 Hello and welcome to Top 10 Emerging Technologies, a show from the World Economic Forum that looks at the performance of some of the most promising technologies from the past decade. I'm your host Alice Hazelton and in today's episode we're talking about artificial intelligence that can argue and instruct, which made it onto the 2018 list. To get us started let's take a look back at this conversation from one of our events in 2016 where former Dean of Carnegie Mellon's School of Computer Science Andrew Moe discusses just how smart AI could become. The question in probably everyone's mind here is will machines become smarter than humans because at this rate wouldn't that be possible soon? We're all going to argue about what we mean by smart but one by one you are going to see things which we thought required our own personal ingenuity turning out to be things which can be automated. Many professions which we thought were smart and I'm going to actually go out and eliminate here and say the lawyer profession or the doctor physician profession there's a lot that can be automated there and those careers might diminish. There are some other areas which we're going to be using AI to help the humans who will remain in charge such as teaching small kids or nursing or things which involve care and really deep social interactions with other folks. So I do see quite terrifying changes in the makeup of the population but the things for people to do armed with these intelligent assistants sitting on your shoulder are actually going to get more interesting not less interesting as a result. Here today to tell us more about AI that's capable of arguing and instructing is Sophia Velastigy Chief Technology Officer at Microsoft who specializes in artificial intelligence. Hi Sophia thanks so much for joining us today. Hello and it's a great honor to talk at the World Economic Forum as well as talk about how AI will impact all of us. So many of us have digital assistants like Alexa and Siri in our homes today but I think any of us who've ever tried to have an argument with one of them know that they fail pretty miserably. So tell us about the technology that underpins these devices and exactly you know how AI is or could become capable of arguing or instructing humans. The main technology that make up digital assistant is machine learning around natural language understanding with context and sentiment additive is able to convert spoken words into text reference the topic and the sentiments individuals have to activate that conversation. Although we're still on the journey to conversational agent we're seeing success with Shao Ice and GPT-3 which are the most sophisticated bots for instance Shao Ice is able to have 0.5 billion conversation in just three months after launch with the average of 23 dialogue turns which is even higher than human conversation that's about 10 minutes of conversation. In 2018 I know that there was an AI that out outdebated a human in a live demonstration on a topic that the AI had never learned before for someone like you who works in AI can you tell us just how significant that moment was. The significance of an AI system that outdebated a person shows the potential to assist people in deeper conversation with deeper contextual awareness can be a system along with responsible AI can remove bias and enrich our lives. Looking ahead say the next five to ten years what can we expect from you know these AI that can argue and instruct. In the next five to ten years we expect conversation AI will be integrated further into assisting people in their job and life there'll be an increased hybrid digital agent and human engagement just as we were able to perform online searches leveraging significant AI capability this will occur on conversational AI side. We'll have the flexibility to interact in a way that is more natural for us with equal performance and that's super important in allowing others to be more inclusive and feel comfortable so people from age seven to seventy feel comfortable engaging and speech recognition and contextual understanding is still in its journey to improve so we'll see more advancement there. What are some of the technological hurdles that need to be overcome for us to get to that point that you've described. There are five points that I want to highlight number one is mobile internet increasingly inexpensive and capable mobile computing devices and connectivity across the globe which will bring us to number two which is 5G and beyond to increase the communication between the edge devices to the cloud where there's large scale algorithm resize and that includes the performance of the conversation that will be happening with the AI system. It'll also reduce the latency so the response time can be more context conversation and more natural to what we're used to we as human for instance expect 100 to 300 millisecond response time. Number three is to increase that cloud computing to process the conversation like non-traditional computing method quantum to neuromorphic computing which mimics our own nervous system and number four is internet of things or IoT which is a network of low cost sensors which adds another layer signals to increase understanding for instance if we're having a conversation and I can use camera system to understand how you're feeling and emotion I can then respond in kind in the right sentiment and number five is to further investment in training AI systems for a specific subject matter for increased contextual awareness and accuracy. One can say that most of these are already available but the major difference is the pervasiveness and accessibility for all that will change the scale. It's one thing if 100 people have cell phone it's completely different when billions of people have access to it. I think we've focused today's episode on AI that can argue and instruct but if we look back at the top 10 emerging technologies list from the past decade AI has come up a number of different times you know including for searching for novel drugs or AI that can kind of rival human vision so there's many different potential applications of AI. Speaking more generally like what are you most excited about for the future of AI? As a Microsoft AICTO for dynamic where I work in the forefront of AI innovation I see myself as like a detective working with knowledge and becomes more pervasive and accessible for those who are not data scientists. I'm excited to see ever-increasing growth and customization for business and people engagement so there's not a stock response for everyone but is customized for each individual situation and context to address their unique needs and situation. I see a long journey of growth as AI moves from R&D to applied AI to be infused into our daily life and work. Well thank you so much for joining us today Sophia it's been fascinating to hear all of your responses and what you think AI holds for the future. Thank you and it's exciting to see where AI will take us. As we've heard today artificial intelligence that can argue and instruct has made significant headway since making the top 10 emerging technologies list in 2018. We've heard how digital assistants which are commonplace in many households today are getting better and better at interacting with humans but there are still many hurdles to be overcome before artificial intelligence can really converse naturally with human beings. Perhaps what's more pertinent though is the question of accessibility and equitable distribution of this technology. The effectiveness of artificial intelligence will only increase with the pervasiveness of the underlying technologies but it may not be too long before I have today's conversation with an AI instead of with a human. If you enjoyed this episode then please do join in the conversation on social media and we'll look forward to seeing you next time for another episode of top 10 emerging technologies.