My understanding of Bit coin is well represented bythis comic
But as I understand it bit coin mining is done by identifying a solution to a computationally intensive problem, whose only value is that it is computationally intensive and so limits the number of bit coins.
Currently mining operations are taking 73 TW of power per year, or somewhat more than the energy consumption than the state of New Jersey, all to solve problems that have no practical use.
Meanwhile there are a lot of other projects that also are highly computationally intensive but which may actually advance scientific understanding.
Would it be possible to generate a bitcoin like entity that relied on these calculations? Or better yet, but probably impossible, alter the current bitcoin market to use useful problems instead?
Bit Coin “mining” is not designed for arbitrary computation, and the current network does not support such transactions.
If anyone is interested in running a cool screen saver, I would suggest contributing to one of the other programs like SETI, protein folding, prime numbers, etc. Last time I checked, the reward was your name on a top-participants list; they didn’t pay you money, but that’s OK. (And if they did pay money it would be for actual supercomputer time.)
It may be in principle possible to replace bitcoin by something that does arbitrary calculations, but it’s not clear that this could be efficient from the point of view of those paying for the computer time, and on the other hand if you want secure e-money you can do it using distributed computation techniques that do not require any “mining” at all.
Bitcoin minding specifically relies on finding solutions to a very particular computationally expensive problem, and that problem doesn’t really have any useful applications aside from making bitcoins.
There are other blockchain networks that allow for cryptocurrency to be used to compensate distributed computing tasks. The best-known one is probably Ethereum.
The calculations aren’t just for creating bitcoins. They’re also for tracking transactions made in bitcoins. When you pay some bitcoins to someone for a pizza or drugs or the password to unscramble your computer or whatever, some miner somewhere is doing the calculation that enables that transaction to go through.
Now, these transaction calculations are much more difficult for bitcoin than they are for conventional banking. But that’s part of the design requirements. If they were too easy, it’d be possible to defraud the system. Conventional banking gets around this by having entities that have special privileges to determine who has what, but part of the point of bitcoin is to avoid such centralization.
Is that simply people with distributed tasks paying users with Ethereum or is it actual fundamentally using the tasks as proof of work that organically generates Ethereum, the same way that bit coin rewards its problem solvers.
As the energy produced directly relates to the human production of carbon in the atmosphere it would be far more productive to just target the primary needs like many distributed computing efforts do.
Crypto-currencies being modeled to do work would be less efficient and would really only encourage wasteful energy use.
Yes, just base it on distributed consensus without any sort of mining. [Historical note: “mining”, aka hashcash, was originally designed to combat e-mail spam, not for currency transactions.]
As for my research funds, I would always choose to invest them directly in building or hiring a supercomputer, as people point out. Supporting some sort of financial network would not even be feasible, let alone efficient, or remotely ethical.
“Distributed computing” on Ethereum is radically different from the kind one might associate with SETI@home. SETI@home hands out different pieces of data to each participant in order to increase total computing throughput. Ethereum computing throughput is relatively fixed and extremely slow. The distributed aspect is intended to resist tampering or censorship and all participants process the exact same software and data. Ethereum distributed computing is performed by all nodes on the network, not just miner nodes, and participants are not rewarded directly.
DoesFoldit come close? It’s a protein folding game that is used to find possible foldings for proteins. According to wikipedia, they’ve had three successes so far.
I think the question the OP has is basically: could you design a cryptocurrency system where the proof of work accomplished some social good. It seems to me that it’s possible, but challenging, and it’s not clear to me that it’s a net bonus in the long run.
In order for something to be a good proof of work system to base a cryptocurrency on it needs to be both scalable to react to changing compute power on the network and quickly verifiable as correct with a minimum of data transferred.
The primary technical barrier that limits the use of cryptocurrencies is that each transaction takes some number of bytes, and every node has to store and process all those bytes. The nice thing about calculating a nonce is that it takes up only 4 bytes per block and can be very quickly verified to be a valid solution.
I believe that this makes things like Seti@home not reasonable. For one, there’s tons of data to be crunched, and for two, I’m not sure that you can actually verify that the result that someone gives from crunching that data is correct without reproducing the crunching. I could be wrong on this part.
