 I'm Hongshan. Fortunately, or unfortunately, this is a real blockchain talk. So we heard a lot about blockchain in the past days. So yeah, Nagamoto's original design is amazing and cool. It's based on a really simple haji in the quality. It has a lot of features. At least six of them, you can list more. So based on these features, we can see the protocol is really secure. For example, the protocol has lower communication complexity. And that means that it can be executed over a large-scale network. Ideally, it's supposed to be very difficult for the adversary to control the majority of mining power, because the miners are widely distributed. The protocol, as you see in the previous slide, so the protocol is non-negative, which is great in the sense that it's resilient to network delay. There are a lot of other features. I have no time to go there. However, there is a caveat. If you were in Professor Mikali's talk on Saturday, you will see that Bitcoin blockchain are dominated by a very small number for bigger miners, which is bad. In addition, Bitcoin mining is a huge waste of computing resources, waste of electricity. We want to fix that. There are several ways. You may try to use SGX. Even SGX, the other hardware is trust that there is still a single point of failure concern. You can use multiparty commutation, but there's a caveat. Some researchers try to use haji in the quality. And Professor Mikali tried to use haji in the quality and the BFT. So here we do not want to use the BFT, because the BFT requires a stronger little assumption. So how to do? So our first step is to have a new alternative haji in the quality. Instead of having the randomness as a solution, we ask miners to generate unique signatures as a solutions. This looks very similar to Professor Mikali's construction. There's a big difference. So in our design, the haji in the quality is pretty simple, easy. Multiple miners will be chosen. Multiple miners will be chosen. Then they run the BFT to define a block. So here, the haji in the quality is very difficult. Only one miner will be chosen as in Bitcoin. We try to mimic Bitcoin. So this design is already pretty good in the sense that if malicious miners follow the longest chain, it only extends a long chain, then we can see it's secure. It works. However, in proof of stake setting, malicious miners may extend a long chain, a single long chain, a third long chain, and so on. So this is called nothing-a-stake concern. So proof of the understanding of nothing-a-stake concern is quite limited. We discovered that if you play the nothing-a-stake strategy, you can benefit. However, you can amplify your stakes. However, your American rate is bounded by constant E, which is 2.7. And based on this, we can play nothing-a-stake attack against nothing-a-stake. Only the players play nothing-a-stake strategy to defend against the malicious nothing-a-stake strategy. So we can get a nice protocol. Of course, there are other concerns. We will have a fine-grained, whatever strategy. Multiple strategy. OK, sorry, one more time. OK. So I had more. We can improve the throughput. Finally, we have a project ongoing. It's a very early stage called a fractal with someone you know. OK, that's it. Thanks.