assurance contract in mining process

[cunicula]

tl;dr bootstrap currency adoption by forcing miners to donate money to merchants they like

As many of you have noticed, I am a proof-of-stake evangelist. Proof-of-stake alone is pretty interesting. However, in order to boost interest and attract a developer, I am considering additional modifications to bitcoin that might synergize nicely with proof-of-stake.

One of the core problems in developing an exchange network is convincing merchants to begin to use it. Empirical analysis of other payment systems suggests that the primary obstacle to wider use is merchant adoption costs (I'll post the link again if you like). Consumers, by comparison, face much smaller adoption costs. An institutionalized bounty system in the blockchain might help defray merchant adoption costs and convince them to accept cryptocurrency payment.

It could work as follows:
1) 20% of the block reward is set aside as a bounty and not distributed to the proof-of-stake miner
2) Whenever a block is found, the proof-of-stake miner votes for a candidate address (likely belonging to a merchant) to receive the bounty
3) Every hundred blocks, there is a check to see if any candidate address  has attracted votes above a certain threshold (say 30 or more of the last 100 blocks)
4) Provided that the min threshold is met, then all of the bounty is sent to the candidate address with the most votes
5) If the min threshold is not met, then the bounty accumulates and is added to the next 100 blocks.

The blockchain can enforce these rules.

The idea of this is to direct resources to activities that further the collective interest of stakeholders. One core interest will be merchant adoption. Merchants could be asked to set up a bitcoin payment system and post their address. Stakeholders and merchants could then campaign to have other stakeholders use their votes to send the merchant the bounty money. The system could also be used to benefit charitable causes (though getting 30% agreement would be hard).


I think institutionalizing this in the blockchain will make it a lot more widely-used, then it would be as an ad-hoc add-on with voluntary participation. Here, participation is forced on all currency owners instead of being voluntary. Voluntary participation suffers from the tragedy of the commons. I also think that stakeholders will have much better incentives to vote so as to further bitcoin adoption than proof-of-work miners would. After all, stakeholders are the direct beneficiaries of currency appreciation.
  
Comments?


Edit/Note: I have actually suggested this before, but it was ignored. Recent discussion with BubbleBoy reignited my interest in this.

[BubbleBoy]

4) Provided that the min threshold is met, then all of the bounty is sent to the candidate address with the most votes
[...]
The idea of this is to direct resources to activities that further the collective interest of stakeholders.

What's to stop the largest pool to vote itself and share the extra 20% generations with it's miners ? If the pool owns 20% of the market, then by voting itself the payout would double, and the profit would skyrocket since the extra half revenue does not require any hardware. That's surely in the collective interest of the pool's stakeholders (miners); if they desire to give the money away to promote the currency they can freely decide to do so when they hold the money.

[cunicula]

4) Provided that the min threshold is met, then all of the bounty is sent to the candidate address with the most votes
[...]
The idea of this is to direct resources to activities that further the collective interest of stakeholders.

What's to stop the largest pool to vote itself and share the extra 20% generations with it's miners ? If the pool owns 20% of the market, then by voting itself the payout would double, and the profit would skyrocket since the extra half revenue does not require any hardware. That's surely in the collective interest of the pool's stakeholders (miners); if they desire to give the money away to promote the currency they can freely decide to do so when they hold the money.

1) Not a problem at all. Stakeholders are the voters. They promote their own interests. Good. They are not proof-of-work miners, so they should incentives to behave in a way that promotes the collective interest of the currency. If the majority of stakeholders feel that it is in there best interest to redistribute coins to themselves, then that is exactly what they should do. If they feel they would be better off by expanding the pool of users through merchant promotions, then that is what they should do.

2) Pools are not very practical in proof-of-stake. Variance is reduced by a couple orders of magnitude under my proof-of-stake system. Pool's reason for existence is no longer there. Moreover, the pool would need possession of 20% of all coin in existence to command 20% of mining power. People won't trust pools to hold on to huge sums of money unless they are very trustworthy. If they are very trustworthy, then fine, let them make decisions.

3) If you were really worried about it pick a higher minimum threshold for agreement. Say 50%, then the pool would need 50% of all currency in existence. In this case, it would already be a monopoly power anyway.

[cunicula]

; if they desire to give the money away to promote the currency they can freely decide to do so when they hold the money.

