On Todays Episode of Let's Talk Bitcoin... Andreas M. Antonopoulos and Stephanie Murphy join Adam B. Levine for an in-depth discussion on the ups and downs of consensus reality and the two "big" methods (Proof of Work and Proof of Stake) which we use to define it. Topic 1: Proof of Stake and Proof of Work Discussion Topic 2: Ethereum Update and Upgrade Path Partial transcript from about 35 minutes in... ANDREAS M. ANTONOPOULOS:"The problem with proof of work as a system is that it's so poorly understood and it's very difficult to see what benefits it offers.... Immutability and resistance against state level colluding actors are benefits you don't really see until somebody tries to attack them. The fact that they don't attack them, in bitcoin, is not something you can celebrate because it's like proving a negative. Another way to put it is that no one really pays attention to a surgeon until their patients start dying. It's very difficult to say 'look at how good I am because none of my patients have died lately', but it's very easy to point to the surgeon famous for performing an operation with the 300% mortality rate. He did an amputation in 25 seconds, during which he managed to also amputate some of the fingers of one of his surgery assistants... And it was such a horrific scene that one of the audience members died from shock. Then the patient and the assistant died from sepsis days later, making it the only surgery with 300% mortality rate... Now, you can point to that and say "That's not a good surgeon" but it's very difficult to look at a surgeon and say "Well, nobody's died yet so that's a good surgeon". This is a general problem with security, because security working means that nothing bad happened. And the absence of something bad happening is not a provable thing. SO you can say that Bitcoin is secure. How do you know? well, it's hasn't been successfully attacked yet. Could it be successfully attacked? Could someone amass 51%? And we can point to examples of smaller systems which have been attacked like Classic Recently, but the problem is it's really hard to see that value. Same thing with immutability." "It may be the case that only one virgin birth industrial, planetary scale proof of work system could be bootstrapped. It already has been. Only one is affordable and in the end maybe only one is needed, because it can offer immutability services to other chains through checkpointing thereby allowing proof of stake to be even more robust. That's not an indictment of proof of work, in fact, it demonstrates the one system that did bootstrap and offers that qualitatively important immutability. Qualitatively different from the immutability you'd get in proof of stake, is extremely valuable because you can only possibly have one, and we can only afford one. It's environmentally very expensive and energy expensive. Then the question is, is it worth it? I think proof of stake can be very powerful for smaller blockchains, especially combined with checkpointing in a proof of work chain. But I think, again, the immutability you get is qualitatively different. One of the important things that most critics of proof of work don't understand, is I can take a computer that's never seen a bitcoin blockchain. I can give it one block, the most recent block. It can look at the proof of work and tell me if that's real without any external network connection. How does it know? It knows because the proof of work embedded in that single block, computationally, would have required 45 Exa-hashes per second to compute in ten minutes. So there's only one of two possibilities... Either somebody was running a lot of general purpose computers for two years to produce just that one block in order to fool you, or it was a block produced in 10 minutes by the only network able to deliver 45 exa-hashes. The bitcoin network. You can make that assessment without any external information just by looking at that one hash and the nonce. Then saying "H
