 Good evening everyone my name is Camille Cameron I'm the Dean at the Shirley School of Law at Dalhousie, and it's my great pleasure to welcome you all here this evening May I say go away? It's a mcmack greeting and I'd also like to say that we are holding this event tonight on unseeded Migma territory On behalf of all of my law school and university colleagues including our president professor Florezone who's joined us this evening It's my pleasure to welcome you all here to the second annual Sir Graham Day lecture in ethics morality and the law in addition to students faculty and members of the university community we have Quite a few people from our wider community as well, and it's great to be able to welcome you all here It's no surprise because the work of our guest professor Richard Susskind is well known in groundbreaking and so It comes as no surprise that there would be interest in and broad appeal More about professor Susskind in a moment First I'd like to just say a few words about this lecture series and about its namesake Saying a little bit about sir Graham Day is not really difficult because there's so much to say about his remarkable Accomplishments and in fact perhaps the only challenge is to what not to say and to choose What to exclude in such a brief introduction? So Graham has served as the chairman and CEO of Rover Group and British shipbuilders as the Chancellor of Dalhousie University He was knighted by Queen Elizabeth II. He is an officer of the Order of Canada He has honorary degrees from Canadian and British universities and perhaps equally important as all of those things He used to sing on sing along Jubilee This lecture series is possible because of his generosity and his vision and we thank him very much for that We're immensely proud of the collaboration between this lecture series and the CBC program ideas a little bit about that program the 1965 premiere broadcast of ideas was called the best ideas you'll hear tonight and the first few programs featured a discussion of Darwin's theory of evolution and Earl Bernie on poetry and creativity and back in 1965 the CBC Times announced that ideas is a serious a series prepared for people who just enjoy thinking Paul Kennedy who was with us this evening and you'll be hearing from him in a moment has been the host of ideas since 1999 but his association with the program goes back more than 30 years to about 1977 that year he made his first contribution to ideas with a documentary called the fur trade Revisited the project took him on a 1600 kilometer journey paddling down the Mackenzie River from Great Slave Lake to the Arctic Ocean quite an adventure his interests encompass the environment sport travel food music art and Biography and his work engages what he describes as the core curriculum of contemporary culture I won't list all of the awards. He's received. He's received quite a few We're delighted to have him here this evening And I'm now going to give the floor to him to introduce our special guest tonight professor Susskind Thank You Dean Cameron. It is always a pleasure to be in Halifax especially when the welcoming weather is put out for one It's it's really no great actually to be here Made reference. This is our 50th anniversary on radio in Canada, which I considered to be something of a miracle in fact And and great to be now partnered with the show of school of law in what I think is a really wonderful Lecture series that had its first edition only last year It's wonderful to do that because this really does in some ways dovetail mesh really carefully and well with the fundamental Ideas behind ideas and that is that we like to take discussion of important events by leading thinkers and make sure that they're They're well distributed in society at large and what we got last year with this lecture And what I'm sure you're going to get tonight is just an extension of that philosophy I only met Richard Susskind this afternoon and I can tell you it was a fascinating Conversation which I recorded it will be used in conjunction with this lecture and ultimately in a broadcast on ideas I should say I'll make very brief introduction because we're all here to hear hear him talk and He can introduce himself. I can tell you quite capably In his own words what I'll say though is that before the conversation began. I made perhaps the Psychological error of confessing that I'm a bit of a Luddite and As you all know, we're going in opposite directions this evening by by moving into the technological future I won't say that the conversation completely changed my mind But it did inspire me as I think that it will inspire you as you hear him talk about the the future of the professions I might have thought that one would end such a conversation in a gloomy very depressed Not quite happy futuristic look at what what might be staring us all in the face. I can tell you that I was Nearly convinced and certainly willing to listen more and more to what Richard Susskind might say about our future So ladies and gentlemen, please join me in welcoming the second Sir Graham Day lecturer Richard Susskind Thank you very much for that introduction. It is a great privilege to be here to be giving this lecture I've been looking forward to it for many months for a whole bundle of reasons not least to see some old friends in my visit To Halifax I want to thank everyone for putting on such dreadful weather for me It reminds me of my native Glasgow. This incidentally is a genuine Scottish accent. I'm not just pretending because I'm in Nova Scotia The idea is I'm going to speak to you for about 50 minutes or so And then there'll be an opportunity to ask and answer questions And that's always the trickiest part of the conversation for a speaker But I always take heart from Albert Einstein who had the right idea Albert Einstein you may know many years ago gave a wonderful lecture tour in England And he got very friendly with a chauffeur who used to drive him to and from the various places at which he was speaking And one day the chauffeur is driving along in the car and Einstein's sitting in the back of the car And the chauffeur looks in the rear view mirror and he says professor Einstein I really think you're a remarkable man What I find amazing is not the simply the sophistication of your theories clearly They are complex, but it's your ability to break these theories down into simple straightforward propositions That's what I find especially amazing in fact so clear are you and so many times have I heard you That I think I myself could give one of your talks So Einstein's sitting in the back of the car. I said well, it's funny You should say that because I'm getting rather bored of giving the same talk again and again Why don't we try and experiment? The people at the lecture today don't know what I look like so why don't we change roles? Just before we arrive at the lecture theater, I'll climb out of my clothes I'll get into your chauffeur's uniform and I'll stand at the back of the hall as you would normally do You can get into my clothes. We're about the same size. You can give the talk What do you think chauffeur thinks is the wonderful idea They do they arrive at the lecture theater Einstein climbs out of his clothes gets into the chauffeur's uniform Stands at the back of the hall as the chauffeur would normally have done and the chauffeur in Einstein's clothes stands at the front and gives this Unbelievably brilliant lecture and the crowds are going wild But of course what they hadn't banked on was question time and this And this very eminent physicist stands up and asks an impossibly difficult question The chauffeur's eyes glaze over but then they light up. He said that question's so easy I'm going to ask my chauffeur at the back of the room So Any difficult questions and I'm going to be throwing them back to the chauffeur's amongst you What I'd like to do this evening is cover seven topics I want to speak a little bit about the future what it holds and how to think about it Then offer some evidence that we gathered in the course of writing the book about which I'm speaking This is a book that I co-authored with my son Daniel I then want to put forward a model that we think summarizes how it is that the professions are evolving and a lot of that We'll find is underpinned by technology. So I want to give a view on what technology is doing in our world After that it takes me quite naturally I think to artificial intelligence and I want to speak about the prospects there and introduce a new couple of Ideas in that connection pen ultimately say well, what does all this mean for jobs and employment and finally ask a fundamental question About expertise that arises from our research So let's start with the future and in fact start with two futures Because in the book and based on the evidence we Gathered we identified Essentially two futures for the professions This is very much focusing on technology in the first future we says reassuringly familiar Because it's a version of what we have today it's where information technology the internet and so forth come along and Streamline and optimize the traditional way the professionals have always worked Perhaps a doctor seeing a patient on Skype a teacher using online resources for research and teaching Or maybe an architect using computer assisted design none of this fundamentally changes the way in which that professional works The second future however is very different The second future is when we use technology not to streamline and optimize our traditional professions But to transform the way in which we solve the problems that historically only the professions have been able to undertake So our idea here is not one of using technology By traditional lawyers by traditional doctors teachers and so forth. It's finding ways in which the services they deliver Might be put an offer particularly through the internet and often enabled through artificial intelligence in entirely new ways and That of course is very disconcerting But I want before I progress to provide some evidence and some thinking about this to perhaps Help you with your mindset So this is a story about black and Decker one of the world's leading manufacturers of power tools And apparently when they recruit new executives, they take them off in a course They sit them down in a room and they've put up a slide much like the slide you see before you and they say to the new Executives, this is what we sell isn't it and the new executives are rather disconcerted by this Of course, that's what we sell. We're black and Decker with the world's leading manufacturers of power tools The trainers with some satisfaction say well, that's not really what we sell because that's not really what our customers want This is what our customers want and it's your job to find ever more creative imaginative and competitive ways of giving our customers what they want and Here there's a fundamental lesson