 I'm not sure what I can do about curing a hangover, but maybe that's for next year. As it turns out, the organizers of Slush and their infinite and provocative wisdom asked me to answer this question here. So I'm going to give it a go, and hopefully we can all get there together. But by way of introduction, I wanted to start with an anecdote which hopefully will give you a little knowledge about my own experience and how pervasive and disruptive technology can be and how quickly technology's ubiquity can establish itself and how technology can consistently and repeatedly deliver these black swan moments where unconventional thinking together with purposeful technology can have an impact on humanity. And it all seems obvious and predictable when you look at it retrospectively, but in the beginning it's quite disruptive. So from my own experience, I kind of began my career in the hospitals, and I don't know if any of you have watched Scrubs or ER or any of those shows, but kind of at the time you would go and you would see your patients in the morning with your attending physician, and the night before you'd carry your call books from kind of room to room to room kind of reading up on your patients to kind of get the best differential diagnosis. And we convinced our university to allow us to put amazingly terminals on the ward floors, and we populated those terminals with information from our textbooks or from periodic literature and so on and so forth, and suddenly this became the norm. And thus, we created a company out of this initiative, and the idea was considered state of the art at the time, much like AI is today, but today that technology is entirely commonplace. Clinicians use iPads, and we all use this type of technology to get information for ourselves. Thus almost three decades ago, we created a technology which became WebMD, now universally respected resource for medical information. The technology started to structure this information and allow people like us to consume that information, but structuring information is not enough. And in biology, it is a big data problem, but it's also a complex data problem. And if you look at something like Facebook, Facebook is big. But in the parlance of technology, not very many entities and not very many different types of relationships. So the entities are people, and the relationship is kind of nose. So it's very, very big, but it's this person knows this person knows this person, whereas in biology, it's much more complex. Those entities can be a gene, it can be a chemical, it can be a protein, and what relates them together is equally complex. It could be up regulates, where this gene up regulates this process. It could be down regulates, it could be phosphorylates, it could be isomerizes, any number of different things. And this knowledge, or rather an exceedingly small proportion of this knowledge, is used to drive scientific research. And I know this because my last company before Benevolent was a biotechnology company. In fact, it was Europe's largest biotechnology company. And the scientists would sit around the table and they would discuss ideas of how they're going to go about treating disease, but they would have only a snapshot of the information available. But they did a great job, you know, because they would read something they put in the back of their mind, they read something, they put it back in their mind, they read something else, and they'd have this kind of eureka moment where a disease and its underlying mechanism is linked together. But these are exceedingly rare events in science. In fact, there was a nature publication earlier this year which said that there's 550, only 550 of these things in modern medical history where a disease has been accurately linked with its cause. But now, instead, we're able to use technology to augment this human intelligence, and most importantly, assimilate all the information to make a really smart scientist even smarter. And I talked earlier about computers, how they're in hospitals, how we have them at home, and these are now the known, the norms. And the reason for this anecdote, I think, is that whilst it's no longer state of the art, it is now the norm. And to patients and doctors of tomorrow, AI will be the norm. You know, patients and their families are very, very well read. They have a deep set vested interest in understanding their disease as best as they can. And the promise of AI and indeed what we're capable of doing at Benevolent is the meaningful understanding of that information. Okay, so where MND democratizes human health information so that we could consume it, AI is democratizing the understanding of that information for us in the future to consume. The ability to hold vast amounts of information and grasp its meaning significance, I feel, is where technology is going to be taking us. We can look very closely now at the underlying cause of disease, you know, without having to be domain expert. As I say here, AI helps us to be scientific experts without the need to be medical specialists. And in healthcare, AI is the new black. And in the next five years, we're going to see more transformation in healthcare than we have in the previous 50. And much of that transformation is going to come from technology and much of that technology is going to be AI driven. And those early days kind of in the hospitals to kind of becoming a technologist and then ultimately a technology investor and then back into biotechnology. And then I've kind of come full circle. I'm now back in technology again. And looking at solving what I believe is perhaps the greatest challenge and opportunity for any industry. It's the effective understanding of information. What on earth does all this stuff mean to us? And today, every discussion about the power of technologies and the changes it brings begins with data, the information. And information is the world's new natural resource. It has the capacity to change all industries, especially the healthcare industry. But it presents a huge challenge because there's a lot of information. This is big mismatch between humanity's ability to generate that information and actually understand what it all means. And in healthcare, this is particularly the case. When I look at the updates to our own knowledge graph, we ingest 10,000 new life science publications every night and dissect