 I'm just going to take a quick 15 minutes. I don't even think I need all that time to welcome you to this conference. So welcome. So this is approximate quantum computing. And the title was from Advantage to Applications. The best thing about a conference is to have great speakers. So thanks for all the people that have come. And they've come to give the talk. There's going to be a great, great lot of talk over the next three days. I also want to thank David Gossett, who really, from outside, did a lot of the organizing and got you guys to come. OK. So why are we doing quantum computing? I like to say, well, not why, we all know why we're doing. When I go back to classical computing, I think back in the 40s, and this is what classical computers look. And we know how classical computers have evolved. They're everywhere in our life. We use them either on mobile phones or we do big computations of them. And you look at what the pictures look back then. You look at what a quantum computer or what our version of a quantum computer looks like. And it looks really like an industry. So you ask the question, what's going to happen in the next five, 10 years? And this is what we really want to try and get at with this conference. But before I go into that, I want to take a little bit of a time and thank Charlie Bennett, who's at IBM. This year, he won the Procedures Direct Mental Award with David Deutch and Peter Shaw. I think most people would know that Charlie's been at IBM for a long time and been working on quantum computing. So if we can just take a minute to thank Charlie. I don't know if he's here. Yeah. And I think, yeah, I think it's just great that we have someone like him in our group. So the first conference that we did on this was in 2013. As you see, there are actually a few people here. I think Ed Farrie's here, Alan Asimov-Guzik's here, Robyn's here. And it was a small crowd. We ran it again in 2015 with another great set of invited people. This time, we focused more on the challenges. It looks like it increased quite a lot in this year. This year, it even looks much bigger than what we have in the last two years, the last two that we've done. So what do I mean by approximate quantum computing? I agree that this is a made up word. I don't. It's not strictly defined. But the next few years is this sort of time where we're not going to have fault tolerance, but we're going to have something which should be more powerful than classical and how we understand that is difficult. So what I mean by when I say approximate is I mean, number one, it's programmable. It has the ability that we can program what we do. Number two is I can fit the quantum computation within the coherence of my device. And why do I care about this? Well, I care about it for a number of reasons. One of them is, like always, to keep studying error correcting codes and fault tolerance and understanding how we make these devices bigger. Some heuristic algorithms that we've been doing, and I know we'll touch on more later from our side and others will touch on it. What I'm calling error mitigation, which honestly is a lump-hole thing for everything that is not error correction. So dynamical decoupling, smart sampling, and things like that. How do we verify and validate a quantum system? And lastly, can we prove something as a quantum advantage about it? So this is why we chose the name approximate quantum computing. And I don't know. I think this defines to me what we mean when we say approximate. So what are some of the things that we've done that I just want to highlight briefly? I hope all of you have had a chance to look at the IBM quantum experience. I think we put this in 2016, and it was a very good demonstration that quantum computing can move from this phase of being a laboratory demonstration to get it out to a lot more people. Since it's gone online, we were very happy that we managed to get it in all seven continents, which I think is pretty cool that someone in our tactic are actually ran a quantum computer. And from a science perspective, over 35 people have actually used this to write archive papers. Some of them publish, some of them ranging from how we view and understand quantum computers to how we look at inequalities. But it's really allowed to take these devices for it traditionally existed in the hardware and give it to people to try things. And we're moving forward in this evolution. We put out what we want to call our quiz tick, which Andrew will actually take an hour on tomorrow. An hour tomorrow to go over and give you a tutorial on how to use this. So we've started to make this quantum computer not only that you can just launch a simple circuit, but program all through a different language. So if you're interested in seeing how that can be done, please go to Andrew's tutorial, and you'll show you how we're doing it. And we are going to keep building this up to give you more flexibility to do a lot more things with the device. We're embedding first in Python, but we're working towards embedding it in other languages. OK. You've probably heard that we have increased the number of devices, the number of qubits in our device. This is a layout of the 20 qubit device. Each one of these are the qubits. At the moment, it's in the stage that it's been characterized, and we're doing multi qubit benchmark experiments on this. And Michael will go into this with a lot more information in her talk. Some other things that we've done is we really want to understand what heuristics we can do. I love that we got a really nice quote by Alan. So I like to put it up. Thank you, Alan, for that quote. But this was an experiment that we recently put out showing that we can start to do these heuristic algorithms with these devices. And we're interested in seeing where that will go. And lastly, there's many more work. And Sergey will talk a lot more about the theory work. But I want to keep reminding people that we're going to take the focus just off the number of qubits and put the focus on what is some metric that is meaningful for these devices. This metric shouldn't just be how many qubits or how good a fidelity. It's got to be a combination of both. And what we are proposing and what we want to work towards is this quantum volume. And we want to aggressively go up the diagonal of this line. Because just making more qubits, we know that if we can't do many gates, we can simulate that. So it's understanding this and understanding this with both theory and hardware in some way that's going forward. So with that, I want to thank you, welcome you, and I hope you have some fun. And enjoy the conference.