Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

A faster, more efficient cryptocurrency

news.mit.edu

Post reply on HN
Print view View original

211–220 of 839 posts

Re: A faster, more efficient cryptocurrency

#211

Post by x2f10 »

naveedi wrote: post #38

Full disclosure: I work on the cryptocurrency in this article, Algorand. There are a lot of questions and speculation here about this paper and Algorand. I would be happy to try an answer them to your satisfaction. Some context may be helpful first, though. This paper is an innovation about one aspect of our technology. Algorand has a very fast consensus mechanism and can add blocks as quickly as the network can deli…

Q: How erosive is it to the image of cryptocurrencies that there's so many coins? It seems that every week there's a new coin that does XYZ better. Often, it has no infrastructure (hardware or software) other than a limited number of exchanges that are used as speculation.

Re: A faster, more efficient cryptocurrency

#212

Post by rohan1024 »

paulgb wrote: post #160
SideburnsOfDoom wrote: post #134

Earlier quoted context omitted.

> My family and friends said i was stupid, utopist, being conned, wasting my time, and so lany things. ... All great things start like that. "But the fact that some geniuses were laughed at does not imply that all who are laughed at are geniuses. They laughed at Columbus, they laughed at Fulton, they laughed at the Wright Brothers. But they also laughed at Bozo the Clown." -- Carl Sagan https://www.brainyquote.com/qu…

“First they think you're crazy, then they fight you, then you change the world.” - Elizabeth Holmes

It's not exactly same but the quote credit goes to https://www.brainyquote.com/quotes/mahatma_gandhi_103630

Re: A faster, more efficient cryptocurrency

#214

Post by 52-6F-62 »

semi-extrinsic wrote: post #177
scient wrote: post #95

Earlier quoted context omitted.

But muh decentralization! /s In reality most of the exchanges and sites dealing with crypto were (and probably still are) built by absolute amateurs with no checks and balances on their apps from a security perspective. Its scary as hell that people trust those sites with their actual money.

The recent QuadrigaCX fiasco shows this very clearly. The entire exchange (one of the biggest in Canada) was somehow just running on the encrypted laptop of this one guy, and after he suddenly died in December, there is 250 million Canadian Dollars owed to 100k+ customers that nobody can recover, because the guy with the password is dead.

Minor correction: the exchange wasn’t running on a laptop. It was run on systems you’d expect it to run on.

What was stored on the encrypted laptop was all of the cold storage wallets containing the majority of the digital assets.

No less idiotic, but an important factual distinction.

Re: A faster, more efficient cryptocurrency

#215

Post by thesausageking »

EthanHeilman wrote: post #198
thesausageking wrote: post #171

There's no free lunch. These gains come with a tradeoff in security. Bitcoin's proof of work which uses an algorithm and processing power to secure it. Vault/Algorand throws out PoW and replaces it that with a pure "proof of stake" system that requires 2/3rds of the currency's value is held by honest actors. In addition, if the network were ever to go offline, there isn't a deterministic way to get it back online.

>In addition, if the network were ever to go offline, there isn't a deterministic way to get it back online. I haven't heard anyone say this about Algorand. Can you provide more details? Lets say a bug causes all the Algorand nodes to crash. After a patch is released 90% of the network comes back online. What prevents the selected users from arriving at consensus over the next block?

If the network goes down completely, as it's coming back up, groups of nodes will start adding blocks before all of the other nodes are back online. When connectivity is completely restored, you end up with different parallel chains, each of which thinks it's the main chain. With Bitcoin, there's a simple, deterministic way to decide which chain to trust: the one which has had the most work done it. With Algorand, it's not clear what would happen.

Re: A faster, more efficient cryptocurrency

#217

Post by beatgammit »

skohan wrote: post #83
ses1984 wrote: post #48

Earlier quoted context omitted.

The anarchist in me wants to believe that cryptocurrency will diminish the power of the state but I don't actually have enough faith in humanity that we will come up with something better than crypto-oligarchy.

The story of cryptocurrency over the past couple years has been a really interesting case-study in the application of Libertarian principals. It's interesting how something which was built on the idea of decentralization by design has given way to a clear power structure, where those who have the sway to control the fate of the technology have largely used it to pursue their own personal enrichment over the health an…

Sorry to be "that guy", but libertarianism != laissez faire, it's a principle about governance of people. The gold standard is just as libertarian as cryptocurrencies, and that doesn't require governance to function, and governance through a private centralized bank that everyone opts into is also libertarian. Governance certainly helps in making things more convenient, but it's not required or necessarily desired.

