 Thanks for that Daniel. Take it away. You want to speak down here or I'll stand up there. So anyway, good evening everyone. I'm Daniel Austin. I'm chief architect at PayPal. Everybody got a PayPal account, right? Very good. All right, so this is a really fast talk a little bit about me in the past. I was a physicist. I worked at a place called CERN. The LHC lives there now. I ran into this guy named Tim Berners-Lee. It's all been downhill ever since. I'm going to talk today about a subject that makes us two of my favorite topics, physics and computing. So we're going to talk a little bit about quantum computing. You're not going to understand a word of this. It's okay. Next year if I do these talks Bianca's offered to cater us some Red Bull, it'll be useful. When people ask me about quantum computing, they ask why do we need quantum computers? What's the difference between your analog computer, the computer you're carrying around in your wallet, your desktop? It's fundamentally qualitatively a different animal. We use quantum logic to describe quantum computing. It's very different. The answers are not just yes and no. They could be yes, no or maybe. And we don't always get yes or no results. We get probabilistic results. There's a 60% chance that, you know, it's going to rain tomorrow. Well, maybe not in San Jose. But we get probabilistic results. There's also a lot of quantum mechanical effects in here. We get teleportation. We're going to talk about that in about 45 seconds. Some super dense coding, non-causal computations. That means that the cause and the effect aren't necessarily related. Multiversal parallelism, alternate universes and all that. A lot of craziness goes on in quantum computing. Quantum computing is based on circuits which are made up of gates. These represent mathematical operators that operate on some Harry function that nobody can solve. So we use little pictures in order to illustrate these things because if you actually tried to do those integrals, well, it wouldn't end well. The gates that we use are very similar to the ones that you studied in school. Anybody who took an E-class, took a class in electrical engineering will recognize this concept. Basically, these are gates. A signal goes through them. They're changed by the gate and then we get a result. These are a couple examples of the sort of gates that we use in quantum computing. This is an operator. This is actually what we call a Hadamard gate. It's a good friend that H there. We'll see it on a couple more slides. Basically, these are logic gates that we use for quantum computing. I wanted to show you what a standard quantum circuit might look like. This is actually what an algorithm might actually look like. Oh, and if you think I talk really fast, I learned that from Tim. No, seriously, anybody who's heard Tim Berners-Lee talk knows that he talks a mile a minute and that's where I learned it. Seriously, this is what we call the Deutsch algorithm. All of this stuff here with the fanciness, what it does is it turns a one into a zero or vice versa. Nothing more complicated than that. You can imagine that if you're going to go and build Google or Facebook, it's going to take you quite a few of these. We can't do any of that just yet, but it won't be very long. One of the craziest things people ask me about in the quantum world is quantum teleportation. I mean, this can't happen outside of the movies, right? Area 51 is over here, you're over here, can't go back and forth. It doesn't really work that way in the quantum world. We can teleport information, not substances, not objects, but real information from A to B without crossing the space in between. The current record for this is about 143 kilometers. They did this in the Canary Islands from the island of Tenerife to another island and were able to teleport some information across that gap without passing through the space in between. How that happens is a little complicated, but the schematic of the device is here. It involves a beam splitter and some mirrors. Alice and Bob are going to exchange some qubits. Qubits are the quantum equivalent of bits, basically yes, no, and maybe, not just yes and no. The diagram is a little complicated. I don't want to get into that too much. Here, once again, is the quantum circuit description of that. What's our current state in quantum computation? I mean, this all sounds like science fiction. It's great, but it's maybe the wrong convention. Maybe I should have gone to the science fiction convention. No, it's all real. It's here. Google is actually going to run one. I'm not kidding. What's our current state of play? We can read and write 512 qubits, the D-Wave quantum computing. I think it's the D-Wave 2 model now has a 512 qubit chip. You can buy it. Thank you. That means time is up. This is the last slide on this one we're carrying on. We have to go to the next one. The next slides go right in. It's fine. This is really just part two. If there is a segue from one to the other, I'll allow that in extra time here. There was a little segue there. Do you want to mention it now? No, it's fine. Go ahead. Basically, the current state for quantum computation is we can do 512 qubits. We can move quantum teleportation about 143 kilometers. We launched and put into production the first quantum computer in the world today, Lockheed, did, NASA, and Google just purchased a second one, a D-Wave 2. It'll be put into production later this year. The second part of the talk is really about Grover's search, which is a good example of a quantum algorithm. In fact, it's the fastest possible search of a database. Since there's no SQL conference, I thought this might be a good topic. Grover's search is actually provably optimal for searching a random database, which is kind of remarkable. Just to give you an idea of the complexity difference here, we can run Grover into the one half time, whereas normally we run it in O to the N or worse. Actually, most systems in the real world do worse. The gates, once again, I'm going to skip these because I showed them to you earlier and skipped to the meat. This is what Grover's search algorithm looks like. This is the fastest possible search algorithm in this universe. Just to give you an idea of what's happening here, we're applying some operators to some inputs. There's some ones and zeros over there. Those are qubits. They come in here. They go through a Hadamard transform. Then we do some work on them, and then we invert all the things that we did, and then we take a measurement of that. That's what this little object here at the end is, is a measurement gate. We finally actually got a result. Of course, it was a probabilistic result. If you wanted to write code for this, there's multiple languages out there. I use QCL for most things. This is an example of QCL for Grover's algorithm. You can kind of see once again what we do here. We use the H gate, that Hadamard transform. We do an inversion. We check the phase of the bit that we're looking at. Then we undo the inversion, undo the Hadamard transform, and lo and behold, the resulting bit has the answer that we want. It's actually true. That's exactly how it works. Once again, people ask me, you know, so prove it. It's the fastest search. Prove it. Well, one slide isn't going to contain that. This guy's going to be all over me if I try to prove it in this shorter period of time. But just to give you a flavor of the search, of the proof, any alternate algorithm will have to compute just as many iterations as Grover's does. And we can prove that and therefore their equivalent. The one thing about the Grover's search is it does require an oracle, which is a black box. The black box really doesn't have any function. It doesn't change the quantum nature of the algorithm, but it is a required part of this and people ask me about that. Computer scientists talk about quantum computing as if it was some branch of computation or something. They're not physicists, right? So they're worried about complexity theory. And one of the big questions around Grover's search is, does it prove that NP, the number of normal problems, so to speak, is contained in the bounded quantum probabilistic problem set? And the answer, in my view, no. NP is not contained in BQP, Nobel Prize forthcoming next year. The number of iterations is optimal based on a constant around the square root of the number of entries. If for k entries, I'm going to have to do this many iterations over the data set to prove it. Once again, the slide is too small, as for Matt said, to contain this proof. Just summing it up, Grover's search is the fastest possible search algorithm, runs in this number of iterations for k entries. Quantum computing algorithms are based on gates and circuits, defining operators that change the state of qubits. These algorithms are qualitatively different. This is not your grandfather's computing. It's fundamentally different. If you can teleport something from one machine to another, that ought to give you the idea of just how crazy it is. I mean, goodbye network, right? There's a lot of progress in this field right now. We're starting to see the first commercial implementations. New quantum language, Kipper was released earlier this year. It's actually really useful. It's built on Haskell. I don't know if anybody's played with it. And I thought I would leave you with a small bit of advice around quantum computing. I can actually read that. Okay, for my last 20 seconds, I'm speaking tomorrow. I'm proud of me. I'm speaking on Thursday afternoon at 1 o'clock. Even bigger topic. Reconceiving the worldwide web as a distributed NoSQL database. Okay. I'm going to redistribute it. Great.