it and determine what it all means. There's a paper published 30 seconds every 30 seconds. So if you can imagine, you know, if I don't continue to draw it on and on up here on say, so be about a thousand of them by the end of this talk up here. Yet the average research scientists can only read about 200 to 400 papers per year. And it's obviously going to be in an area where they have a bias interest. So this is a real problem for scientific discovery. It means that patients are not getting the diagnosis, the care and the medicines that they need. In short, scientific research is being held back because despite living in a knowledge age, we're not getting smart, smart or fast enough. But what if a scientist could read those 10,000 papers? And furthermore, what if he could read the 25 million papers that were published before that? What if he could read the 90 million patents that demonstrate that scientific innovation? What if he could read all the proteomic databases, the genomic databases, the chemistry databases? And what's more, have digital recollection of what that information means and be able to reason on top of that knowledge. You know, imagine the sheer speed and scale of discoveries that could be made. Imagine the positive impact on society and on medical research. And I believe, and I truly believe, we'll only know what we know in the future with the help of a machine brain. We're here at this conference. There's a lot of AI. I've spoken to a lot of people who are working on AI. And it's everywhere. You know, it's all over the news. It's Skynet. It's everywhere. But I think that commentators and industry experts all agree on kind of two main themes. One, it's going to be very big. And two, it's going to be everywhere. In a recent report by Credit Suisse, looked at forecasting kind of 154 specific use cases in 29 industries. And revenue generated from these is going to be in the trillions by the mid-2020s. And healthcare will perhaps see the highest rate of application and experience the greatest benefit. And for good reason, actually, because we need it. Drug discovery offers an area where technology can truly disrupt the industry and have, you know, significant cost and time savings because the current system isn't working well enough. If we look, you know, there are 13,800 odd diseases of which 5,000 of them are treated. So there's 8,000 diseases that have no treatment whatsoever at the moment. And here is the big opportunity. You know, this is ultimately, it's the patients who are paying the price, who are funding the cost of failure and the profound consequences of not being able to address this in a more meaningful way. I'm conscious that I've spent a lot of time talking about the potential of AI without really explaining what I mean by it. And for this, I'd like to explain a little bit how I think of our own company in this respect. And I'm talking about narrow artificial intelligence. I'm not talking about general artificial intelligence. We're not trying to create a sentient being. We're not trying to supplant human intelligence. We're trying to augment that intelligence. And I somewhat think of the way that our knowledge graph works. It's kind of like a periodic table of yesteryear. You know, you have your elements that appear in that table and you know much about it. But you also know that there's great big spaces of that periodic table of yesteryear that are missing. But you know attributes of what those elements are. You know, kind of roughly where it sits in the periodic table. So you know, if it's a gas, you know, if it's a metal, you know, what row it's in. So you know how many valence electrons, what a satomic number is, you know all sorts of attributes of these things without actually having discovered what they are. So you start with a bunch of knowns that we extract from the knowledge. And then we say, well, if we know all this, we should also know this. And it's that we should also know element of it, which is forming the basis of a hypothesis to treat disease. Which brings us to the big question. It appreciate that it's not a hangover cure, but it's different. And in early 2016, we embarked looking at a cure for ALS. ALS is also known as motor neuron disease. It's a devastating disease. Life expectancy is around three to six months after diagnosis. And it's a very complex disease. There's three different genotypes. The sporadic ALS is a lot of things going on. And we have, as scientists, kind of through the history of time up until now have never really been able to figure out what's happening. And now imagine a machine brain in five hours propose five hypotheses to treat this disease. Those hypotheses were then taken to Citran, which is a center of excellence for ALS in the United Kingdom. It took them a year to test those hypotheses such that they were confident. And in May of this year, they reported out those results. And in fact, it showed that the machine found a way to prevent the death of further neurons. And this is the first time that this has ever been shown before. So we've kind of got to the point where we consider these are AlphaGo moments. You kind of know AlphaGo. You're not an AlphaGo champion. You're not a Go champion. You play a Go champion. You beat a Go champion. It was probably because of the technology. And in our case, there is no cure for a disease. A machine brain proposes a cure for a disease. And lo and behold, it is curing the disease. We call those our AlphaGo moments. And we've had 11 of those in the last 18 months. So we're kind of at the point now, instead of this dystopian prophecy that machines will be ending humanity, we have a machine beginning to try to cure humanity, which I think is a super interesting time. So I know that AI is going to have a ubiquitous part in health care. It's going to play a crucial role in the human condition. And I like the quote by Andrew McCaffey, who's a professor at MIT. It's kind of an analogy of the Industrial Revolution and AI. And he talks about how the invention of the steam engine overcame the limitations of the human body and infinitely multiplied the power of our muscles. And in the same way, AI will overcome the limitations of our individual brains and multiply our mental powers. But to the specific question that I was asked, can a machine brain cure a disease? My answer is it probably already has. Thank you.