Live data from Hacker News

Ask HN: Are there any substantial examples of blockchain solving a real problem?

news.ycombinator.com

71–80 of 192 posts

Re: Ask HN: Are there any substantial examples of blockchain solving a real problem?

#71

Earlier quoted context omitted.

Well than this at least makes it decentralized and cheap (though of course that central entity could provide that as well in theory...)

Yes, it makes it decentralized, but so what? The whole point of this discussion asked by the OP above is to question why (and IF ) decentralization is an advantage - what practical benefits does it achieve for this particular use case? Decentralization is a trade-off; it almost always incurs an efficiency cost compared to an equivalent centralized solution, so the question is if it brings some specific benefits that…

I agree.

> timestamping services are trivially cheap already

what does it cost, where can I do it right now?

Re: Ask HN: Are there any substantial examples of blockchain solving a real problem?

#72
Certificate transparency (http://www.certificate-transparency.org/) solves the real problem of allowing website owners to check that no one issues illegitimate certificates for their domains. I believe it is technically a blockchain although it doesn't have most attributes traditionally associated with blockchains (decentralized, PoW).

Re: Ask HN: Are there any substantial examples of blockchain solving a real problem?

#73
post #61

Earlier quoted context omitted.

A well-run, transparent, centralized entity can easily become an opaque and corrupted entity over time. A well designed decentralized system will maintain its integrity indefinitely. In practice we've only really seen this play out with Bitcoin so far, but that's because the space is young and people still haven't really gotten their heads around the exact strengths and weaknesses of a blockchain.

> A well-run, transparent, centralized entity can easily become an opaque and corrupted entity over time. And the opposite can also happen. Humans are good at quickly adapting to circumstances and fixing "bugs" in organizations through firing/hiring, new organizational structures, or other means. That's what markets do. A "perfectly designed" decentralized system may be superior, but I don't believe it's possible to…

> A "perfectly designed" decentralized system may be superior, but I don't believe it's possible to perfectly design an incentive structure from the get-go. The world is incredibly complex, and once you have a bug or mistake in your decentralized system, it's doomed because it can be exploited. There are already countless examples of various tokens and blockchains being hacked/exploited. A centralized, human-led organization is less efficient, but more resilient to such things.

Part of the issue here is that the blockchain space as a whole has been _super_ eager to rush ahead of the safety margins and build everything they can imagine rather than build everything that's comfortably within our understanding of how it works.

I would describe a lot of the building happening today to be similar to Mesopotamians trying to build aircraft carriers. Like yeah, boats are one of the most important inventions in history, but you can't just go from river boat to crossing the ocean in one step.

Re: Ask HN: Are there any substantial examples of blockchain solving a real problem?

#74
post #61

Earlier quoted context omitted.

A well-run, transparent, centralized entity can easily become an opaque and corrupted entity over time. A well designed decentralized system will maintain its integrity indefinitely. In practice we've only really seen this play out with Bitcoin so far, but that's because the space is young and people still haven't really gotten their heads around the exact strengths and weaknesses of a blockchain.

> A well designed decentralized system will maintain its integrity indefinitely. That’s a giant leap of faith. If the system is maintainable it is also corruptible .

And that's why you see advocation for systems like Bitcoin to ossify, and get to a state where the protocol can no longer be adjusted.

I totally agree with you, if someone has the ability to make good changes, that same someone has the ability to make bad changes. And a blockchain is best able to deliver on its promise as a tech when nobody has the ability to make bad changes. Therefore, mature blockchains really should be rigid entities.

Re: Ask HN: Are there any substantial examples of blockchain solving a real problem?

#75
The main use case for blockchain is as a financial instrument to represent shares in a community's economy. A community economy is like a national economy except that the value of its currency does not rely on the coercive legal powers of a government. Aside from that, it functions exactly like any normal economy; if participants in this economy (token holders) export more services than they import, then their cryptocurrency's total value (market cap) will go up relative to other national currencies.

