 Hey, Andreas. Thank you so much for everything that you've done for the Bitcoin community and an open blockchain. My question to you is, just how secure is Google Authenticator? Is there a way that you can hack it as a middleman or something like that? Thank you. That's a great question. Google Authenticator, which is a one-time password to factor authentication system, is basically a system that has a private key that installs on your phone. How secure that is depends on how secure the phone is. If you put something like Google Authenticator on an Android 5 phone that doesn't have a hardware security module, or as it's called in Apple terms, a secure element, then that key is floating around in memory. There are other applications that you may have downloaded and allowed far-too-wide permissions or have some bug. They can reach into the memory, pull that private key, and then they have your two-factor. Now, you hear all of that, and this is exactly the problem. This is the point at which a regular user goes, well, screw that, I'm not using two-factor then. I'll just use one factor, because clearly, as Andreas just said, this Google Authenticator in one obscure scenario could be compromised. The truth is that using that will make you thousands of times more secure than not using that, because the alternative is you try to do something else, like use SMS two-factor, which is far, far more vulnerable. The truth is that most people have devices that are quite secure, smartphones nowadays are quite robust devices. They're not as easy to hack. The smartphone you have is probably the most secure device you have. It can store keys for small amounts. The truth is that having two-factor versus not having two-factor is the option. Yes, there are obscure vulnerabilities. Better yet, use a hardware one-time password to factor a device. UB key, now Google is making one. These are hardware signing devices that store the keys on hardware. My laptop has a little thing on the edge, which I just have to tap every time I want to log in, and that stores the hardware keys. That's better than, but both of those solutions, thousands of times, better than not. That's really the calculation you need to think about. Hi, my name is Karen, and the project that I work on works out in border towns in frontier markets, which is some of the places we are tough. One of the things we are running up into is the complete lack of hardware security and affordable price for these markets. We know of people in the space that are working on this, but there's sub-$10 cards that don't exist. And even then, onboarding merchants who are fairly sophisticated that use feature phones with SIM cards for their online banking, but they still can't access credit. This is the kind of security thing that I deal with day in, day out. Uh-huh, absolutely. And the simple answer is, you have to wait. We're not ready for that yet. And it's important to realize that that doesn't mean it's not coming. It's very important to realize that judging the technology or the market by what's available today, misses the evolution of this technology. We're talking about a technology that, like many others, is affected deeply by Moore's law. It's affected deeply by exponential trends in technology development, in the kind of interaction between multiple different projects that all push together in the same way, in the lowering cost of electronics and consumer devices. If you miss that point, you'll look at it and judge it the wrong way. Just like in the early 90s, if you looked at who owned cell phones in those days, when a cell phone was the size of a suitcase, I had a cell phone in 1991. It was about this big and you retracted the antenna, and the battery lasted for an awesome 15 minutes of talk time. But I had it because my job paid for it, because I can guarantee you I couldn't afford it. The only people I knew who had cell phones at the time were millionaires. I think what's ironic is that if you look at that, you're like, this tool isn't helping the people who need help, because you miss the fact that this tool will cost $10 in 20 years. In fact, it goes from being a status symbol of being rich to the exact opposite, where the status symbol today is, if you're rich enough, you don't have a cell phone. You have a secretary carrying a cell phone next to you. Rich people don't wear Bluetooth headsets. That's basically the transition of technology from the ultra-exclusive to the ultra-available for everyone and the drop in price. The first hardware wallet I bought, I paid a Bitcoin for it. You're welcome, Alina. It was worth it, but that's not a price of which we're going to take this market mainstream. There's already some hardware wallets that are dropping in price into the $20 or $30, and I hope we'll see more of those. My question is a little bit different. Maybe it's off the topic, but it's related to the architecture of P2P. So last year, January, February, that Meltdown and Spectrum came up. The computer research scientists, Google and others, found out with flaws in the CPU process and the ability to secure the old play. It's interesting if we look at the price of crypto coins since that point in time, they plunged into a dot-com like prices. My question is, do you believe that we need to solve computing insecurity at the trusted execution environment before we're going to come out of the flow and have the architecture flow be intervened? That's a really good question, actually. Trusted execution environments are the core processing part that allows us to trust that our own computer is not hacking us while we're using it. I don't trust my own computer. Fortunately, there are solutions. The most effective solutions in cryptocurrencies, in fact, are low-tech and offline. There are solutions that involve storing backup seeds as English words. Before we had that great technology, we used paper wallets. We also have hardware wallets. While the hardware wallets themselves may have problems with the trusted execution environment, they interact with the outside world through a very narrow, well-defined channel, usually a USB serial channel. It has a very specific protocol. It's extremely difficult without access to the actual hardware in order to compromise it remotely. We don't need a fully trusted execution environment in order to deliver very high levels of security and privacy on peer-to-peer networks. Fortunately, not.