Scaling risk is a major risk. It's got to be fixed. Actually, there's more than one way to deal with this. We've already talked about changing consensus, which is a major step, but there's other ideas out there. I want to talk about vertical and horizontal scaling. A vertical scaling basically centralizes to some degree the transaction processing. Think of having some super-nodes that are handling the actual key transactions and there's a number of cryptos out there that have a situation like this. This is exactly the trade-off, the trilemma; a trade-off that I talked about. If you look at the Ethereum network, you've got all of these nodes out there and you've got miners, and it's very decentralized. Whereas with this vertical scaling, it is more centralized. There's some key nodes that are really important and you get from that speed, so Solana can actually do 50,000 transactions per second right now. Another one is Algorand, so again, this is like a really difficult trade-off that you become less decentralized, but you greatly increase the transaction speed. There's another idea here, and that's horizontal scaling and it's sometimes known as sharding. Sharding is a word that you need to know and it's a word in my word cloud that we'll see at the end of the course as usual. Think of this as dividing a block chain into different pieces, so there's multiple chains and then there is a master chain that connects them, so you can go from one shard to another shard and this is actually also part of Ethereum 2.0. Just to be clear here, 75,000 transactions per second is not enough in terms of the future for DPhi. This effort has got multiple components to it and all of these components are focused on increasing the transactions per second and increasing the latency. Again, Ethereum 2 supposedly just with the sharding, we can do 50,000 transactions per second, which is a massive improvement compared to what we have today. The sharding that will occur is 64, that's the proposal right now, 64 different shards. This has not been settled yet. There's a lot of detail that hasn't been worked out, but it's definitely moving forward. Sharding's got a lot of risks, of course, one risk is just the way that the shards actually communicate with each other. Often the idea is that certain shards are assigned to do certain things. DPhi might be on one of the shards, but there needs to be communication. It's also a risk that an adversary potentially could take over one of the shards and cause damage, but I think that's not a big risk because the proof of work is replaced by proof of stake, and with the proof of stake, what would happen is, given the stake you put up, you would be randomly assigned to propose a block in a randomly chosen shard. This idea of amassing enough computing power to take over a shard, I think is low-probability nevertheless, it is a risk that we need to consider. This is all being worked out. This is something that can be very important. I say that the Ethereum 2 is going to use 64 shards and that's a proposal, but there's no reason that you can't go beyond 64. It is like, in a way, parallel computing.