I think cryptocurrency is the best use case for blockchain technology. It is a very narrow use case but an extremely important one. It gives a group of people access to powerful financial tools that are otherwise completely out of reach for regular people.

The idea that communities can have their own micro-economies is going to be radical and completely change how we do business.

One specific use case which I've identified is the ability to use a cryptocurrency to represent shares in a business. It offers small businesses powerful financial tools which are not available to normal businesses of that size. For example, you can use business profits to do token buybacks (which are like share buyback which corporations use); with decentralized exchanges, it's even possible to do buybacks without relying on any intermediaries (e.g. no accountants or brokers). Token buybacks are a simpler alternative to distributing dividends to shareholders. Token buybacks (and subsequent burning of bought tokens) allow a company to pay shareholders through capital appreciation by diluting the cryptocurrency's total supply; this gives small businesses access to the same financial tools which big corporations use to reduce their shareholders' tax liability among other things (capital gains are typically taxed much lower than income and only paid when they are realized; I.e. when tokens are sold). Also, having a tradable cryptocurrency gives a business exposure to influential outsiders (e.g. journalists, wealthy investors, politicians) who have an incentive to participate in the underlying small business and then use their influence to help grow that business and its cryptocurrency. 'Airdrops' can also be carried out in highly targeted ways to create a favorable economic and political environment for the business.

With cryptocurrency it's also possible to pay contributors (employees) using the token instead of fiat. It also helps make sure that contributors' incentives are aligned with all other stakeholders. I'm currently part of a business which works like this. We have 7 contributors and everyone has only been paid using crypto for the past few months. We have a buyback-and-burn deflationary mechanism in place so everyone knows that each token is (on average) guaranteed to always go up in value. You can also pay for marketing and other services using the token - I was quite surprised by how many people were willing accept our cryptocurrency in exchange for services; it feels like the majority of people.

Re: Ask HN: Are there any substantial examples of blockchain solving a real problem?

#76

I still like it as "proof of existence", proofing that you where in possession of a certain information at some point in time. (by publishing the hash on the blockchain). I think this is already relevant for court cases.

This use case, especially for legal proof in court cases, has been solved by digital signatures with timestamps - essentially, you hash the document; and a third party issues a digitally signed certificate that they verified that hash at this specific day and time). The technical infrastructure and legal process for that existed before Bitcoin was invented and the blockchain boom, I recall doing some work on these th…

I agree, this is not a very good use case for blockchain. For example, you can use a bank statement or utility bills to 'prove' your home address - It's not as solid a blockchain proof, but the courts will accept it... The fact that courts of law don't depend on blockchain-quality proofs means that blockchain does not actually solve any new problems in that area. It can add a bit of extra convenience in certain cases but the value is unclear.

Re: Ask HN: Are there any substantial examples of blockchain solving a real problem?

#77

Earlier quoted context omitted.

Yes, it makes it decentralized, but so what? The whole point of this discussion asked by the OP above is to question why (and IF ) decentralization is an advantage - what practical benefits does it achieve for this particular use case? Decentralization is a trade-off; it almost always incurs an efficiency cost compared to an equivalent centralized solution, so the question is if it brings some specific benefits that…

I agree. > timestamping services are trivially cheap already what does it cost, where can I do it right now?

Have the Internet Archive archive a document and it’s hash. Courts have accepted Internet Archive snapshots as evidence.

More costly services would be something like DocuSign.

Re: Ask HN: Are there any substantial examples of blockchain solving a real problem?

#78
I'm an old man, so let me show you how to frame thinking about blockchain.

I experienced the 90's, so let me tell you why email will never take off, since we have a trustworthy postal service:

1. Almost nobody you know has an email address, but everyone has a real address.

2. People want to have something they can touch, a personal, perfumed handwritten letter. Not a cold virtual thing with printed letters.

3. You can enclose a polaroid picture in an envelope, this is impossible with email.

4. If you really want, you can send a VHS video tape with the post. This is impossible with email.

5. They say email is faster, but if you need someone urgently, you just call them.

So therefore, email will never become as popular as sending envelope mails. I rest my case.

