems. When "decentralized" systems are subverted, you will discover that they were centralized after all when it emerges that someone (unilaterally or by way of some vote) can make "fixes" that affect the entire system (forking the blockchain, dropping some "rogue" transactions, changing the algorithm, etc).

The difference between centralized trust and decentralized trust is you can actually audit and reason about the system. In centralized trust you are typically hoping that human actors obey rules both written and unwritten. In decentralized trust the system is solely responsible and the code of the system is the exact rules it will follow. The tricky part of decentralized systems of course is have their designs been done flawlessly, so there are no vulnerabilities. (A very hard thing to do) In practice, perhaps decentralized systems do not have to be perfect, but just better than the centralized alternatives.

If you trust a central authority such as the government, then blockchain doesn't solve a problem you have.

That doesn't change the fact that many people don't trust the central authorities, often because their situation differs from yours such that it is not rational for them to trust the central authorities. In some cases blockchain solves their problems.

Blockchain solves real problems. They just aren't problems you (or most people on HN) have.

I think Conway's law is something important to note in this context, where there are benefits to running centralized businesses over decentralized (ie communication). Decentralized communication between people is a problem itself, namely, decentralized decision making and the speed at which decision are made.