Proof of Work as Currency Unit, no block chain required.

[bytemaster]

All of the fancy crypto, block chain, etc may be far more complicated than necessary.

Suppose that the fundamental unit of currency was 'one mine pool share'.  
Suppose that mining pools accepted work from anyone who was willing to pay to have the hash found.

There is a finite amount of 'work' that can be done in a given period of time.  This represents the monetary base.

A merchant would sell goods and services priced in 'mine pool shares' and the buyer would need to generate enough 'shares' for the merchant.  It is impractical for the buyer to generate those shares himself (like mining for gold in your back yard, possible but too slow/rare to be practical).  So the buyer will hire a mining pool to mine for him.  Deepbit currently averages 50,000 shares/block or $0.01-3 per share.   50 BTC * $30 *10 = $15,000/hour is what people earn on deepbit selling their CPU time.   So a merchant gets work from a pool and gives it to the buyer who buys work from the pool.  The merchant is paid for the work by the pool and the buyer gives money to the pool.

Real value is transferred from the buyer to the merchant, but no block chain was ever necessary.  Clearly it helps that bitcoin has created a default 'market' for proof of work, but there is no reason why the bitcoin block hashing would be needed long term.  Super efficient hashers could 'create work from nothing' and pay people $1 to solve it, but charge people $2 to have it solved for them.

Everyone switches to exchanging proof of work instead of transferring balances with a public transaction DB.  Currency exchanges would turn into working in one currency and selling in another.   So I could get work from deepbit earning bitcoins and then pay people in $USD to find the hash.  The people I pay in $USD may turn around and hire someone else in BTC until ultimately someone provides the CPU time in exchange.  

Completely decentralized, completely anonymous.

You think a government could outlaw 'buying and selling' CPU time?  

[asdf]

The only reason the blockchain exists is to prevent double spending. How does your scheme address this problem?

[thedrs]

also the block chain removes central management

[interfect]

I could see someone paying Bitcoins for GPU time. Instead of mining, miners would probably be happy to compute something else for which people actually had a use, if they were compensated appropriately.

[BubbleBoy]

How does your scheme prevent the original buyer to continue reusing the same proof of work instead of doing the work ? The same for any party that receives a copy ?

[bytemaster]

Everything would work just like the existing mining pools.

No double spending is possible because it is not possible to 'double earn' mining shares.

There is no central management.  Anyone can start a 'pool' and work/hire anyone else.

The reason double spending is not possible is because all 'proof of work' is time limited.  You do not get to 'save' proof of work to 'use later', you must always compute it 'fresh'.  The only way to 'save' proof of work is as an IOU.  Suppose I mine 1000 shares for deepbit, deepbit would either 'owe me' 1000 shares or how much BTC/USD he was paying.

When I go to purchase something from 'Not Silk Road', then I need to do fresh proof of work, or get deepbit to redeem proof of work for me by doing new work today in exchange for the work I did for deepbit last week.

The work is simply the medium of exchange.  To 'own it', you must purchase hardware + electricity and generate work. Because all work is time-limited, (must be completed between now and the next Block), you cannot store more than you can do in 10 minutes except as credits with someone else and if you do not 'sell it' it becomes worthless.

The benefit from this approach is that 'micro payments' are now possible by doing 'small' proof of work between two parties and no blockchain is or 'network' overhead is incurred.

[Pieter Wuille]

You're effectively turning each transaction into a separate block? This will lead to the problem of block chain splits like we have now, on the level of transactions.  You'd have to redo the work if your transaction was ignored by others. As the frequency of transactions increases, this problem would become worse, up to the point were it is hardly possible to get any transaction through.

[TraderTimm]

Your proposal makes no sense.

Please read the whitepaper:

http://www.scribd.com/doc/34237903/Bit-Coin-Whitepaper

This may help as well:

https://en.bitcoin.it/wiki/Myths

[bytemaster]

There is no block chain, there are no splits and I have read the white paper, I have written my own 'block chain based' multi-currency program, and otherwise am 100% aware of how and why bitcoin works like it does.

My proposal only 'makes no sense' at first, but here is what we already know.

1) People will find hashes for 0.0000793/share or $0.002/share on deep bit.
2) As a merchant, if I gave you work I got from deepbit and you solved it for me then I would earn $0.002/share and no 'block chain' or public 'transaction' was required.
3) As a customer, you are lucky to have enough CPU power to generate $10/day.  However deep bit has enough CPU power to generate $100,000/day.
4) If the customer had a means to hire deepbit (or some other mining pool) to do work for them then the customer could easily 'pay' for anything at a rate of $100,000/day.
5) The demand for a 'transfer medium' would bid up the available GPU cycles in order to accelerate transactions or support 'more' transactions.
6) This will cause the price for mining to go up until there is enough mining available to meet market demand for a medium of exchange and would stop when it was no longer profitable to mine on behalf of others.

