More gold-like cryptocurrency mining rules

[cjp]

It has been written by some that Bitcoin's issuance of new bitcoins to miners is modeled to resemble mining of things like gold. However, I think there is a difference, and it is possible to make a Bitcoin-like crypto-currency which better resembles gold. This might decrease price fluctuations.

The supply of gold depends on the gold price(*): when the gold price rises, gold mining becomes more profitable, and therefore more people enter the business, old mines are re-opened & so on. The opposite happens when the gold price drops. This dampens fluctuations in the gold price a bit, since increased mining increases supply, which tends to decrease the price. The reason this happens is because some gold mines are more profitable than others, and the lesser profitable ones are only profitable when the gold price is high. Also, some people are probably more skilled/productive than others, so that some people can only make a living(**) from gold mining when the price is high.

In Bitcoin, OTOH, supply is only a function of time, and does not depend on the price. I guess the result could be that there is less of a supply-side dampening of price fluctuations than in the case of gold.

I think the gold situation (with more and less profitable mines) can be modeled in a crypto-currency by making the block reward dependent on the amount of PoW. You should see the to-be-issued currency as a to-be-mined precious metal, and some of it can be mined easily (requiring little PoW), while other parts requires more PoW. The required PoW should go to infinity(***) as the total to-be-mined currency approaches the maximum, so that no amount of PoW can ever let the total amount of currency exceed the maximum (e.g. 21 million).

Then, at any moment of time, the total PoW of all blocks and the total mined currency should match, according to the pre-defined curve of currency-as-a-function-of-PoW. Each block contributes a bit to the total PoW, and should be rewarded accordingly with new currency. The average time between blocks can be kept constant, just like it is now, by requiring an adjustable minimum difficulty per block

The result would be that, when the currency price drops, mining becomes less profitable, some miners quit, and the difficulty drops. As a result, the amount of newly created currency per unit of time would drop too. This would dampen the price drop. The opposite would happen in case of a price increase.

(*) If you see gold as the absolute norm of everything, then don't read this as "gold w.r.t. fiat", but rather "gold w.r.t. typical cost of living".
(**) or a better living than alternative occupations.
(***) or to breaking of the PoW hash function, which is actually a finite difficulty

[DhaniBoy]

of course there is a difference between bitcoin mining gold, gold mining requires permission from the government and the country, the government surely apply higher taxes for the mining of gold, while for bitcoin mining does not require permission from the government and the state, mining bitcoin not have to pay taxes because the transaction is no on the internet, but both have something in common, namely the mining above both require a lot of fund ... 

[zinger]

What do you mean by "gold-like cryptocurrency"? You mean its a physical coin?

[OrientA]

The difference between gold mining and difficulty adjusted block reward is that in gold mining, you know how much gold you will mine. But as the difficulty changes all the time, you do not know how much coin you will mine.

[brian_23452]

Your premise is completely false.  The large price fluctuations in BTC are almost exclusively demand driven (i.e., speculation) and have very little to do with supply.  And in any case, the supply is not fixed with respect to time the way you think it is.  Obviously, were prices to rise substantially, some people currently holding BTC would sell, and the reverse is also true. 

[pattu1]

Gold would have gone through price fluctuations too, during the previous centuries (although not to the extent of bitcoin).

Total gold production in 2013 = 2.9KT, while estimated total gold mined so far = 174 KT.
So the total supply is increasing by ~2% per year. Bitcoin is increasing by a much higher amount, but after the next block reward halving, it should slow down. 

[cjp]

The difference between gold mining and difficulty adjusted block reward is that in gold mining, you know how much gold you will mine. But as the difficulty changes all the time, you do not know how much coin you will mine.

My proposal would be more like having all gold miners work in a single, shared mine, where everybody is allowed to grab what he can. You know in advance how much gold there is and how difficult it is to get that gold, you just aren't sure you'll be the first to get it.

If the road to depletion of the to-be-mined amount is slow, then the reward (expressed in amounts of the cryptocurrency) for a given amount of PoW is relatively constant. In fact, for some kinds of PoW/reward curves it would be completely constant for most of the time, just like Bitcoin's block reward is constant most of the time (except when it halves). You know your own hashing power, so you know how much reward you can get with that.

An open question is: given that Moore's law exponentially increases mining power (different from physical mining!), is there a PoW/reward curve that allows mining to happen "slowly"? This issue does not exist in Bitcoin, since it completely compensates Moore's law (or any other development in hashing power) with its difficulty adjusting algorithm.

Your premise is completely false.  The large price fluctuations in BTC are almost exclusively demand driven (i.e., speculation) and have very little to do with supply.  And in any case, the supply is not fixed with respect to time the way you think it is.  Obviously, were prices to rise substantially, some people currently holding BTC would sell, and the reverse is also true. 

It's true that mining output is just a part of supply. If it's only a small part, then the corresponding effect would be small too.

[brian_23452]

Right, but the point is, you're proposing completely re designing  how the whole thing works.  Assuming there is a problem, that it is real, if your proposal will have only a negligible affect on the problem then why bother?

[cjp]

Right, but the point is, you're proposing completely re designing  how the whole thing works.  Assuming there is a problem, that it is real, if your proposal will have only a negligible affect on the problem then why bother?

Well, I was just brainstorming a bit; it probably isn't a really important idea.