Tragedy of the commons. Say I hold 0.1% of all bitcoin. I send money to a merchant to get him to adopt bitcoin payment. My motivation is to boost bitcoin prices. I get 0.1% of the benefit and pay 100% of the cost. The other 99.9% of the benefit goes to other people who own coins. They pay 0% of the cost. Result... No one does much to promote adoption. It is not individually rational.

Under my system everyone is forced to contribute in equal measure when some kind of majority (30%,50%, whatever) agrees that they should. Stakeholders who own more bitcoin must contribute more because they stand to benefit more. Democracy -> Solution to tragedy of the commons.

It's worth noting that some similar voting system could be used to set txn fees under proof-of-work mining. This would result in monopoly level fees preventing the tragedy of the commons problem that leads to a highly insecure blockchain. However, it would replace the no security problem with the huge fee problem. Proof-of-stake is the only way of removing both problems simultaneously.

[Vitalik Buterin]

What's to stop the largest pool to vote itself and share the extra 20% generations with it's miners ?

One option is to remove the need for pools by making mining rewards more even. Make a two-tier mining reward system: if you make a block that hashes to less than T₂ it's a normal block and it gets, for example, a 10 BTC subsidy, while if you make a block that hashes to less than an easier target T₁ but more than T₂ you can push the block as a special type of transaction to be included in another block and get a subsidy of 30 * T₂/T₁ BTC. The remaining 10 BTC gets democratically distributed as described. Obviously the block hashing algorithm would have to be modified to something like a hash of a hash since you can't have T₂/T₁ blocks full of transactions in every block, but if this or something else makes mining viable without pools it would make the coordination problem of finding a sufficiently large subgroup for capturing the democracy rather difficult.

Another option is to use approval voting. If the top legitimate candidates have >50% approval then a mining pool would need to have majority hashpower to overtake them.

[coretechs]

Your entire "improvement" is based on your assumption of how bitcoin will ideally be utilized in the future.  What if it doesn't become a global currency, but instead develops into a competitor to the ACH system or something else?  Merchants and charities may be completely irrelevant.

You want to put a tax on block rewards that is redistributed by the miners who are being taxed to begin with, and they can vote to redistribute it to themselves.  Sounds like a lot of unnecessary complexity that gains everyone nothing - aside from requiring additional participation from miners who, for the most part it seems, just want to profit.  Why not find ways create incentives without drastic changes that would completely alter the functioning of the protocol?  I highly doubt the block reward mechanism will ever change.

[cunicula]

Your entire "improvement" is based on your assumption of how bitcoin will ideally be utilized in the future.  What if it doesn't become a global currency, but instead develops into a competitor to the ACH system or something else?  Merchants and charities may be completely irrelevant.

You are confused.  My idea is quite general. The person you are motivating to adopt bitcoin could be a bank...  The underlying assumptions are that there is a network externality and that merchant adoption is costly. Are you disagreeing with these? What bitcoin develops into is not relevant. It just needs to have network properties and be costly to adopt. These conditions are more or less guaranteed given the nature of the technology.

Sounds like a lot of unnecessary complexity that gains everyone nothing - aside from requiring additional participation from miners who, for the most part it seems, just want to profit.  

What miners profit from is determined by the incentive structure. They will always "just want to profit". The currency designers determine the incentive structure. A well-designed incentive structure should allow miners to profit by performing socially useful functions. The current design does little to further this aim.
 
I want to use an incentive structure that allows miners to profit by pursuing socially beneficial goals that benefit all currency holders. I would like to change the identity of miners so that miners interests will be identical to those of investors in the currency. I would like to change the reward structure so that it is privately profitable for miners to pursue activities that benefit investors. This is extremely similar to how publically-held companies operate. Bitcoin should adopt the same institutions. They have proven to be amazingly successful.

Why not find ways create incentives without drastic changes that would completely alter the functioning of the protocol?  I highly doubt the block reward mechanism will ever change.

You could find ways of creating incentives without drastic changes to the protocol. I haven't heard of any promising suggestions. The protocol is so flawed that it will need drastic changes anyway. I don't see any point of proposing band-aids to patch together a crumbling edifice. It would be futile.

[cunicula]

Another option is to use approval voting. If the top legitimate candidates have >50% approval then a mining pool would need to have majority hashpower to overtake them.

I prefer the approval voting system. Note that this is already in the original post.

It could work as follows:
...
3) Every hundred blocks, there is a check to see if any candidate address  has attracted votes above a certain threshold (say 30 or more of the last 100 blocks)
...

Thus, a top legitimate candidates would need to have >30% approval. It could be made into 50%. I think anything in this range (30-50) is fine.