No double spending, just like there is no 'double mining'.  
All work exchanges are P2P.  All 'conversions to local currency' can be with trusted peers.  

[rebuilder]

What happens as the difficulty rises, and block rewards go down? What happens when no more Bitcoins are mined? Maybe I don't understand this, but it seems like this would be an inflationary currency. Also Bitcoin isn't about paying for CPU time as such, that's just needed to secure the network.

[bytemaster]

It is simply a commodity money and the supply is limited in the same way the supply of gold/silver is limited.  It takes work.

The money supply is not 'inflationary' in the 'money from nothing debasement way' because no one can ever 'store' work except as an IOU of the form, "you generated a share for me today, I will generate a share for you tomorrow".  You cannot store a 'pre-computed result' and ultimately, most people would opt for 'I will generate a share for $0.0x' and you credit my account with $ or gold or silver or what ever other item you will pay me with.  In the future you would charge me the going rate to generate a share for me.

So the currency would be 'inflationary', but inflation only matters when people can 'create something from nothing', in this case inflation can only occur by adding more CPU power which is hardly free and all such additional CPU power further secures the value of proof of work because the work becomes more and more 'scarce' for the individual even as it becomes more available as a society.  This drives demand to purchase a share of 'current global cpu power' and not a share of previously generated hash numbers.  The hash numbers merely prove that you 'own' a certain amount of CPU power.

The fact that bitcoin has value and that there is a reward for finding blocks is irrelevant to the value/scarcity of GPU power.  Forcing people to acquire GPU power and proving it by having them solve proof of work creates a means of exchanging value.  Merchants simply become 'resellers' of GPU power earning money from one individual and having their customer do the work.  

1) Bitcoin is not about paying for CPU time, but exchanging shares of a limited stock secured by CPU time.
2) The value of having a limited stock secured by CPU time has been proven by bitcoin, and now miners have popped up who pay others to mine for them (pool operators).  This establishes a market value for a mining share which is in turn tied to the market value of a bitcoin.
3) If we start trading in mining shares which have an established market value, eventually bitcoin blocks become irrelevant as the 'demand' for shares outstrips the supply of GPU power and people can exchange 'work'.  

  

 

[kjj]

I had to read this several times, and I still have no idea what you are trying to do.

I think it really just boils down to a really convoluted way to move dollars around.

Maybe you should write down a step by step description of a transaction.  Identify each of the participants in the transaction, what they are holding and exchanging at each step, etc.

[bytemaster]

Lets assume DeepBit + Bitcoin Pool are two 'work exchanges' among many.

DeepBit pays people $0.01 per share solved and Bitcoin Pool will find shares for others at $0.01/share.

Joe wants to buy a computer from amazon.   Amazon goes to DeepBit and gets work for 100,000 shares and asks Joe to do it.

Joe has a beefy mining rig and decides to just do the work and after 2 weeks has done enough work for Amazon to earn a computer and Amazon has sold all of that work to deepbit earning $1000 for Amazon and costing Joe significant CPU/Time/Energy costs.

Now Sam also wants to buy a computer, but does not have a beefy mining rig nor a desire to wait 2 weeks to make payment, so when Sam gets work from Amazon he hires BitcoinPool to solve that work for him.  Now bitcoin pool is currently solving 10,000,000 shares every 10 minutes.  So Sam gets the work done for Amazon in 10 minutes and Amazon has earned $1000 from Deepbit.

After Joe has paid for his computer, he still has his mining Rig and so joins either Deepbit or Bitcoin Pool earning $1000 every two weeks.

AmazonEurope prefers to be paid euros, so AmazonEurope gets work from DeepEuro which pays .0075 euro / share.  Sam wishes to buy a computer from AmazonEurope while on vacation, so he gets work from AmazonEurope and hires BitcoinPool in $USD. 

Work is work regardless of who does it.  It is entirely fungible and divisible.     

[TraderTimm]

Okay, I get what you are saying. The problem is your parent post topic subject line.

"Proof of Work as Currency Unit, no block chain required" - is a bit contradictory. Mining pools rely on the block chain or else they'd have no work to do

A better title would be:

"Monetizing Your Hashrate - Trading Hash Capacity In a Secondary Market"

Good luck, I think the more direct way of either mining or buying bitcoin outright is a bit easier.

