 Thanks for joining us here in Geneva for the AI for Good Global Summit 2018. My next guest is Dr Adrian Weller. He is program director for artificial intelligence at the Alan Turing Institute. Thank you for joining us today. My pleasure, Charlotte. So Dr Weller, when we talk about artificial intelligence today, what are we talking about exactly? It's a very exciting time for AI. We've seen tremendous increases in the capabilities of AI systems over the last few years, and that's leading to many applications that are influencing many people throughout society. So it's a very exciting time to see the sorts of things we're able to do. But we also have to start thinking about some of the ethical considerations. So before we move on to the ethical considerations, that's very important indeed. What are the current applications of AI we can make or use to deliver the UN sustainable development goals? Well, first, let me just point out some applications that we see all around us all the time. Many of us, of course, use a smartphone, and interestingly, smartphones are used all around the world, even in developing countries. We use these to do things like listen to us in one language and in real time be able to translate into another language, which would have been science fiction just 10 years ago, or to be able to recognize objects or faces. Actually, these very same devices are helping to contribute to the SDG goals. For example, we've had some great talks at this conference about things like being able to use a phone, to identify plants or particular good areas of soil for where you'll be able to plant, what sort of things you should plant in particular areas of terrain, or even to help diagnose diseases, be able to take a picture of a growth on your skin, and from that be able to predict whether or not there's a cancerous growth. So the main gains currently are in terms of productivity and efficiency? Well, I'd say there's a mixture. So the sorts of applications I was describing are really more about tasks of perception, where we've seen tremendous gains through deep learning methods. Of course, there are absolutely great areas to be able to improve and optimize efficiency of infrastructure, and that's going to be very exciting, for example, in transport systems all around the world. Also, AI can really help empower people from all over the world to connect in new ways to be able to do jobs which they might not have been able to do before. But of course, there are other very exciting areas such as healthcare, where everyone I think is very excited about the potential. I think that most people will agree here that AI can be a force for good, but you've mentioned earlier concerns about ethics, for instance, and trust. There should be concerns about ethics, of course, with everything that we do, including AI. So AI really enables us to do all kinds of new things, and whenever we have new technology, it can potentially be used for good or ill, and AI is just like other technologies in that respect. So we need to think very carefully. I'd say that one theme that can be helpful for helping to think about how we can address these kinds of issues is to think about how can we try to be sure that we can trust these AI technologies. So we don't want to trust them too much. We don't trust them too little. We want to trust them the right amount, and so it's just that we need appropriate measures of trustworthiness to be sure that they are helping us and not taking advantage of us. And some of the themes that we explored in the track yesterday relate to topics like fairness. That is, how can we be sure that an algorithm is not going to discriminate unfairly against any individual or subgroup or transparency of systems? How can we understand what they're doing? How can we know that they're respecting people's privacy appropriately? Or how can we be sure that they're not unduly influencing us? So we've seen concerns in social media around personalized messages, which might be prompting us to just nudging us a little bit to do things which perhaps are not in our direct best interest. So we need to think carefully about these issues. So is the AI for good summit a good platform to discuss these issues? I think it's a fantastic platform. It's wonderful to see so many people come from all over the world to think together about these topics. AI is a global phenomenon, and it's fantastic that we think about these issues together. Each country is thinking to some extent about how to optimize its own ecosystem. And I think that's actually very, well, it's very helpful for us to try to coordinate how that works. But it's also very useful to see different countries doing slightly different things, because actually none of us know exactly what the right thing is to do. And by different countries taking slightly different positions, we can learn from each other about what actually is going to work well. I think that's something I've heard quite a lot at the summit from talking to some of the guests and the presenters that AI and the successful implementation of AI would have to be a collaborative effort. Well, I don't know that it necessarily has to be, but I think, of course, we want it to be. I mean, what we really want is these technologies to be used to benefit all of mankind, and they can and they should be used for that purpose. And so by collaborating together, we're more likely to achieve that. Of course, there are also exciting opportunities for smaller groups, startups, for example, to create some new product and to help lots of people. And we should embrace that and allow that to continue. Dr. Adrian Weller, thank you very much. Pleasure, thank you.