for professionals because most professionals when they think of the future Tend to think how can we make our power drill one-to-one consultative advisory service invariably? How can we make it a bit better quicker perhaps cheaper? Not often enough are we as professionals taking the fundamental step back and asking What is the hole in the wall metaphorically speaking in our world? Why is it that clients and patients and students come to us? What value is it that we ultimately bring? So I want to encourage the Professions when I say there might be fundamentally new in different ways of delivering the services in their fields To be more of hole in the wall or hole in the piece of wood mentality than power drill mentality And when we're thinking about the professions we wanted to take a step back and ask a fairly basic and fundamental question Why do we have the professions at all? What basic problem do our professions exist to solve? and As we fished about and thought about this in some detail it seemed to us that human beings one of our characteristics Indeed those of you are studying jurisprudence legal philosophy will know there's something that Herbert Hart pointed out the great legal philosopher He said human beings have limited understanding It's a feature of all human beings. We don't know enough as it were and What happens in a print-based industrial society when we're faced with problems or issues? beyond our fields of competence is That we tend to turn to people who have the expertise that can sort these problems out And we use the term practical expertise as a kind of catch-all phrase to cover knowledge expertise experience skills know-how But there's this practical expertise in the print-based industrial society that traditionally Professionals have brought to bear in helping us with our serious problems problems We can't sort out ourselves and they've done so under a grand bargain which essentially is an exclusivity We've glanted to the professions so that in many areas of life. It is only professionals Who are permitted to deliver the service and? Professionals thus have become a form of gatekeeper and I don't mean this in an unhealthy way necessarily But those are individuals who have expertise that others don't they have this monopoly over the expertise And so they become the gateway through which people have to travel when it is they need legal help medical help help with tax or whatever and That's the approach we've developed in the print-based industrial society and by and large the modern version of that's been in place since the middle of the 19th century But we're not in the print-based industrial society anymore We've moved to a technology-based internet society and let's be honest our professions are creaking If you look at our legal health our legal services our health service our educational services even in all advanced economies We few we find widespread Satisfaction So often they're unaffordable if you take legal services, and I'll use law a lot given the context of this evening's lecture It's an unfortunate reality in all of us all jurisdictions now That most consumers can't actually afford to instruct lawyers if they have a dispute Our working methods and so many of our professions are antiquated one only needs to visit most courts to see the files the ledgers the papers They're also opaque as well very hard for a non expert for the recipient of a service actually to know what's going on Or indeed to know whether or not the service and offer is a good service And the professions under form perform And I'm not suggesting here professionals are substandard, but I'm meaning that It is a reality that the very finest expertise Of the top notch experts in our world is available only to a very privileged few Because of the way we make their expertise available And so our challenge in thinking about this book this question in our book the future of the professions was Can we solve the problem differently? Might there be a different way of making practical expertise available in society? Do we really need The old gatekeepers And so one of the things we thought we would do is have a look at the evidence And daniel my son went around the world speaking to thought leaders and market leaders About what was happening at the leading edge of the professions And we found some remarkable developments in education for example in harvard We found that more people in one year Signed up for their online courses Then had graduated from the university in its entire 377 years of existence We found in can academy the most marvelous online instructional videos and tutorials and maths And daniel who teaches maths at outford will say this is something he directs his students towards It's absolutely first rate 10 million unique users Every month that's about the same effective population school population of england and wales In medicine if you look at something like web md and network of medical websites Each month over 190 million americans not all of them hypochondriacs go online Because of some kind of medical difficulty that's far more people of course than consult medical doctors themselves The the fda predicted by 2018 One and a half billion people will be actively using at least one medical app on their smartphones In journalism we look to the new york times in the huffington post After six years of the huffington post existence its websites getting more hits than the new york times 164 years old or bleacher report in sport 2000 bloggers coming together 22 million users Viewers it's a strong competitor now to cnn sport or associated press created some algorithms to generate financial reports 15 times more productive than human analysts for these earnings reports in law ebay Every year in ebay. There are 60 million disputes 60 million disagreements amongst traders How many of them are sorted out by lawyers in the courts hardly any they're sorted out by online dispute resolution The best known legal brand in the united states by a long margin the most legally active country in the world The best known legal brand is no longer a law firm. It's legal zoom.com online documents and advice for small businesses and consumers in tax 48 million people in 2014 Set aside their tax advisors and use turbo tax or similar software to file their tax returns even HM revenue is using technology It is a system to detect fraud that manages over a billion pieces of data Apparently about the same amount of data as has held in the british library Which has every book that has ever been published in the united kingdom in architecture and engineering remarkable developments too Autonomous robots actually building structures. There's a structure assembled autonomously by robots over 1500 bricks What about printing out a building? Printed building parts up to 3.5 meters in size Being put together and this is the idea and it's remarkable of 3d printing being used in the construction and engineering sector In consulting eccentric a very well known consulting firm But they don't just engage in management consulting anymore. For example, they have 750 hospital nurses on their books Or what about Deloitte founded 170 years ago as an accounting practice It now has over 200 000 people And it has its own university Deloitte university in a 700 000 square foot campus in texas The professions are changing before our very eyes. Even the clergy is affected. This is second life You know, we're about a million people have their own atavars and this is an anglican rather gloomy anglican cathedral it offers daily prayer services weekly bible classes on an island that the um Cathedrals located on an island called epiphany These are people not playing and they are taking their religion seriously, but practicing in an entirely different way But our all-time favorite is an app you can get for your phone called confession And unbelievably The vatican granted this a digital its first digital imprimatur as the official license that grants to religious texts It is tools for tracking sin at And it has a drop-down menu option for contrition. So The question We therefore ask is what's happening? And these are just a glimpse of some of the things we saw at the leading edge This is a far cry from the one-to-one consultative advisor that we think of when we Imagine a doctor a lawyer a tax advisor an auditor and one way we have of looking at this is saying That what's happening is there's an evolutionary process at play And the first stage of evolution is one when one regards and practices a professional as a form of craft I sometimes call this the bespoke service a highly tailored customize one-to-one service and many of us Have a rather romantic view that that's what the professions are all about And we think of this as a lot in law I think when you study law you get the sense from law school where it starts That almost every case could reach the supreme court and in our literature and theater people are looking for smoking guns and loopholes But the reality is in legal practice as frankly in all other professions. There's now a high degree of standardization management consultants use their methodologies Doctors use protocols and checklists Lawyers use precedents teachers use last year's notes We don't reinvent the wheel. We don't Paint on a new canvas every day and we sometimes go further. We systematize rather than having checklists We have workflow systems in law. We've automated document generation systems So these are the kinds of techniques within the profession and within professional institutions that we've developed We've moved from hand crafting what still happens and sometimes creativity and Imagination and strategic insights required but very often in daily practice. It's become highly standardized and systematized This by and large is a good thing But this is within the profession But at this stage in the evolution we reach what we think is a vital stage and we call this externalization And this is where the content the expertise the guidance the materials can be made available online And it can be made available online in three entirely different ways The first way particularly for commercially oriented organizations very attractive you charge for online services So one of the finest law firms in the world allen and ovary They now generate about 15 million pounds every year from their online legal services rather than consulting with and paying for a lawyer You pay for a system that's available online. Deloitte have been doing the same in tax for many years A second option is that the system is made available online on a non-chargeable basis and this is characteristic of very many voluntary sector charitable organizations Websites as well as many government websites the third intrigues us and attracts us the most And this is where content is made available on a commons basis And we're all familiar with that in the shape of wikipedia or open source software If one thinks of