[bytemaster]

Actually, the title is appropriate because while bitcoin mining represents an 'initial market', the proof of work has value by itself.  Bitcoin would merely bootstrap it and eventually efficient miners can 'create work from nothing' in order to close the gap between demand for work per hour and the amount of work that can be done in an hour.  If there is only 1,000,000 shares/hour capacity and yet 2,000,000 shares demand then a miner can charge more than he pays and thus create work 'from nothing' and still make a profit.

[bytemaster]

The proof of work market would also provide a means of paying to validate 'namecoin' or any other 'block chain'. 

[ffe]

I don't understand what happens after the first transaction.

You get proof of work from a miner. You send it to a merchant for a real T-Shirt.

You cannot double spend because you need "fresh" proof of work to buy another T-Shirt. Hence block chain is not required.

How does the merchant spend what you gave him? What keeps him from double spending? How about the guy third in the chain. How can he spend the coin but NOT double spend?

[bytemaster]

The merchant does not 'spend what you gave him', he simply forwarded unsolved work from deepbit to you.  When you give the merchant the solution, the merchant gives the solution to deepbit and deepbit credits him with either BTC, $USD, or 'share credit'.  The merchant gets paid when he 'withdraws' money from his deepbit, bitcoin pool, compute4cash, etc account.

You, not having a sufficiently fast miner approach deepbit (or another miner) and ask them to solve the work for you.  They do this, but require payment in the form of BTC, $USD, or deducting a 'share credit' from previous work you did for them.


Mining Pool Operator creates random work... 'find $nonce for $data such that hash($data+$nonce) < $target'  and will pay someone $1 to solve it.
Mining Pool Operator can 'solve it' in 10 minutes at an electric cost of $0.01, but Joe Desktop Owner would require '6 months' to solve it at an electric cost of $1.30.
Mining Pool Operator offers to solve that work for $1.10.  Cost to Mining Pool Operator $1.01, profit $0.09.  Thus the 'transaction fee' in this case is $0.10 which is only this high to make it clear who does what and why.

Merchant agrees to solve the work for $1, passes the work to a customer, whom hires Mining Pool Operator to solve it.  Cost to customer $1.10, merchant nets $1.00 and Mining Pool Operator nets $0.09 and power company/ISP nets $0.01.

Clearly a high margin for the Mining Pool Operator, but this is where market forces come to play.  SuperBitPool comes along and says 'I will solve it for $0.55' which forces MiningPoolOperator to reduce his payout to $0.50 and he will only charge $0.52 to solve it.  Margins get squeezed until the difference between buy/sell price for work approaches the transaction cost and the 'value' of that work approaches the electric+capital costs.

[ffe]

the merchant gives the solution to deepbit and deepbit credits him with either BTC, $USD, or 'share credit'.  The merchant gets paid when he 'withdraws' money from his deepbit, bitcoin pool, compute4cash, etc account.

If deepbit can give BTC then we need a block chain or he can double spend or he is acting as a bank an you have to trust him. If deepbit gives $USD, where does he get it? From you? If you, why don't you directly pay with $USD. If deepbit gives out 'share credit' he is acting as a bank again and you have to trust him.

[bytemaster]

Everyone currently has to trust deepbit who acts as a bank for BTC already between the time you send in results and when you withdraw BTC.

The key is that each individual 'trusts' someone.  Currently everyone has to trust Mt.Gox, Tradehill, or some Over The Counter individual in order to 'cash out'. 

To run an exchange you must accept cash or bitcoins and perform other actions that are borderline illegal.

As far as I can tell, there is nothing illegal with paying people to find hashes for you in any currency.

So now to sell bitcoins, you pay someone to hash for you in bitcoins and you work for someone who will pay your $USD to find hashes.  You have no idea whether the other party is earning Bitcoins, Euros, Gold, or Arbitraging $USD. As long as you trust your employer and your employee trusts you the transaction is valid.

The main benefit is greater privacy and the potential for micro-payments and less centralization in places like Mt.Gox, the block chain.

Imagine you wanted to 'pay' Tor nodes for priority bandwidth allocation.  With bitcoin your privacy is completely blown for several reasons:

1) Each node would receive 'small payments' from the same 'set' of coins in your wallet.  Thus a single identity could be traced through the network.
2) The price-for bandwidth would result in micropayments that would be too costly.
3) The latency in the bitcoin network would make it hard to make new connections and 'validate' payment.
4) Merchants take an 'exchange risk' while holding bitcoins.  Ideally a merchant would want to contract for work from a pool at a 'fixed $/share' and then pass that work on to customers.  Merchant now has 0 risk to currency flux.

To be fair, I think there is a place for both Bitcoin and a Proof of Work currency.   Think of everything that can benefit from 'secure hashing' and all of the different block chains that could be 'validated' for a fee if there was a market for such decentralized proof of work.