 Enrico asks, your thoughts on the Kiva protocol and distributed ledger partnership between Sierra Leone, the UN, and Kiva.org? Enrico says, I have been a lender to the microfinance charity Kiva.org for many years. Hey, so have I. I've been a lender on Kiva for almost ten years now. This year, Kiva is partnering with the government of Sierra Leone and the UN agencies to provide a nationwide digital identification system. Designed to help the country's 7 million citizens access the financial services they need to improve their lives, the development called the Kiva protocol is based on hyper ledger technology and is intended to solve the problems of identity and credit history for the 80% of the country's citizens who are unbanked. While I have no doubts about the good intentions of Kiva, I'm concerned about the choice of technology. Would you clarify the pros and cons of distributed ledger technology in such a situation? What are the chances of this project making a positive contribution to the financial future of the people of Sierra Leone? Here's the thing, this distributed ledger technology is centralized. That's the key insight. One of the reasons it's centralized is because the very process of providing identification, traditional human ID, in places like Sierra Leone, is itself centralized. So whether you use a decentralized blockchain or a centralized blockchain doesn't make a huge difference if the source of the identity documents, the source of the truth, the things that you can't control by consensus, but has to be input into the system by an external party, let's call this an oracle for consistency purposes. The government of Sierra Leone in this particular case is acting as an identity oracle that is providing identity documentation for its citizens. That is a centralized function that is prone to corruption, vulnerabilities, lack of access, and all of the other problems of identity systems. Whether they take that identity information and put it in a database, in hyperledger, or in a blockchain, it doesn't really make any difference to the truth of that information, nor does it make any difference to the availability and access of that information. If that ledger is being run with all of the inputs being a single governmental agency, then whatever truth goes in, that's what truth comes out. And if it's a lie that goes in, a lie comes out. You can't create truth with a blockchain. You can only preserve truth that is already there, or you can enforce the things that are subject to the consensus rules. Of course, human identity isn't managed by consensus, certainly not a decentralized consensus. What is the benefit of this? Arguably, if you were to build a database, one of the interesting things about that database, or interesting questions for that database, is how open are the application programming interfaces that give access to that database to a variety of agencies, and even private organizations, to access the identity information that is contained in that database? One of the advantages that you have by using something like Hyperledger, arguably, is that at least you have an open API, a known API that can be used for cryptographic validation of integrity, validation of the records, making changes to those records, and all of the other things that a DLT can do. So, is it slightly better than a database? Marginally better than a database. Could you make it better by putting it on an open, borderless, decentralized, censorship-resistant blockchain? You could make it marginally better, but it would become significantly less efficient. Here we have a trade-off between the efficiency of using a Microsoft SQL server or an Oracle database on one side, all the way up to using, let's say, an open public blockchain, Bitcoin, Ethereum, whatever, in order to install this identity information. Here's the bigger problem, though. If you give access to these APIs to public-private partnerships or private organizations, how do you protect the privacy of the people whose identity is on this blockchain? The problem is that open access to these databases effectively exposes everyone whose information is on this system to everyone who has access to this system. It becomes a surveillance database. Now, is this a good trade-off to have in a country where 80% of people don't have banking, identity, or credit information, and therefore do not have access to loans? The reason I invest in Kiva is primarily because of the opportunity it gives me to extend lending to people who don't have identity documents through the use of a network of relatively decentralized, non-governmental organizations and charities that help people present their needs, document them, and then use a number of techniques to help them repay their loans, including building communities and group lending, consortiums, cooperatives, and things like that, which provably decline default rates. If you have a group of people applying for a loan together in a community to build something that benefits the whole community, they are far more likely to repay their loan because they can both support each other and act as a check and balance on each other in order to make sure that loan is repaid so that the credit of the group can continue to improve. None of that involves documentation. When I invest in ten Kenyan women who are building an agricultural cooperative, or a sowing cooperative to make garments, I have no idea who they are. I don't need to know who they are. It doesn't matter to me whether they're documented or have a good credit rating or not. It's the power of the group that makes it. One of the things I really don't understand is this doesn't help the Kiva bottle, which is rather surprising. What it does is it moves the Kiva model more to traditional banking, where the importance of identity and credit is emphasized. I'd like to move to a world where we can do lending more and more based on the power of numbers and decentralization, where by spreading default risk across many lenders and borrowers, removing the need for identity opens banking to more people. We don't solve poverty by creating more identity. We solve poverty, in my opinion, by making the demand for identity less prevalent, by opening more doors to the availability of funds, credit, commercial opportunities, and liquidity to people who don't have a valid government identity. I'm very skeptical about this. I don't think the problem here is the fact that they're using hyperledger versus whatever else. I think the problem here is the centralized origin of this identity, and the idea that the primary problem with default in third-world and developing countries that have credit problems is identity. I don't think that's the case, but who knows? We'll see. It's a good project. Maybe they'll do some research, and we'll all learn something.