 So, economics of a cryptocurrency and the incentives to maintain a cryptocurrency are among the most complicated of projects one can work on. And the reason being is that there actually aren't very good models, and it's not a very well understood problem. With proof of work, what Satoshi did is he picked several magic numbers out of a hat and said, well, pay you X per block, and here's how it's going to deteriorate over time by a factor of two every four years, and so far that's worked. But it's a tremendously expensive endeavor, and we're not sure if it's actually the optimal solution to the problem. It's a solution to the problem. Furthermore, it's created incentives for mass centralization around small private pools who have subsidized power and first access to mining hardware over everyone else. So it's a natural question of, can we choose a different way of governing a network? So with Ouroboros and in proof of stake in general, we have the philosophy that the people who own the tokens are probably in the best position to decide or at least in a position where they think or deeply about what is a good outcome for the system. What I mean by that is that if you act in a way that can harm the network, the first thing that will reflect the harm that you've done is the price of the currency. So if you're a large holder of it, you're actually acting against your best interest whenever you do something Byzantine in a certain respect. So given that assumption, you can work your way to saying, well, can we use this to actually construct a consensus system where people proportionally by how much stake they have, we can infer how much concern they have to the health of the network, and we can then use this to build a good algorithm. That's one of the underlying core assumptions of proof of stake. Now that assumption alone is probably not good enough. You still need to have other mechanisms inside the system to ensure that you have checks and balances and controls. Things like subscribeable checkpoints, things like bonding mechanisms, things like witnesses and other such mechanisms, which give you some cryptographic certainty and give the user some certainty that there isn't any dishonesty in the chain. The point of proof of stake, though, is that it allows you to start having an explicit discussion about who should be in charge, are they doing a good job, and also give you some memory in the protocol so that if people stop doing a good job, they can be fired.