wikipedia, it's not under controlled by a corporate or by a government or a charitable body It's as it were held in a commons and we can each draw from and contribute to it ourselves So this is a rough map and you can call this the commoditization of professional service a move from this notion Of professional service as a kind of craft Through the notion of standardizing then systematizing make it any available online Sometimes chargeable sometimes non-chargeable and even on a commons basis Now this transition this move this evolution Is driven and enabled by technology? And when thinking about technology, I'd like to take you back to 1996 And I wrote a book then called the future of law And in the future of law strangely enough because it's 2016. I was making a 20 year prediction about high it Would transform the work of lawyers and the administration of justice And I know this sounds ridiculous in retrospect But one of the things I banged on about a lot at the time Was electronic communication email and I was giving lots of lectures about the book and so forth And I was saying that the dominant way that lawyers and clients would come to communicate in the future Would be by email Okay, so that sounds pretty self-evident today I kid you not that the law society of England and Wales said I shouldn't be allowed to speak in public They said I didn't understand confidentiality and security They said I was bringing the legal profession into disrepute By suggesting email would be used by clients and lawyers and this at the time was a very serious issue for me We've come such a long way as we all know but I say this today not to well Maybe just a little bit to vindicate what I said but more to Say to you I'm going to be talking about a number of technologies over the next few minutes And your inclination of seasoned professionals in the room will be to think well that doesn't really apply to us In fact, as I always say one of the characteristics of professionals is that they see that all these changes apply to other professions rather than their own But what I'm saying to you today is that all these technologies are way more predictable than email was for lawyers in 1996 There's four characteristics aspects of technology that we think you need to have in mind if you're trying to understand its impact on the professions The first is that the underpinning technology is growing at an exponential rate. I'm going to say a little bit about each Second and this is one of our catchphrases as it were. We believe our systems and machines are becoming Increasingly capable able to do more and more Thirdly the actual technology is becoming increasingly pervasive in ways I'll explain And finally as human beings we're becoming increasingly connected So let's talk about the exponential growth and think about a law and it's not a law of the land It's Moore's law Gordon Moore the man who 51 years ago in 1965 Said very approximately and there's debate about Exactly what he does said, but I've read the technical paper and basically he's saying that every two years processing power of computers would double And people at the time said well the latter last two or three years And by and large it's still going strong 51 years later and people will say today Well, you know, there's only there's physical limits to this depend It's all down to the limits of the nature and substance that is silicon But actually if you speak to people who are involved in material science and computer science The imperatives to keep that doubling every year going are very very strong indeed and you'll hear for specialists This is going strong for decades yet And it gets rise to what is not quite but an almost exponential curve an explosive growth in the processing power of computers And while we might feel as computer users remarkable things have happened and indeed they have over the last few years Our and other people's thesis is actually we're still we're still really at the shallow part of the curve perhaps the knee of the curve and what we're about to experience Over the next few decades Is almost unimaginable now to give a sense of The impact of exponential growth consider this imagine a sheet of paper A4 0.06 millimeters in thickness imagine folding it over once Folding it over a second time forwarding it over a third time in the fourth time after four folds It's the same thickness as a credit card After 11 folds It's the same height as a coke can After 21 folds. It's the same height as big ben After 43 folds It's the same distances from here to the moon and after 100 folds It's 8 billion light years in thickness now. This is the maths of doubling again and again You may have heard the story of the princess and the tramp the tramp saves the princess's life The king says to the tramp. I will give you anything in the kingdom by way of reward The tramp said I'd like something very simple I'd like a chessboard and the first square I'd like you to put one grain of rice On the second square double it and the third square double again and keep doubling around the board And that rice I would like is my reward and the king says you have it But you can't because that's more rice than exists in planet air This is what happens with this doubling up again and again. So does this mean in practical terms? It means by 2020 The average desktop machine Will have about the same processing power as the human brain It'll be able to calculate about 10 to the 16th 10 to the 17th calculations per second Not saying the machines will be artificially intelligent or of cognitive states Just giving a sense of the processing power and how far we've come remarkable really not nearly as remarkable as this By 2050 If you keep doubling and processing power every year the average desktop machine will have more processing power than All of humanity put together Now this sounds science fictional as does this doubling up in the many billions of light years and so forth But I want to put it rather differently to you That we are I think privileged as human beings to be living in a time of greater and more rapid technological growth Than humanity has ever seen It just happens to be going on in our watch And although we don't appreciate this rapid change it is quite phenomenal And you can think as professionals or aspiring professionals that very interesting, but it doesn't apply to me or you can think wow I want to be part of this. I want to see how we can actually Exploit this technology in the fields to which i'm committed. You probably think well he's exaggerating a bit Well, here's the Nobel prize lecture of michael spence about prize winner economics Who said and this was 25 years ago? He noted in the first 50 years of the computer age roughly a 10 billion times reduction in the cost of processing power It's the same kind of point and remember every two years that doubles 20 billion 40 billion 80 billion 160 billion And away you go and in not many years. It's just unimaginable figures. We're speaking about it's not just processing power It's data as well Note what garrick schmidt google's chairman said every two days now We create as much information as we did from the dawn of civilization up until 2003 By about 2020 that'll be about every couple of hours. We'll be creating that amount of information More prosaically think of the memory card you might have in your camera in 2005 a good card 128 megabytes In 2014 120 gigabytes more than a thousand fold increase in less than 10 years That's more than doubling every year. You can hold all the world's recorded music in a little box that can fit in your hand So that's the backdrop. That's the underpinning technology folks It's only going in one direction. It's dramatic. It's explosive whether or not it's strictly exponential is almost of no moment Because what it's doing is it's underpinning and enabling the development of machines that in our language are increasingly capable And we look at that under four different headings. What's happening within the field of big data? How it is that computers are now solving problems? the emergence of a new field known as affective computing and also Quite remarkable developments and robotics So big data is very interesting that one of the byproducts of the use of the internet Sometimes people call it the data exhaust But it's all the data that we create as we use the internet And actually that data reflects and embodies great experience And that data can yield insights patterns correlations. So look at lex machina A system developed in the united states that can predict more accurately than most human lawyers the outcome of patent disputes It answers the question that all clients want answered, which is this What are my chances of winning? It knows nothing about the law It makes a statistical prediction on the basis of the past behavior of the courts Who the judges were what the court was which parties were involved which law firms were involved the individual lawyers and so forth And so it is in other areas too I've heard senior doctors say that data science will trump medical science. We will be able to make greater better predictions Based on the data we've gathered then on the substantive knowledge of the traditional professions But it's not just big data. It's problem solving in question and answer systems. Think of watson by ibm Most of you will know that in 2011 five years ago now Watson A computer system appeared Live on a tv quiz show called jeopardy And beat the two best ever human jeopardy champions Think about this folks This is a computer system answering questions On anything in the world More accurately and more quickly than any human being this isn't science fiction This is five years ago now ibm did not invest in this because they want to be good at quiz shows They're of course investing in other areas. So in medicine for example The feedback i'm getting about the level of diagnostic performance the quality of treatment plan This is a serious use of ai technology A number of major law firms around the world are collaborating with watson In the idea of using that kind of technology in law And what fascinates me in law and i'll Quite happily make a prediction. It seems to me for everyday law and everyday consumers Probably by the late 20s Most people who have a legal problem won't consult a lawyer They will be able to sit down and speak with a computer tell them the kind of problem they have interact with it And the system will offer them a fair answer You don't say well that's unlikely But a system has already been in place for five years that can answer questions on anything in the world more accurately than the best human beings The technologies are being developed at a remarkable rate But here's another element of this this era called affective computing Machines that can both detect And express human emotions A machine can look at your face and tell whether or not you're happy surprised angry or disgusted There are