Thank you for an alternative opinion!
So far all I've heard is how PoS will kill ETH.
Another aspect of this switch will be, which coin will gain all of those ETH miners? I have a pretty good idea and hope I'm right. :)
You are viewing a single comment's thread from:
I will say something controversial here, I suspect but cannot yet prove that Bitcoin is already controlled by blockchain cartels. Again it's not like you can easily prove such a thing but what is a mining pool? Mining pools concentrate hashing power. A mining club also would do something similar but again can we be sure there aren't cartels or secret miners clubs? What about Blockstream and who are the shareholders of that? What about various big businesses around Bitcoin?
Power in Bitcoin appears concentrated to the point where we see censorship, where we don't know which chain will be the real chain, will it be Bitcoin Unlimited or Bitcoin Core? These are competing interests. I'm not against the hard fork and currently prefer Bitcoin Unlimited but still I think Vlad Zamfir is onto something.
Proof of Stake or Proof of Work provide security based on assumptions. The logic if you are doing propositional logic or proofs or math depends on axioms. So we start with some basic assumptions but if we find out later that these assumptions are flawed then what shall we do?
Bitcoin is supposed to be secure based on the assumption that it would cost an adversary too much resources to capture 51% of the hashing power. I don't think this is true and Dan Larimer himself figured out that this assumption isn't true which is why DPOS exists in the first place. Proof of Work provides an illusion of great security because you see some big numbers (difficulty, hashing power, electricity cost, market cap) or whatever but the actual security rests on game theoretic mechanism design. It's the game theory which produces the security for Bitcoin and not hashing power.
This means Casper if they are right that cooperative game theory and their new starting assumptions are more accurate then it will mean the incentive structure will be better and that would mean more security based on better mechanism design. This means yes you can have greater security in Proof of Stake if as Vlad Zamfir noted that the cost for a cartel to undermine a protocol guarantee is of a sufficient size. This is basically, if you have a large enough economic deterrent this can produce security by the same means that the US in theory is secure from a Chinese invasion because it would wreck China's economy.
We need to keep kicking the crypto money game-theoretic can down the road a few times in 'good enough' fashion until our first sentient AI can build a really good one--or if that is impossible, at least it can be the first incorruptible 'benevolent dictator' that humans seem to desire.
I don't believe in sentient AI or that it would even be desirable. Ordinary AI can already help us build a cryptocurrency.
I was mostly kidding. Just make sure you buy-in to the Skynet ICO. If you are a sharehodler or miner, the robots might give you a pass. :-)