{"id":252,"date":"2016-04-27T14:30:34","date_gmt":"2016-04-27T14:30:34","guid":{"rendered":"https:\/\/www.microsoft.com\/en-gb\/industry\/blog\/industry\/2016\/04\/27\/the-a-i-revolution-will-not-be-televised\/"},"modified":"2016-04-27T14:30:34","modified_gmt":"2016-04-27T14:30:34","slug":"the-a-i-revolution-will-not-be-televised","status":"publish","type":"post","link":"https:\/\/www.microsoft.com\/en-gb\/industry\/blog\/government\/2016\/04\/27\/the-a-i-revolution-will-not-be-televised\/","title":{"rendered":"The A.I. revolution will not be televised"},"content":{"rendered":"
You will not have to stay home. You will not have to plug in, boot up or log out. You may be in the driver\u2019s seat, but who will be driving?<\/em><\/p>\n Well, first, apologies to the late Gil Scott Heron for butchering his famous poem \u201cThe Revolution Will Not Be Televised.\u201d In it, Heron talks about a fundamental shift in power from passive observation to active participation in societal change which won\u2019t be televised, but will be live. More on that a bit later.<\/p>\n Second \u2013 a bit of clarification is in order. To be sure, there is a revolution in artificial intelligence1<\/sup> happening right now. This revolution, however, isn\u2019t the pop culture horror story of smart machines rising up and conquering humanity that we see played out in on our screens, big and small. The real revolution is happening behind our screens (large and small) in the ways that machine intelligence is helping us do everything from prioritising our days to interacting with our customers.<\/p>\n In the first half of the decade, it seemed as though you couldn\u2019t get away from the term \u201cbig data.\u201d With the increase in the connected devices generating a deluge of new information, companies were struggling to make sense of it all. For most, the biggest hurdle was trying to gain a single view of the truth through their data. For a handful of companies, having solved the issue of managing all their data (more or less), they turned their attention to developing algorithms that allowed companies to create sophisticated predictive analytics models to help drive human decision-making.<\/p>\n An even smaller subset of companies has been able to make the shift from predictive analytics to prescriptive analytics, or allowing computer agents to take data and make actions in the real world. And that\u2019s at the heart of machine intelligence. Take banking, for instance. Their evolution from data to machine intelligence may look something like this: first, the challenge was in trying to figure out if a customer had a mortgage, a car loan, a savings account and a credit card with the bank (because that data was often in different systems). Then, predictive analytics would help determine which new products and services that customer would be most likely to purchase, and when they\u2019d be most likely to do so.<\/p>\n Adding \u201cbasic\u201d machine intelligence to the mix, the bank is able to automatically serve up a customised offer to the customer on their mobile when they stop for a period inside a car dealership, or when they visit a branch. Taken a few steps further in the journey, and the bank will be able to not only serve up offers, but with machine intelligence, their automated agents (or bots) will be able to interact with that customer in natural language to solve problems, make new offers, or have a conversation that connects that customer more closely to the brand.<\/p>\n One of the most visible examples of real-world machine intelligence is the driverless car: multiple sources providing real-time data to a sophisticated, yet familiar machine in a way that could fundamentally change the way we live. The driverless car is also an example, in countless movies and TV shows, of our basic fears of \u201cartificial intelligence\u201d: a sentient two-ton, turbo-charged killing machine hell-bent on destruction! (Sensational enough for you?) But sentience and intelligence really are worlds apart, and in the way machine intelligence is being developed, we humans are firmly in the driver\u2019s seat. Machine intelligence research is centred on the ability to understand and extend our actions \u2013 it\u2019s a symbiotic relationship that requires human guidance to learn. So what will we teach them?<\/p>\n If machines pattern themselves \u2013 or more aptly, if we program them to pattern themselves \u2013 on human behaviour, it\u2019s a pretty daunting task. It requires a\u00a0sophisticated understanding of the mechanics of human behaviour which is quite often bewildering and illogical even to other humans. It requires a great deal of trial and error. We can create virtual environments<\/a> in which machine algorithms try, fail, learn and try again. In the less controlled environment of the real world, these experiments in machine learning start to become reflections of society \u2013 the good and the bad. At its worst\u2026well, there are stories<\/a>. But at its best it can connect with some of our most important and basic needs, from food and shelter to love and friendship<\/a>.<\/p>\n So where will this A.I.<\/span> machine intelligence revolution take us? What implications will it have for individuals, businesses and society at large? The truth is, nobody knows. We have a pretty good understanding of what\u2019s now and what\u2019s next, and some\u00a0reasonably good assumptions\u00a0of what\u2019s after that (from the technological to the philosophical). We\u00a0gathered in London on 5th<\/sup> May at the AI Summit to discuss these themes and more. Microsoft UK’s Chief Envisioning Officer, Dave Coplin, talked about the role of human and machine intelligence as we move towards the algorithmic business. And\u00a0although we\u00a0talked about this revolution in London, it wasn’t televised. It\u00a0was live.<\/p>\nFrom Passive to Active Participation<\/h2>\n
Who\u2019s in the Driver\u2019s Seat?<\/h2>\n
Return on Failure<\/h2>\n
Rise of the Humans<\/h1>\n