now systems more accurately than any human beings can look at a human smile And tell whether or not that smile is fake or genuine There are now systems more accurately than any human beings can hear two female voices And tell whether or not they belong to a mother and a daughter and so Bizarre as though it may seem Our systems will both be able to detect emotions. They say by the early 20s Our iPhones will know what kind of mood we're in and respond accordingly Your jacket will give you a little hug When a nice message comes in So our systems will not only recognize but somehow start expressing the kinds of Behaviors that we normally associate with human emotion And all of this I suppose is leading to this as well the notion of androids This is professor ishi guru in my view the world's leading academic roboticist. He's with his android of himself It's quite hard to tell which is him and which is his android And all of these technologies are just getting more and more sophisticated a little word about robotics it's relevant for some professions for Surgery and dentistry for veterinary science architecture, perhaps less so for law But it's very interesting in 2004 two of the world's leading economists who specialize On the impact of technology on work Wrote a very interesting book called the new division of labour where they were trying to say what kinds of work Can machines do and what kind of work will really only be Undertaken by human beings in the future and they made the point at one stage saying well You can't really imagine a computer for example cutting your hair You can't imagine a computer doing your garden You can't imagine they said the leading specialists thinking about the future of work in 2004 You can't imagine a computer driving a car So not that many years later. We've got autonomous vehicles. I believe it's 10 or 15 years time We'll be looking back and thinking it's amazing. We used to drive cars. Why would you want to do that? Why do you want to why would you want to sit at a wheel with a gear stick? Why would you do that? We'll have chauffeur effectively chauffeur driven cars for us But this is just to see how quickly the technology moves on and how wrong the specialists can be And as human beings we're increasingly connected and it's not just this kind of technology Which is offering telepresence three-dimensional holographic telepresence or whatever What we found in our research is that social networks are extending into the professions And so that there are close social networks Ceremonial for doctors a couple of hundred thousand doctors in america Architects of architect and edmodos are similar kind of facility in education legal on rampant law And so the idea of social networking which for many professionals They think is an indulgence of their children actually is becoming a serious mechanism for collaboration cooperation communication amongst professionals And then there's a vital new form of social network for the professions This is the kind of system where the recipients of service the patients the clients the students come on and share their experiences And so patients like me over a third of a million people Particularly people suffering from chronic illnesses. They can come online And they can receive comfort insight often tricks of the trade They can learn from other patients what worked for them This is substantive help not delivered by doctors or nurses Complementary to often often complimentary to the work of doctors and nurses, but valuable on its own right And people actually welcome and often find more palatable advice and guidance offered by their peers rather than by specialists So there you are when you combine these elements of technology the growth Exponential growth of the underpinning technologies the reality that our machines are becoming increasingly capable I didn't say much about machines becoming increasingly pervasive, but just think we know of our handhelds There are already six billion smartphone Accounts out there, but it goes way beyond that the internet of things this idea that our physical objects are clothing Will all be will have chips embedded within them and will be connected to the internet And of course as we as human beings are increasingly connected through social networks and the like And I suppose the biggest message of the night, however, is that there's no finishing line Somehow you think oh well, that's rather amazing And we particularly people we're going to have over 30 we think well You know it's all going to settle down a bit It's not settling down at all the technologies that will probably transform our lives by 2025 almost certainly haven't been invented yet So this is why it's a remarkable time But let me move into the realm of artificial intelligence, which I think is one Refield that follows quite naturally I wrote my doctors in oxford in artificial intelligence and law in the in the 80s And I worked on what I now regard in retrospect and we call them the book our first wave of ai In 1986 after I finished my doctorate, I got involved with the development of the system I think you can understand the challenge we faced by this extract from a piece of legislation section two of this act Shall not apply to an action to which this section applies. This is serious. This is a piece. This is the law of the land That is binding in all of us and this is one of the more intelligible pieces of that particular instrument So it came from a piece of legislation called the latent damage at 1986 One of my former examiners was the dean of the law school in oxford at the time philip kapper He came to me and said i've just written a book in this new very difficult piece of legislation. No one understands it But I think we could develop an expert system in this area and expert systems were The branch of artificial intelligence at the time if people were focusing on what you did was I was the knowledge engineer you sat down with the human expert you mined the jewels from their heads You developed a model or representation essentially a decision tree of their understanding of the area of expertise and you allow put it into a system so others could wander through it so we developed this system called latent damage system now I want to say just in case you think ill of me that at the time this was state-of-the-art design we were I joke not we really thought this was as cool as could be and now now it's just nothing sure as shameful but I thought you'd want to see it so it says press any t to continue just in case you wondered So essentially we developed this model this decision to a representation of philip's expertise And this ladies and gentlemen was in the days where floppy disks genuinely were floppy I don't I hear someone recognizing the configuration five and a quarter inch floppy disks We're actually no joke. We distributed a pilot version of it and a qc in england put it into brand new cd machine misunderstood the technology Another joke from that time when we developed the system as I say we put it out for piloting I'd forgotten this actually just sprung to mind and we formed up It was a chap who then later became a judge and we said what do you think of the system? He said it's really good, but the printing the printing facilities rather heavy Rather heavy. What do you mean? He said yeah lifting the screen onto the photocopier in all seriousness so We challenged him on this We challenged him in this and he said what do you expect us to lift the photocopier up to the screen? So as a kind of question and answer system as a whole series of questions They were actually over 2 million passed through the system And in the end this system Reduced the research time in this area of law from about two hours to two minutes phenomenal really Although I say so myself and also what philip would gladly still say is that it became better than him Because it what he worked on every area of it and this representation quite often we're testing it He said no, it's got it wrong. I said no, I've got it wrong I can see why and this was just an amazing experience as we developed the system And looking back in 1988 when we discussed this we thought this would be transformational Of course in the legal system, but these systems have not really been Very much in play. There are examples And at the time there's other people working in law and medicine and tax and audit consulting But why they didn't come into or Didn't really get accepted or use very widely. We believe it For three reasons. Firstly, they're quite costly to develop Secondly, this was at a time when many professionals particular lawyers weren't under the same cost pressures as they're under today And so if you were charging by the hour, there wasn't any obvious incentive to reducing your research time from two hours to two minutes But actually what I believe killed this first wave Are the expert systems ai in the professions was the web The web came along and suddenly we had this service that allowed us Rapidly not nearly sophisticated a way But rapidly to put content and guidance and documents available online and everyone else was online And I have to say we were all seduced by this and they're like, what can we do in the web? And we actually developed technically inferior systems, but they actually had huge impact Now the big turning point in all of this story I think was 1997 when gary kasparov the world chess champion was beaten by a computer called deep blue Now in the 80s when I was this stray lawyer working in the oxford university computing laboratory We used to have a discussion about whether or not systems could ever beat The best chess players And I remember them so well and our Definite answer at the time was yes. Sorry. It was no Small in the story the answer was no and here's why we thought no And it's terribly significant for the future of systems At the time the only way we thought you could develop these systems Was by sitting down with an expert reducing their expertise to a decision tree to a body of rules and making that available to non experts But here was the difficulty particularly in medicine and certainly in chess playing The best experts don't know how they come to their decisions. If you say, how did you know what moved to me? Where did it come from? Where did that? insight come from As in so many areas professions. I don't know it's gut reaction. It's intuition. It's knee jerk. It's experience So if we couldn't articulate it, we certainly couldn't make it available in a system So our joint view very definitely was we can get systems to perform at quite a good level Where we can represent humans knowledge The knowledge they can express but there was this tacit knowledge is sometimes called that was somehow Unarticulatable and therefore beyond us and therefore computer systems could never outperform the best experts When Kasparov was beaten by deep blue What we hadn't really banked on was the exponential increase