Protein folding might be a bit better. I think the datasize is still large, but not as large as the seti signals. And maybe you can tell the difference between a correctly folded protein and a not-folded one with less effort than figuring it out in the first place?
The block chain is a great tool and lot’s of companies and governments are investigating how they can use it. Things like financial transactions (natch) and food security are top of the list.
Cryptography is a great tool. And applied properly, it can be used in a great many situations. Blockchain is almost never the proper application for any given situation, and it’s almost always deployed by people who have no understanding of cryptography at all.
Slashdot had a blurb a couple days ago about the fall of corporate hype over blockchain.
“Blockchain is going to protect patients’ healthcare records!” Yeah, right.
(Meanwhile, Bitcoin went under $6k, setting a new low for the year.)
Anyway, as to the OP. There is no known use outside of Bitcoin for solving the hashes involved. So no “good use”.
You can create a virtual currency based on solving problems that seem hard, like protein folding. Emphasis on the “seem”. But the problem there is that things are not uniform. Solving one protein folding problem isn’t as hard as solving another protein folding problem. So there is going to be an incredibly large divergence between effort put into something and payoff. Which could lead to people trying to cherry pick which proteins to work on. One could simply work on an assigned protein for a modest time, no solution? Drop it and get another one.
In addition, we don’t know for sure the practical complexity of protein folding. (Theoretically it’s hard. But those are for very artificially defined molecules.) So someone comes up with an insight to the problem and starts racking up coins.
Sure, the hash functions are subject to the same kind of breakthrough, but they are base on Math rather than Physics concepts so there’s a lot less concern about something new turning everything on end.
Arguably, this would be a great outcome for society, so hooray!
Although, not a great outcome for holders of that coin. Which suggests that any sort of crypto coin that uses a problem of unknown complexity is going to suffer in the market. Would you want to hold a coin that might become (more) unstable if someone makes a mathetmatical breakthrough.
That depend on what the insight is. Maybe they discover a whole class of proteins that all fold into basically the same shape. That’s no more useful to society than having a single protein that folds into that shape.
True, and this probably illustrates why there’s not going to be a cryptocurrency with a “social good” proof of work.
Any real computational problem is probably complicated enough that we can’t determine the best way to solve it. Which makes it unstable as the basis for a currency, and difficult to rigorously define in a way that solutions actually advance the social good and aren’t simply finding a shortcut through the work definition.
Remember, bitcoin was published in 2009 (or was it 2010)? Is it the case that research into better e-currencies has not abandoned the hashcash/blockchain method? Say that today I want to roll out a new network allowing thirsty physicists all over the world to instantly pay for coffee and cold drinks using e-credits. It should support 1000 transactions per second, be decentralized, non-hackable by anyone including the implementers, anonymous, and usable via a simple smartphone app or physical card. Is a bitcoin/ethereum-type algorithm the first tool I reach for to put this together?
No. There is no extant technology that can do that. That does not mean it’s possible to do it with blockchains. The reason they can work on a technical level is that they are designed to be inefficient to make the cost of compromising them too high to be practical. The point of mining is to expend energy to win a race to be allowed to update the network, for instance. You could pay for enough computing capacity to perform a “51 percent” attack, but the idea is that would be more expensive than the profit you’d make from doing it. Additionally, issuing and validating transactions requires each participant to keep their own copy of the blockchain. The current Bitcoin blockchain is about 190 GB, and a more active chain would be much larger.
This is where the decentralization really fails. It is not reasonable to expect everyone who wants to participate in the network to download and carry a full copy of the blockchain on their phones and other devices. And you don’t. You can rely on an, er, central entity to host the blockchain, and this is how phone-based Bitcoin apps and the majority of web-based and desktop-based apps work. Once you are willing to do this, there is no reason to keep up the pretense of a decentralized network anymore. Additionally, while theoretical discussion of how cryptocurrencies work implies miners are individual entities, in real life mining is performed by large centrally controlled pools. The result is that cryptocurrencies are no longer decentralized in any meaningful sense, and this is a natural and expected (by rational people) result of economies of scale.