For example, if there are problems with scams or whatever (e.g. impure gold scams, or cryptocurrency manipulation), people may choose non-coercive governance, like banks (e.g. through something like GNU Taler with a bank-backed currency). You could do this with any currency you choose, so which currency you use has little impact on the structure of government and how it interacts with the population, provided the government stays out of it. Because of this, fiat and resource-based currencies are equivalently acceptable for libertarianism.

So yeah, maybe cryptocurrency is tangentially related to libertarianism in that many libertarians are interested in them as an alternative to central, coercive banking, but that's about where the relationship ends. Libertarianism doesn't care if you have governance, as long as that governance doesn't use force.

And since cryptocurrencies aren't really used for real transactions, we can't really see how they would fare in a competitive currency market.

Re: A faster, more efficient cryptocurrency

#218

Post by throwaway415415 »

naveedi wrote: post #38

Full disclosure: I work on the cryptocurrency in this article, Algorand. There are a lot of questions and speculation here about this paper and Algorand. I would be happy to try an answer them to your satisfaction. Some context may be helpful first, though. This paper is an innovation about one aspect of our technology. Algorand has a very fast consensus mechanism and can add blocks as quickly as the network can deli…

A terabyte a month! Wow. At this point, it seems like reducing this for most clients would help. Have you looked into recursive zkp like what coda is doing?

Re: A faster, more efficient cryptocurrency

#219

Post by iamwil »

lukifer wrote: post #144
DenisM wrote: post #87

This gets asked often, but in case something changed recently - what are the use cases of cryptocurrencies that are not: - Attempts to side-step the law (however misguided the law is) - Trivial to do using a typical centralized database (e.g. Federal Reserver or VISA) - Tantamount to waiting for a greater fool to buy out the current players I am not looking for answers of the form "well whenever you want to avoid a t…

It's a very good question. There's an interesting argument that the greatest value is not in the ledger, but in a mechanism for reliable, unalterable history: https://www.ribbonfarm.com/2017/05/25/blockchains-never-forg... It's also worth considering that crypto is effectively the greatest experiment the world has ever seen in free banking: https://en.wikipedia.org/wiki/Free_banking . Hypothetically, this could be do…

What about those with a mimblewimble implementation? As I understand it, those cryptocurrencies know how much was mined, and how much each address owns now, but the intermediate transactions are purged.

Re: A faster, more efficient cryptocurrency

#220

Post by PeterisP »

glass_of_water wrote: post #88
School-Cotton wrote: post #64

Earlier quoted context omitted.

Why do you think moderate inflation is bad?

I don't think the issue is controlled inflation. I think the fear is of rapid inflation caused by governments struggling to make interest payments, which could cause the value of any cash or bond positions you have to effectively go to 0. Similar reason people find gold appealing I suppose.

Intentional debasement is one way of getting out of very serious national financial problems. It is unpleasant, but it's chosen (or, more likely, can't be prevented) because the alternatives are worse. If you make this option impossible for your government, you're setting up your country for some very bad consequences the next time in some economic disaster it would have needed to have that rapid inflation, but can't have it.

Hyperinflation is a symptom of economic problems, not the origin of it. If you find a way to prevent nominal inflation by crypto, it doesn't prevent the related problems. There's a good reason why we got rid of the gold standard, which had all the same problems as a crypto-based economy would have, and was unsustainable because of that.

Since you mention bond positions, if they'd were denominated in non-inflationary crypto, in any such scenario they'd drop anyway - if it wouldn't be possible by inflation, then it'd happen by defaults. Which would also mean a default of all credit institutions. Which would then mean the disruption of most businesses who rely on these institution. Which would then mean nonpayment of salaries. Which would then cause a lack of demand, causing even more economic problems. Compared to that, a semi-orderly inflation (up to the extent that is required) is preferable to a cascade of defaults. The vast majority of the money in economy is anyway in the "I owe you" form, and crypto doesn't change that; even if all the economy would move to a cryptocurrency, only something like 2-5% of it could be in the form of "hard coin that I control" and the rest of it is in debt relationships between various businesses.

It sucks to get paid in some paper that's now nearly worthless - however, the alternative isn't to get paid in hard crypto, in such a scenario the possibility to get paid in full doesn't exist. So if we don't have the possibility to get paid in a devaluated currency, then people either don't get paid at all, or have to suddenly invent a new worthless paper currency and build an infrastructure for it, at a time when they don't really have the time and resources to do it. This isn't a hypothetical example - such "currencies" have developed in e.g. various long-term conflict zones.

Post reply on HN