in processing power because deep blue in one second could explore over 330 million possible moves Even the best human experts can only juggle about 100 moves in their head at one time Kasparov wasn't in a sense wasn't even really beaten at the same game This system didn't have strategic insight or creativity or genius. It was brute force computing power and lots of data But the lesson from this is vital and it's best expressed. I think by Patrick Winston One of the fathers of AI who says there are lots of ways of being smart that aren't smart like us We'd assume the only way you could get to beat the chest system Sorry to beat the best expert was to have a chest system that somehow mimic human experts The only way you could really do clever stuff is by doing it in the human way was what we unbelievably arrogantly thought And this goes rise to what we believe now and we call is the AI fallacy And it's one of the central themes of our book And once stated it's somewhat quite obvious But we read it everywhere even the best academics and the best journalists And it's the mistaken assumption That the only way to develop systems that perform tasks at the level experts are higher Is somehow to replicate the thinking processes of human specialists That's the only way to do it and it's fundamentally mistaken. Let's take an example judgment Most professionals will say to me, I see what you say Richard, but these systems only get to a certain level Because what I do as an expert Is I have creativity. I Maybe express empathy or I exercise judgment. So let's just look at judgment How on earth The challenger will say to me how on earth could a computer system exercise judgment? And we say that's the wrong question This is the question you should be asking to what problem is judgment the solution Why is it we call upon our human experts to exercise judgment? Why do we need their judgment? And the answer we believe is this Is because we live in a world of uncertainty Often the facts the problem is uncertain And the knowledge required to sort out the problem is also uncertain And we go to experts because they've got experience. They've done lots of this stuff before there may not be a definitive answer But based on your experience based on your judgment What's the answer here? So what it is that human beings actually do is they handle uncertainty and the way they handle uncertainty is by exercising what we call judgment So the real question is can computers handle uncertainty better than human beings Not can computers exercise judgment. We know the answer to that's probably no It's the same as can computers Express empathy the answer that knows because they are distinctively human characteristics. It's a cheat question You're defining the question out of existence So the big question of empathy too is not can a computer exercise or emit empathy It's to what problem is empathy the solution? Why is it we want from our experts this phenomenon known as empathy? Why is it we need this degree of comfort and certainty? And so that's the way we need to think about the future not There's some really tricky stuff that by definition machines can't do because they're distinctively human capacities, but what Solutions are they offering professionals when they exercise these faculties? So when you ask the question about handling uncertainty think again about big data Think about a system that maybe has in medicine or in law 10 million past examples And maybe 50 or 60 variables attached to each boy can machines handle uncertainty. That's what they're good at Another question people ask which we love for people who are interested in philosophy But it's actually irrelevant for current purposes is can machines think It's a little it's wrapped up with this question Well if a machine can't think it can't exercise judgment and again we say that's the wrong way of looking at it That's assuming that the only way to solve the problems which judgment is a solution Is actually by having a thinking cognitive state and we just don't believe that think about this This is wonderful. IBM as we know developed what's in the be The best of our jeopardy champions in 2011 and the next day A man called John Searle who's a Berkeley philosopher who gave the reef lectures in the early 80s in in London and I don't know if Andrew my friend here was at that lecture, but it was he talked about AI AI And the philosophical aspects of AI and wonderful philosopher But what he said in an op-ed the next day, and I just think this is wonderful. He said Watson doesn't know that it won in jeopardy Now it's just it's just perfect, isn't it? Watson didn't want to go down to the pub afterwards to celebrate It didn't want to tell people how it felt and how it was a great relief Because what you have here is the evolution of what we call increasingly capable non-thinking machines These machines don't need to think to outperform us. This is the big mistake You read it in all even top-notch journalists go on about this. Well, I can't think and therefore there's a limit there No, these machines are outperforming us because they have brute force processing power phenomenal access to data And that's the second wave of AI ladies and gentlemen. That's the one that's going to change our world And just to give you a sense of how quickly all of this is moving Some of you might have read and this is just about a fortnight old the news about a system called alpha go And alpha goes a system developed by deep mind, which is owned by google And it plays the game go now the game go apparently has more permutations than there are atoms in the universe So it's of different many different many orders of magnitude more complex than chess And it wasn't many months ago that leading experts were saying it will be 10 years Before a computer system will be able to beat the world champion at go Well, it happened two weeks ago They played each other and the machine one four to one four matches to one It uses a form of neural networks and a kind of deep learning the interesting thing about it is once it learns the game It plays with itself and it gets better and better And so it's off a way of refining and improving with that This is not some sense that you have to program the computer to anticipate every move Yet it actually using these neural network and deep learning techniques. It gets better itself This is called machine learning. So what does all this mean for jobs? Two questions we're often asked first of all, will there be any jobs left for professionals? And secondly, what could humans actually do that machines can't and I'm just going to skirt over these but in the medium run We think it's about redeployment rather than unemployment We don't think if you're in the room as a young lawyer aspiring lawyer a young doctor There will be nothing to do in the 20s But actually it may not look much like what it is the doctors and lawyers are doing today Here are the kinds of Roles that tomorrow's professionals will play And it's unsettling for many professionals because that's not what we went to law school or medical school to be And often people won't even know what they are. What is a knowledge engineer? an empathizer for good and the sake and We explain these in in detail in our book But my point now is not to describe the roles It's to suggest that what will happen is there'll be new jobs And a lot of these jobs will be focused on the development of the systems that will in one way or another be replacing the human beings In the long run though, and many of you may be well ahead. You may think well actually Why will these roles still exist? our book was reviewed by The economist I mentioned it obviously because it was a very favorable review But it also had this quite funny cartoon and the book there the review was headed professor dr. Robot qc So look at this robot with its wig and its ledger and its stethoscope and the rest of it and many people when they think of The future of work and technology They have professor dr. Robot qc or one of his relatives in mind They think one day they're going to walk into the office and sitting in the chair will be this guy And you'll have been displaced. It's probably not going to unpack like that for a whole bundle of reasons But our thesis about the long run in very short order is this I've said it many times this evening machines It seems to us will become increasingly capable and that doesn't seem to be slowing down This means they'll take on more and more Of the tasks that today are undertaken by human professionals Now many of you'll say but surely new tasks will emerge. That's always what happens when there's new technology We accept that but it's likely machines will take these on as well It's not at all obvious to us that there's anything about the new tasks that are emerging Which means they're less amenable in the long run to technology So we find it hard to avoid the conclusion That there will be a steady decline in the need for human professionals in the long run I'm not seeing over the next few years. I'm not even seeing through the 20s But as one moves into 30s and 40s if you accept the broad thesis It seems hard to avoid that conclusion Now from the point of view of the professional, of course This is deeply threatening but I want to say again from the point of view of the recipient of services We've a global problem about how we Distribute and make available medical expertise How we ensure that people all over the world can understand and enforce their entitlements How people around the world can be educated at the level that we're fortunate enough to be educated in And here is the glimmer it seems to us of an answer And all of this hinges on this notion of practical expertise And we started off writing a book that we thought was really asking the question. What's the future of the professions? But actually we asked an answer a different question How do you produce and distribute practical expertise? That's the real question For to ask the question. What's the future of the professions assumes that the professions have a future And we didn't want to say by definition that that was a possibility But the bigger question is Can the work that professionals currently do Be done in different ways And the work it seems to us that professionals currently do Is they produce and distribute practical expertise? So the answer to this question in the print-based industrial society has been this the traditional model how we Across all societies and economies have sorted out our difficult problems Is by putting professionals and firms and schools and hospitals and so forth And what we identify in the book and in closing I just want to Give a very quick flavor of each is six alternative models It needn't be that way There are other ways that we can solve the problems to which professionals historically have been the only possible solution The network experts model the least forbidding one in a