RE: Vlad Zamfir proposes cutting the mining reward and the community response gives evidence in favor of his oligopoly hypothesis
Vitalik Buterin's response:
I feel ambiguous about this.
Object level argument first. ETH block rewards are currently $14.3 per second. Next highest is Zcash at $4.77 per second. If we drop 4x, then that would put ETH at second place. If the price drops 3x then we would go down to $1.2 per second, which puts us down to third place below Zcash and Monero. At this point, we may be much more vulnerable to 51% attacks. A drop up to 2x seems safe to me, a larger drop less so.
Second object-level argument. Suppose conservatively the proof of stake switch happens in 15 months. A 4x block reward difference starting 3 months from now is only ~7 million coins, and quite possibly less if we do hybrid proof of stake in the mean time. It may be most pragmatic to just let the issue slide until we can reduce block rewards even further (I am projecting 0.1-0.6 ETH per 14 sec) through a full PoS switch.
Now, the meta-level disagreement. I think that as Ethereum is maturing, we need to start taking norms, schelling points, and stability of policy more seriously, especially on key economic parameters. That is, moving fast on precompiles is still fine, but mucking around with block rewards, staking rewards, etc arbitrarily much less so. This is in part because while Vlad fears what miners would do to the blockchain if unfettered, I have similar fears long-term with what developers would do if unfettered, and we should be acting to limit through social norms what any of these groups can do on its own.
There are two situations in which I support dropping the block reward:
- As part of a hard fork which delays the ice age, and has the property that miners get net expected revenue that is somewhere between the revenue that they would get if the ice age was not stopped and the revenue that they would get if 5 ETH blocks with no ice age continued forever. That is, taking the ice age delay into account, the hard fork would not hurt miners relative to the baseline of inaction. Such a reduction should also have strong community consensus.
- As part of a switch to hybrid proof of stake, that is a step along the roadmap to full proof of stake. If Ethereum switches to an algorithm that can be reasonably described as only 50% relying on miners and 50% relying on stake, then it seems totally reasonable to cut the reward given to miners by 50%, and give stakers 50% the reward that they will eventually receive.
I have to say I think it's an error to focus on stability for something like this over adaptability. I do understand the users require some stability but at what cost does that stability come? In essence, what do you get in exchange for this stability?
Bitcoin lost adaptability for stability and what did it get?
Not sure what you mean by adaptability here.
Could you elaborate?
My premise is blockchains exist to support the community. By adaptability, I mean continuous adaptation to the needs and requirements of the community it exists to support. Developers, miners, the code/smart contracts, all only exist to support the needs and requirements of the community of users.
Bitcoin cannot adapt quickly because it has to deal with special interest groups. What will Core devs think, or the miners? As a result we cannot guarantee the best version of Bitcoin for the users and cannot promote mainstream adoption (growth in the user base). I suppose that is fine for Bitcoin but it is not how I would do things.
A community can agree on some principles but to be locked into any set of rules, norms, or principles, in my opinion is not wise. Bayesian methods, when you receive new knowledge you update your best practices, your algorithms, your procedures, it's not supposed to ever end. So I believe stability can happen but not at the cost of innovation favorable to the long term growth and interests of the community.
If the community decides it doesn't want to innovate anymore and I were an Ethereum developer I would probably see it as my time to leave. I would choose to develop for the community which favors continuous innovation over stability as long as the innovation furthers the objective of improving the project in accordance to whatever was promised in the beginning.
I understand what Vlad seems to be saying too, but it seems like a fear based argument. Maybe I'm reading it wrong but it looks like he's saying if these tough unpopular changes to the reward mechanisms are made down the road when powerful entities are involved it could cause a lot of these entities to become upset and this could disrupt stability in the ecosystem or cause all sorts of problems.
But I would see it like, maybe that is a problem in itself. If developers cannot do what is best for the community without worrying about upsetting certain powerful entities and or members then it restrictions options and reduces adaptability. I do think everyone should have a voice but what if the situation is a change is made or Ethereum loses some security or competitive advantage lost? When the "mechanism design" or incentive structure is flawed I think its important to always be capable of modifying it as new knowledge indicates a need for a change.
Great discussion btw, apologize for late reply. I agree with a lot of what you said. I am not too familiar with all technicalities of ethereum, but I believe we are a lot better off and a lot more ahead with the one we have here - at least compared to bitcoins. Mining and how it evolved in my opinion is just a waste of electricity and investment into asics for the sake of mining. I like the way we mine here a lot more, and I feel that person on twitter that replied to Vlad with threats is a bit ridiculous about it. This technology is still in its infancy, he seems to be too emotionally involved in his investments. :P