Re: Ask HN: Are there any substantial examples of blockchain solving a real problem?

#79
post #49

Blockchain has existed as a widespread technology for what, a little over a decade now? And people are already proclaiming its fundamental uselessness? Seems like a massive lack of patience and long term thinking. It took nearly a century for the first airplane to go anywhere and another half-century for it to be widespread and practical.

For information technology that's really long. It didn't take us 10+ years to find a useful purpose for smartphones or the internet. If for many years so much money is thrown at it and so many intelligent people are working at it and nothing fundamental comes it, I think it's safe to conclude that it's nearly useless.

> It didn't take us 10+ years to find a useful purpose for smartphones or the internet.

It took decades before the internet and mobile phones had any significant market penetration, and then within the space of a few years there are fewer people without it than with it. Technology adoption typically follows a Gompertz function: initially slow, but then rapid adoption before declining as there's no larger audience to adopt it.

Bitcoin presents a fundamental shift in how we treat money which can have a knock on effect on every other industry, just as the internet did. If you think it's time to call it off after 10 years (despite continued new interest over that time), this is a bit short-sighted.

The issue with "cryptocurrency," is unlike Bitcoin, they don't actually provide us with any new innovation. They are scams designed to enrich their creators, who pretend that changing a few variables in a codebase can replicate the invention of Bitcoin. All "cryptocurrency" are useless because they don't have the permanent fix for inflation that Bitcoin has. They introduce inflation by design - they are fiat money made digital. Bitcoin's key innovation is the fix for inflation, which is now impossible to replicate because Bitcoin already exists.

Re: Ask HN: Are there any substantial examples of blockchain solving a real problem?

#80

Earlier quoted context omitted.

Yes, it makes it decentralized, but so what? The whole point of this discussion asked by the OP above is to question why (and IF ) decentralization is an advantage - what practical benefits does it achieve for this particular use case? Decentralization is a trade-off; it almost always incurs an efficiency cost compared to an equivalent centralized solution, so the question is if it brings some specific benefits that…

I agree. > timestamping services are trivially cheap already what does it cost, where can I do it right now?

This depends on where you live (i.e. for which legal system that assertion is intended) and what type of identity/signing process you're using, as that timestamping is usually not a separate, isolated service but as an extra add-on to the ordinary digital signing/verification of e-documents.

For signing e-documents as an individual with the gov't ID chipcard I get the timestamping services for free (there's probably some limit to quantity, but it's large enough). But that, of course, depends on the national infrastructure - so it's simple and cheap for me, but I have no idea if you can do it right now easily - there are a bunch of companies that offer digital timestamping for something like $10 per a bundle of signatures, but I have no particular recommendations as I haven't tried them.

For larger companies the general practice seems to be that providers of document flow/recordkeeping management software also bundle the digital signature process, integrating the timestamping services there along with other stuff such as verification of digital signatures on incoming documents, and charge some trivial monthly fee for that integration; I believe my employer is paying something like 50 eur/month for that, which is a bunch of money for an individual (but it's free for individuals) but it's tiny compared to, for example, the cost of drafting and filing a court document explaining how you made and published your document hash in some unconventional way. In any case, that means that the company has a pre-existing solution that works reasonably well and has no obvious benefit that a change to blockchain might bring.

And to preempt a suggestion that this nation-specificness is a failure of centralized solutions and a decentralized solution can be global, this is not really the case - since the key purpose of e-signatures (and timestamps) is to integrate with national legal systems, a decentralized solution with no integrations doesn't work anywhere. E.g. here's a EU list of trusted providers https://webgate.ec.europa.eu/tl-browser/#/ - it's decentralized in some manner (there are many providers providing very different services in very different ways), but it's centralized in the sense that the law acts as a gatekeeper of what signing providers are considered trustworthy. You can have a "we'll put this stuff on the blockchain" signing provider, but doing the exact same thing yourself won't be as trusted, you'll need to (expensively) demonstrate how you're doing that and prove that it's actually the exact same thing.

Post reply on HN