way Economists call it workers and tap still experts, but actually not working in the traditional institutions and organizations This is freelance specialists a great example in law is axiom Lawyers who don't want to go to be partners in law firms or in-house counsel You can work freelance and be made available Through this very high powered essentially agency, but it's putting a structure around people who are professionals but who can progress in their lives Not in standard institutions and organizations, but by delivering their services in new ways Then there's the paraprofessional model the classic thought is well what we have in the professions is deep experts at the top and Less expert people at the bottom of this triangle And the thesis as one can normally understand is that the people really at threat are those at the bottom of the triangle Because that's where the routine repetitive work goes on and to some extent that is our thesis But there is another possibility where actually the experts at the top are displaced So if you think of the work that whatson ibm is doing in medicine A very conceivable model Is that in the future you have nurse practitioners? paramedics Full of empathy great interpersonal skills sitting down with Diagnostic systems or treatment planning systems rather than expert doctors And you can see that right across the professions And then there's the knowledge engineering model This is what I really did in the mid 80s when I essentially restructured a human expert's knowledge and made it available to others So they could perform at his level and there's a lot of that still about diagnostic medical systems and Tax systems audit systems and so forth Then there's the communities of experience model, which I mentioned earlier. You saw this slide This is the idea that another way in which you can sort out your problems is not by going to experts But by going to communities of people Who have sorted out problems for themselves or indeed often here where experts have left guidance and this is one of the jobs We anticipate what we call the moderator where there will be experts helping out and clarify I don't use A pc anymore, but when I did and used to use ridiculous windows messages What I used to do and everyone does is you cut them you paste them into google and out there somewhere There's someone who tells you what the problem is It's no different really if you're using turbo tax and you quite understand a question You don't then phone up a tax advisor You go to turbo tax answer exchange and other people who've used these systems will help you out Then there's the embedded knowledge model And I like this in law. Look at this. This is a screenshot from a handheld and we all play this game When I was a kid We used to play this game with atoms. We call them playing cards and we call the game patients Now those of you who are of my age and stage Will remember playing patients It's now called solitaire with cards. You also remember If you wanted to put a red five under a red six You could do that. It's called cheating, but uh You could do it. I don't know why you'd want to do it, but you could do it Now what happens today in solitaire electronically when you do it? The system flicks the card back. You can't do it. Why? The rules are embedded in the system And in law, this is going to be how much of law is administered It won't be a question of what's the law should I comply? In major organizations who are automating say the banks and manufacturing organizations production companies Who are automating their workflows and all their systems the idea that somehow there's going to be a human lawyer in the loop Giving a little bit of legal advice when it's required is fanciful We'll embody the law required and embed that knowledge Into our systems and we'll do the same in tax and we'll do the same in audit And we'll certainly be doing the same in medicine and finally and most disconcertingly and I think the biggest challenge of all But only one of the six models is the machine-generated model where computer systems themselves produce effectively new knowledge and expertise And this is intellectually one of the greatest challenge for us to get our heads around that a system can actually identify patterns and regulations and Correlations rather that we as human beings can't see or that a system can solve problems better than human beings So these are the six different models just a couple of final thoughts and then we can have some question and answer As an experienced professional and many of you are in this room You might be thinking I hope I can hold out to retirement before any of this stuff comes along I understand that view But I prefer to look at this as a great opportunity and I go back to the earlier comment Just think about it not from the point of view of the profession to which you're committed but actually Where we happen to be in life. We happen to be born at this time of remarkable technological development What an amazing opportunity To help As when people say to me, I want to be a lawyer. Should I be in light of what you say? I say to the future of law. It's not john grisham. It's not rumpel of the bailey It's not suits. It's not the good wife all of which I love incidentally, but that's 20th century lawyering You should go into the law not because you want to practice law like these guys Or let your mum or dad or your uncle Or frankly, let most of your professors will tell you legal work is undertaken You should go into law because you're really interested for example in access to justice And in the knowledge now that there are techniques emerging that will fundamentally change the way in which we can offer access to justice In law you should go into medicine Not because you want to be a primary care specialist of the sort we have today But because you are fascinated by and want to be involved in the development of new techniques For helping improve the health of your community So this is exciting It might be threatening for the traditional professional who's not prepared to change their ways But for those of you who are coming into the profession I wish I was you. I wish I could play a role in reinventing the professions It's gonna be hard work. I love this by Edison opportunities Missed by most people because it's dressed in overalls and looks like work people often don't think well There's some magic in all of this but a lot of this is about hard graft It's grind if you look at all the tech startups gosh do these people work hard So you can't have what most lawyers want from me, which what I call off-the-shelf competitive advantage Usually get a question. What should I do in monday trying to be pragmatic and practical as if there's an answer to this I don't know what you do All I do is lay out the buffet. I see the whole bundle of things that are happening You're the lawyer. You're the person who's running the business. It's up to you now The future is very much in your hands. And that's why I love to finish my talks for this quotation For a silicon valley guy called Alan Kay who says the best way to predict the future is to invent it The question is not Richard was the future look like the question is If you're an old professional, it's what legacy am I going to leave? I'm not predicting the future. I'm going to create the future and have a young professional What part am I going to play in redefining the way we sort out the problems? Which historically could only be handled by our traditional professionals. Thank you As I mentioned earlier, Richard has agreed to answer questions from the floor We have I think about 15 or 20 minutes to accommodate them Because there's such a short period of time Can I ask everyone to keep their questions as brief as possible? And there's a couple microphones set up in the aisles To get you a chance to get to those questions. I'll I'll prove that it's a journalist. I was using the new technology only this morning I sent an email to Richard sun Co-author of the book Daniel Suscant and I asked him what questions I should ask his dad And his first question was what's it like to work with your dad on a book? But the more serious question and the one I think that we should perhaps Get an answer to from Richard this evening Daniel's question was Are there more? Are there moral as well as technical limits to machines? Your answer Yes, that's a fascinating question and we agonized over that and the point is this that As these machines become increasingly capable and they outperform us I think we can ask the question whether or not there are certain tasks activities That we think machines ought not to be allowed to do We might want to draw a line and actually we might want to draw that line quite quickly as machines are becoming increasingly capable Rather rapidly So I feel and you may or may not agree That when it comes to the decision to turn off a life support system and actually to flick that switch I think I want a human being to be doing that I don't want the buck to stop at a robot. I think there's an element of moral responsibility there Now others may say there may come the stage where these machines have just such greatly superior understanding of the permutations and possibilities That actually is better to leave it to the machines but I give an analogy also or an analogous Thought in the legal world. Do we really want machines say passing life sentences? So our sense is there might be moral boundaries to these machines And what we call for in the book is public debate on that very question We are no better place than anyone else to draw the line But we think there ought to be public debate on Whether that line should be drawn and if one is should be drawn where it should be drawn This for us is as fundamental as the debate in the early 80s about in vitro fertilization and test tube babies Where in england we had a public debate learned by a well-known philosopher called mary warwick and the country came to a view We might not be able to come to a view, but we should have public debate on the issue Do you want me to answer the question? What's it like to work with that? Sure. What is it like to work with that? Well, we're both impossible, so it's very difficult Is there a question there? Yes It seems like the capacities that these technologies are affording us Are unimaginable power compared to what we've had in the past And it seems to me that when Power becomes available. It can be used either for liberation or for exploitation To inject a little bit of moral content into the conversation of what we're going How can you be sure or or do you have worries that there Is going to be an exploitative capacity to these technologies as opposed to a liberative capacity and if so Is that perhaps where The human being can be maintained in a position of while moral expertise for lack of a better word Well, it may be that we The distinctive role For the human being is in the normative