 Welcome to the AI for Good Global Summit 2018. I'm delighted to be joined by the technologist, Johnny Penn. Johnny, thank you for joining us. Thanks for having me. Johnny, you are a Google Technology Policy Fellow and an AI researcher at the University of Cambridge. So to start with, let's cut to the chase. We're here to talk about AI for Good. So tell me how you think or you envisage AI could be a way to speed up the delivery of the UN Sustainable Development Goals. I think there's a lot of different ways that AI can come in useful to help accomplish the Sustainable Development Goals. You know, in the ways that it's being implemented out in the world now, it's kind of, people talk about an intelligence explosion, which is why they worry about killer robots that one day we're going to create a machine that is able to teach itself and it'll be more intelligent than anything we've ever confronted. I talk about a creativity explosion. I think that the sort of uses that AI, the sort of things that AI makes possible, are often you see them executed in really creative ways. So team at Stanford, for example, that figured out an app that could tell you whether a mole on your body is cancerous or not. So you can just take a picture of it and we'll use machine learning to categorize whether it's cancerous or not. Team in Tanzania found a similar sort of way of, if you have a leaf in a field, you can tell whether a certain bug is eating it, depending on the photo you take of it. And so in those sort of kind of pattern recognition tasks, I think that's going to be a really effective tool to start accomplishing the SDGs. But I think that it also has its challenges because we need to make sure that we're funding research and getting the best researchers, the sort of education they need to go out and create those things. So you don't believe in the technology as an end in itself as a way to enable to unleash all this human creativity? Yeah, yeah, I study the history of technology and I think of technology as a means, it's not an end. I think a lot of times businesses talk about technology like it's an end. And actually, when we think about what sort of future we want with AI, a lot of people get intimidated and they don't know really how to identify with it. And then what I tell them is just, think of your values, what do you want out of your life and what do you, for you and your friends and your family? And then how can we use AI to help accomplish that? So something for me in my life, I would love to see education be free. I would love to see healthcare be free. I would love to see the three-day weekend come into effect soon. Things like this that I think AI can help us accomplish because it's going to take sort of work away. But if we just fill the jobs that disappear, we're not necessarily moving towards the future that people might want. And so, yeah, whether or not AI can do good, I think for me it comes back to, is it helping you get closer to the sort of life you want to live? Earlier, you were talking about the future. Youth is the future. And you were actually a panelist on an interesting session dedicated to youth engagement and the development of AI. What can you tell us about it? Well, you know, a lot of, I got to go to this, the biggest AI event in the world. It's called NIPPS last year. And I met a lot of different researchers. They're young people in the early 20s who are some of the, you know, the kind of young bright minds in the field of AI. And they said, it's such a shame that this technology that we've created, which is being compared to the next electricity or the next fire in terms of like, how influential it'll be for our species over time. They're saying, we're using it to sell ads. And what a shame that is. And I think for the future with young people, I think it's important not to alienate them with the idea that AI is kind of purely mathematic because we're going to have to spend a lot of energy kind of fitting AI into the world. And that's a task that's going to take people from all different backgrounds. I mean, the first prototype of artificial intelligence built in the 20th century was built by a political scientist with no formal mathematical training. You know, so if we don't include kind of all types of young people and just embrace diversity generally, I think we risk kind of overfitting our society to the needs of artificial intelligence as opposed to our needs as human beings and of the world, you know, that we live in. So once again, yeah, it brings me back to, you know, leading with our values. And one of the values I would like to kind of advocate for is that young people co-design the future and not get kind of handed this, you know, massive no jobs, a dying planet and a kind of tough economic crisis that let them tell us the sort of future they want and they should have the power to kind of create that future. Johnny, you have done a lot of work on human emotion, haven't you? And fulfilling your dreams and having a vision, a positive vision for mankind. What do you think of people who see a field of AI, robotics in particular, human or robotics, as a solution to replace humans where the human touch is sometimes missing? I'm thinking of nursing homes, hospitals, children for instance. Do you believe that's the answer? A lot of people say now that we should be using AI in education because it can give a kind of higher level of education to people who otherwise wouldn't have that level. And I think that's true to some extent but it doesn't have to be the only way. Exactly for the reason you're talking about it. We need to embrace systems that keep the human at the core. I mean in my education growing up I've had teachers who saw more than I could see in terms of like the potential or my interests in something and they would coax it out of me not just as like rote learning but in making me believe in myself and think that I can accomplish something that I at the time didn't. And yeah, there's so many things that you lose from taking education and putting it into a system that we just have to make sure that if we are gonna go the way of having iPads that teach young people everything then we also need to provide all sorts of other create an ecosystem around that of different caretakers and mentors and things like this that can lift them up and bring them into a better future. And I think basically it comes down to the question of purpose is like can a machine tell you what your purpose is or is a human better suited for that task? So it's not that we can't use AI in education but we just have to make sure not to overfit because if you look historically we've overfit to scientific hubris in the past and it doesn't always work as well as we hope. So I'm optimistic but I'm also kind of I would hedge that we have to take a few steps in addition to just creating great tools, great AI tools. So AI, yes but let's not forget the human touch or input. Yes, yeah. My dad says that friends let you change. When you have a friend you can test out different versions of yourself. I think that sort of stuff, those kind of weird human intricacies that you get just from personal interaction are gonna be really hard to put into a machine. So it doesn't mean we don't use AI. AI is gonna be great in some ways but it's gonna mean that we do other things as well. Johnny, thank you very much. Thanks for having me.