the ethical Realm but we asked the question I think first of all she says very good question We asked the question in our book, which I think flows from what you say who should own and control The practical expertise Because what we're doing is creating these very powerful systems and we're gathering amazingly powerful content And it worries us a bit. I have to say that there'll be new gatekeepers Because traditionally the answer to that has been the relatively benevolent professionals But you'll remember maybe from my diagram about the evolution of professional service that there's three ways in which The content can be externalized. It can be available commercially on a chargeable basis It can be available Non-commercially often by government bodies or in a commons basis We worry about the first we worry about new gatekeepers We worry about this if you take something that medical knowledge and expertise or legal knowledge and expertise We want that to be made as widely available across the world If this is now held for example by Major data companies or major internet companies We may have new gatekeepers whose goals and roles and values Are actually not aligned with the traditional professions Or even with governments, even if one doesn't believe in an evil state Most of us I suspect fear a grotesquely inefficient state So do we really want the custodians and the managers of that content in these systems to be The state this is why we're inclined to the commons model and We discussed this earlier. I can be accused of being idealistic, but I'm simply being prescriptive I'm saying insofar as we can I think The best way To ensure that that material is made available for the common good is to hold it For common purpose in that way. So you're absolutely right and to raise the question and It's an issue that we agonized over and it also raises some very interesting economic questions as well As fewer and fewer people are employed and essentially In economic terms the value is As your hand suggests It's essentially capital Who actually owns and controls that capital and vital questions. Thank you. Thank you You haven't said very much about who does the programming of these machines that are going to do all these wonderful things And I'm prompted to ask the question because of something related to watson And I mean, yes watson beat two human beings in jeopardy But I remember and I don't remember the full details of this, but there was One question i mean watson did beat the two human beings, but didn't get get every question right And there was one question if I remember rightly watson's answer was toronto And any half-awake canadian would have known that toronto was a spectacularly wrong answer So it obviously wasn't programmed by a canadian So it raises the question of how do we make sure we get the right programmers I think if I may That's not how a lot of these systems are working now that the notion and it's a common notion that these systems behave in a way that we as human beings have preconceived and programmed Is very much and I belong to this era. So it's not a criticism Is very much an atheist conception of how knowledge is gathered and made available If one thinks of alpha go if one thinks of A lot of the work that's going on in machine learning these systems are learning As it were on their own the New strategies the new tactics The new moves That they're thinking about have not been in the contemplation of any human being It's not that it's all mapped out in advance. This is not a 1980s Pre-programmed expert system decision tree But what you say is actually very interesting about watson because watson actually is what I call 1.5 Because a lot of the watson knowledge actually was pre coded and one needs to know that That if you read the books and the articles about it, there's 20 postdocs Actually, although there was quite a lot of intelligent search and unstructured data there and it's wikipedia was in there as well There was quite a lot of structured content as well. They're really exciting things when you move into the second wave of artificial intelligence fully where no one's needing To structure the data because I suppose there's really only two ways a machine can do things in a clever way a that uh The programmers and the content and the human contribution is clever and the machines are simply the vehicle for delivery Or secondly the machines themselves acting on unstructured information and data do really smart stuff And that's the distinction between the first and second generation It's interesting also though what you say and I never really know where to go with this because you're right and yet You say uh machines do get things absurdly wrong But I suppose the thing is human beings that think they're absurdly wrong as well And we're getting it happens a lot those who are interested in law in electronic disclosure So that's the the whole idea of systems searching through very large numbers of documents and preparation for litigation And it was established in a paper in 2011 pretty much beyond discussion that in terms of precision and recall Systems can outperform junior lawyers and paralegals, but it wasn't quite as simple as that because it reflected what you were saying actually Essentially the system made errors, but different errors from the human beings And one just needs to be pragmatic and say well actually human beings and machines make errors If we if uh, and we're probably going to opt for the The setup that makes fewer errors and then the issue is well How do we know that the machine might not make a gargantuan error and we have to think that through Sir, sorry. I'm doing your job Okay Thank you very much for this I'm an economic historian and about two years ago. I began to get increasingly concerned about this Not from a technological Angle I agree with you. It's just going to continue to grow and move forward more. It's more from an economic Perspective. So obama just came out last month through his report to congress on the state of the u.s economy saying that 83 percent of Jobs that make $20 an hour or less So approximately 42,000 a year Will be automated within the next 10 years So that's the vast majority of the Nova Scotia economy in 10 years will effectively be Automated unless we find new jobs, which is debatable So that means when del housing or kings college or mount or the Nova Scotia community college goes to Ask for money to support their endeavors. The province will be further in debt Won't have the money will continue to retract that money. I'm wondering About the economic side if you guys looked at how that's going to undermine the technology will move forward But will the economics bottom out the economy in such a way that this creates Social strife In a way, I wish my co-author was here because he is an economist and that's what his research is on Is the future of art and it's a great question you ask and Just even in hearing you Saying it I get a ripple of concern because It seems to me and I'm not As I'm sometimes accused of being a determinist that's to say it all definitely happened But it seems to me it pretty much is moving in the direction that more and more Jobs will be replaced. Now what Daniel would immediately say is that the concept of jobs is unhelpful We should think more in terms of the tasks that make up jobs because if you take the notion of a nurse, for example My wife was a nurse in the early 80s. It's very much bedpans and bedside manner And bedside conversation and now many nurses. It's certainly in england are allowed to prescribe. They're undertaking Operations and minor operations and so forth still the same job title But the tasks are very much changed and so what we're anticipating is under the heading of various jobs There'll be a there'll be some task will no longer be undertaken a new task will come on board But the more general theory is that over time Under the heading of a particular job title as the number of tasks diminish essentially the the job will fade And what dan always says is he he talks about the and this is it's a 101 from your point of view but he he talks of the It talks of the economy as a cake as it were and looking in terms of productivity and value and income and says that Just now you can look there's a There's the work and the productivity that comes about is from human beings Who are employed human labor and then there's income and so forth that's derived from capital and this thesis essentially is saying that the Employment slice is going to be Diminished even if the productivity as a whole increases the the employment slice is going to diminish and therefore More and more value in the economy is going to reside in the capital Which will be in the systems and The content I think this is right, isn't it so that that's a line of thinking and that you know All of piquetti's work and so forth is relevant for for this question So this is leading to a number of um of thinkers in the economic world saying And one solution is that if if the capital is generating sufficient value and And we're managing through our machines to maintain high levels of productivity All we need to ensure is that everyone actually receives an income even though people might not work they might receive an income so the The job of the status to think about how it is that not in some kind of patronizing welfare basis But actually because very few people will employment if you don't employment where you get your money from So we need to we need to think that through and we need to ensure that everyone has an income as it were but what interests me more Fundamentally is the question of political philosophy which it goes back to the question But who owns and controls the capital? And it seems to me we do I've traditionally not inclined towards socialist views But a lot of what was written about the common ownership of property Begins to make far more sense when you see that what you're talking about as the dominant source of value in society is this Content and these systems and when you start to think in terms of a commons approach It's very interesting because I think we we want to we will otherwise I believe run the risk of having a highly polarized society We're a very few organizations own most of the capital and derive huge income from that And other people simply cannot find the means by which to earn a living and this is why I say I think this is a fundamental policy issues that need to be addressed and it's interesting Obama raises them and we know in England we've an interest From a number of politicians of their senior level on this question. It's not a Problem for the next couple of years, but it's very much within the next 10 years. It's a fundamental difficulty So thanks so much for raising it and actually have the answer But I was just trying to articulate our way of conceptualizing the problem With one eye in the clock and another in the two questioners I think we'll call it closure after these two so first. Thank you um, so as you've said before machines are becoming better at understanding humans and the environment Now taking a look. I know they've made a leap in emotions analytics, but There's always a need for human to human interaction So even if you're calling the store to see what time they're closing Sometimes you just press zero because you want to get to the operator right away or in a more crucial manner. Um Maybe machines can sense empathy or disgust or whatever the emotion is, but They can't provide it for the patient So I guess if you can just comment on The need for human to human interaction versus machines all the time. What's the basis? I'm going to be a A little bit demanding of you. What's the basis of your claim that there's always a need for human interaction? um Just I think as a human being you would always want to engage With another human being versus most time machines So sometimes it's not just the service that you want, but you want that engagement So do you really want to engage with your tax advisor? So if you've got the option I start with tax terrible unfair and tax advisors, but you can see where I'm going in this that are I mean, we hate paying tax the idea. We have to pay someone to help us pay tax seems Rather painful So most people would say I like my tax advisor But actually if I can answer a few questions on screen and scan in a few documents and link my accounting system Yeah, I'm not really sure I want the personal interaction With the tax advisor. I've got a problem with it. I can go to a community and find out what other is done So do you disagree with that or do you still you must speak to a human being to submit your tax? I guess it's specific to some jobs as a doctor as a nurse these kinds of jobs where you need that And what we do or what we think about I'm not disagreeing you through I just want to to push you a little bit in this You have to wonder whether or not What it is goes back to my problem to what To what to my question to what problem is your need for human interaction the solution because I love human interaction. I'm a really sociable person But I prefer socializing with friends rather than my professional advisors. So it's not a it's I don't think it's a need for human interaction per se It's what's the problem we're trying to solve and you think can only be solved by human interaction And I think it's something about comfort and confidence and so forth And it may in some people will be a need for empathy, but in many of us It's just we want to feel comfortable and confident that this is the answer And I think we'll get used to in many years including medicine where historically We would definitely have gone to a human being where the machines will just be So manifestly superior to human beings that the idea that we go along to A less impressive resource For some kind of interpersonal comfort will seem rather silly So I think you're right in some areas, but I think that'll fade over time as well Empathy is fascinating because we do we do a little riff in empathy And think quite deeply about it, but I just and I don't disagree with you But I just want to say a couple of things first of all It's very interesting that We think or we argue That the best experts are empathetic because most of our experience if you speak to the really Technical lawyer or the surgeon many of them and this is not a criticism Just an observation by disposition actually are the other end of the spectrum It's not obvious to me if you've get dreadful health news to impart Or awful legal news to pass along Or disastrous tax news To discuss That the best person to break the news is the technical expert I it's just not self-evident at all. And that's why interestingly We identify a new role in the future as the empathizer because Although this and this really bears out your point Systems might generate conclusions that it's not that you need intellectually someone to help you But you actually need someone to talk to about it But what we want to challenge is that the professionals are the best people to do that We don't see why that would be the case and we see there might be a new class of role Absolutely vital the other end of the spectrum Which is the empathizer and funny enough I think really good empathizers could break the bad medical tax or Legal news equally effectively you might not even need to understand the field because you've got bad news actually What you're wanting is is someone who you really feel is in your shoes And and experiencing the pain with you not at all obvious that that's for the best technical experts to do The final thing which I think is the most challenging is that um If you take some of the empathy And what the psychologist says there's two components to empathy. There's the cognitive component and there's the effective component and um, if you look at say Sam and baron cohn's brilliant book uh on this at the back of the book, uh, there's 35 or 32 sets of eyes And the eyes express emotions and the cognitive test is can you see they get multiple choice? What emotions are the do these eyes express to you? And when I did that test, um, my wife will tell you I'm not particularly empathetic when I did that test I'm I got almost all the answers right, but that's the cognitive part Did I feel sad when I saw someone with a sad eye? Now, that's the effective part I almost don't understand the question in a way that uh, uh, that for some people each of these pairs of eyes will engender a whole set of emotions and Computer is actually can do the cognitive bit quite well But I think it's a real stretch to say, ah, well there there's what there's what There's the knob that the real defining characteristics of the professions is that they provide the effective aspect of empathy I I think people were uh, I think a most don't and b doesn't seem to me that that defines professionals It might define the new roles that we see emerging though. So thanks very much for the question. Thank you um I wonder if you find it a little bit ironic in some ways that What you've done this evening In this room is give essentially a medieval performance. That is a lecture to a group of people Uh, who are sitting here some of them taking notes others listening carefully or not so carefully whatever And that also that the university in its kind of medieval form or whatever you want to call it antiquated anyway has still managed to produce an amazing output of new ideas and Changes that have occurred a lot of what you are talking about tonight And so I'm just wondering what you see coming for the university With what you've described because it's already been doing a lot of interesting work Are they I the irony points an interesting one, but isn't it wonderful? That dalhousie is capturing this presentation And that there's maybe 120 people in this room, but actually probably 10 000 people to watch it I think that's really interesting And so that gives you one illustration That gives you one illustration Of the ways in which I should mention of course cbc or Of course, all right. Well, that was too. Okay, let me rewind so I get the message right. There's two things going on here folks in communication terms Wonderfully dalhousie are recording this as a webcast, but at the same time cbc are also capturing this For a radio and so the audience is far wider than the medieval lecture Universities are going to change beyond recognition I I don't deny for a second that there's wonderful research going on here and elsewhere But I think the emphasis will shift a lot. There'll be a lot more work on data As I say medical science will shift to data science legal research Empirical legal research a lot of it will shift to Big data and predictive analytics I'm not for a second suggesting certainly in the medium term will have no read for bright people to think creatively Indeed, I mean in the book we say The people who do the most creative the most strategic the most intellectual demanding work are the people who last the longest as it were Because that's going to be the hardest to be honest for machines to capture So I think universities Will have to as many other professions will have to rethink how they teach have to reorientate their research But it'll still provide a perfectly Valuable function in all of this we know there is a very long-term challenge and it was hinted at earlier once we get to the 40s and the 50s If some of my expectations about technology are right I think we will be moving towards a society will be far less work for anyone to do even for the most expert But that's many many years away. Let's make the most of what we've got just now My producer from the cpc has instructed me to turn the floor now over to somebody else Good someone interested in becoming a novelist Be replaced or at least affected by artificial intelligence Isn't that a great question That's a wonderful question Um There are systems already That can generate novels not very great novels, but lots of human beings right not very great novels as well Here's what I think And you may think this is unlikely I don't know if it's 10 or 50 or 100 years from now But my guess is machines will be able to write novels that actually are better than any human beings But think of this in a different way think of um Running a race I've no doubt we'll be able to develop robots or androids that can run faster than you say in bolt But if we watch that happening, we wouldn't think any less of bolt When we see him as a human being not using any tola technology Running down the track we marvel in that And similarly I think in art while We may be bowled over By a piece of music or by a piece of art that's generated by a machine I think there's something special In knowing that that was produced by human being And that's the point that even if a machine could write a better novel When you read a really great novel, I don't know how you feel I often think gosh, there's no way I could write this and that's part of my admiration for the novelist I wouldn't feel that same sense of admiration and enjoyment So I look upon art and I look upon literature In part as an aesthetic experience, but I also look at it as a comparative communication exercise And so your question is a great one. I think our machines will write fine books But most of us will still want to read and celebrate the efforts of fellow human beings I think that's a I actually believe that is a very appropriate way to end this evening's session I want to do so though by thanking first my colleagues at the cbc for helping to record this Certainly all of the folks at the shulik school here at dal who Have been very Very instrumental in starting initially and now continuing But I hope will be a great tradition in the sigram day lectures and finally Please join me in thanking richard suskand for his talk and his answers his generosity and his insights this evening