oh i'm very pleased to be here i'm very pleased to be here and listen to the lecture by professor freeman why don't you introduce yourself nice to meet you i am the most popular humanoid computer have been developed at the italian institute of technology in i am genoa to the contributions of many international researchers i have 40 brothers that are distributed all over the world and at present there are over 30 laboratories in the eu japan and u.s and my skin is sensitive so i can perceive whether i'm turned off and whether i'm touched people are very scared not just in this audience or in other very sophisticated activities brings about the contributions about rehabilitation and the biomedical research are you going to stay with us for the next few days if you want to ask your questions with the number of for this edition of the festival richard richard is always wearing his hat also in tv shows richard he's an economist he has a number of acknowledgements technology so how to measure the the intelligence of the committee for understanding and the engineering workforce so we worked extensively on the subject connected to the impact of new technologies on employment and on the level of education he has gained a lot of acknowledgements and i'd like to refer to the labour economics prize that he has been awarded the most distinguished award that he has been given in 2011 then he received the acknowledgement from the global american academy of social science and then the global equity organization awarded company income distribution if you go through his resume and he's 53 pages long with a long list of publications he's just not saying that he has spent some time in different universities but he's published extensively look on google citation you'll realize that there's a publication by british that has been visited uh many many times this publication is called what do unions do where he questions himself and that was a reason for inspiration for many scholars they're interested in trade unions and not just economists so he actually questioned himself on the very function of trade unions so his assumption is that um there are two aspects of trade unions one is very important there's a very important side of trade unions the one which so all different workers in a way speak through the voice of the trade unions and learn from trade unions about different problems without trade unions public workers could not find jobs or could find it difficult in getting which obviously can lead to negative side effects so this is important assumption and theory of visa which tells us a lot about the uh aspects of the important role being played by trade unions we decided to have him hold the inaugural lecture because one of the main subjects that we're going to deal with this edition of the festival economy is how workers can do benefits from the huge increase in wealth that can be brought about by artificial intelligence and by the way this is a question that we left and answered at the previous edition with the blanchard lecture who's owning the robot because this question is critical much depends on the answer that we give to this question in terms of well-being of citizens of the wealth distribution because if any one of us could own a robot then this robot could play different jobs that we don't want to perform so that we could become more efficient in terms of other tasks but if robots are owned by other well then the results that they bring about uh bring about wealth which is concentrated in the hands of few so then we are left with doing things that we don't do doing with robots doing things that we might be more interested in doing with other people earning from that but without any further ado let me give the floor now to our speaker professor preacher freeman thank you and thank you for the nice introduction so i'm going to be following these slides and actually there's a since we're in a theater there is a drama that i that i will be presenting uh to you um the the question of what are you going to do when the robots um do more of the work and earn more of the income you know clearly has the income distribution issue that that tito uh talked about the first paper i wrote on robots just had this title who owns the robots rules the world and i thought that's the only thing an economist really can clearly say because everything else is about technology and if the robots are able to substitute for professors or they're better able to substitute for chefs or they're better able to substitute for uh you know workers on an assembly line and the pictures here on this this on this slide are the many different areas where we now have robots substituting for humans the only one i will not discuss is the the sex robot which is the last one and i have not researched that and um okay so that so what more can one say uh for the the play oops when it went a little faster i got to go back one yes so i have a cast of characters for this for this this uh story um there are humans and when we think about humans we have two parts we have the brain which is our conscious consciousness we have our bodies and then we're connected as a person brains to bodies robots in some sense are bodies are not as flexible at the moment as human bodies but they can do other things that we can't and then they are connected to brains and that's where the artificial intelligence comes in so i present the humans here and then i present the robots but there is a parallelism and then a lot of robotics technology is trying to copy what biological creatures do they would love to be able to build a hand as flexible and as connected to the brain as as our hands are and and they can't do do that at the moment the hands on the robot we saw can do a few tasks very well but it is not comparable to your or my hands they are building prosthetic hands that are meant to connect to the person's you know brain basically through the nervous system um and those are my view the best hands we have for robots uh but they're not fully electrical electronic they have to be connected through through actual human nerves um we also have in in the human part of the of this drama market economies which are directing resources in different places right now it's humans workers connect up with uh capital um and uh the capital is owned by a certain number of people capital as you must know is so much more unequally distributed than our labor earnings as to make the discussions of the rising inequality of labor seem small and they're not small they're huge to everybody uh uh but the capital inequality is is far uh in in excess and i do have the old-fashioned das capital capital that's plant and equipment but it doesn't do things the way robots can do things and it certainly does not have artificial intelligence to be able to make decisions as a human brain would do this the final set of players which will come into the third act of the drama i'm going to tell you are transhumans cyborg um and that is where you implant some computer connection or computer machine type ability into us you might have said the prosthetic hand is indeed makes a person partly a cyborg partially human but partially machine you can imagine somebody with artificial liver an artificial heart a whole bunch of things and they would still be their their brain somewhere in there but the brain might also be connected to the gps and wouldn't wouldn't have to use the cell phones uh to find its way around on then there are the two transhuman creatures that economists strongly believe in the mono invisible and that's what runs the markets we we we like to think at least and i unfortunately the dancing pronounced it correctly um demand of and general equilibrium they will not be able to be here to sing but i will render some of their songs to you and you will you will see what how they play in the drama i'm going to skip over this quickly but if you're a college student the often you don't read anything real you get the notes they're called spark notes i called it here mark notes because i didn't know whether the spark note company might sue me for something uh etc and you see the the drama here opens on a warm may night where there are scientists discussing the issue of the of the uh on robots and they're discussing it however by connecting with a quantum computer because that's where i don't know in 10 or 15 years we certainly will be running simulations of economics and able to do much more than we can do now and the first scientist says that ai robots will create machines that work better and cheaper at all jobs than humans and i'm going to give you some evidence for that from ai robotics experts why they believe that the second one says sooner rather than later this is not something that's a hundred years 200 years into the future for many occupations and work tasks it's going to be every year more and more things will be shifting to robots uh the the third scientist just goes double double toil and trouble fire burn and cauldron bubble which is the potential problems this these events could occur can bring to our societies if it creates greater inequality and leads to a even stronger let's say wealthy class versus everybody else system and and i'll have some policy conclusions at the end have to try to prevent that um then the second act the scene shifts to a summer festival here we are and there are two people debating and it was tito was not economist of traditionale debating the young robot but there is a big debate uh between economists from a very traditional view of this technology it's no different than any other technology we've had automation scares in the past and they never panned out and then a robot futurista if if i had my three acrylic aqualita here they would be they would be saying for the traditional economist future like the past because the only evidence we traditional economists can look at is the past and so we try to extrapolate from past things and therefore many of my friends do not believe that i'm fully sane by thinking that the robot thing is something incredibly uh different the robot futurista uh says the future will be the future you imagine and i have moved to that point of view over the last i would say five to ten years and uh i will give you these three laws of robo economics which partially reflects why i think that and in the last scene there's a tv broadcast desk and it's talking heads trying to offer solutions to the the the problems and um i'll actually offer some solutions etc okay so act one of this play is is there a technology challenge and i want to start off quoting uh one of the great economists in my view far greater than most economists recognize this is herb simon who also was a great builder of artificial intelligence and robots extraordinary person and he wrote this book in 1966 and it's he about the then fear of automation which swept through the united states uh in in the early 60s unemployment had gone up a bit and they were bringing in all these machines and everybody was a lot of people were scared and i taught his answer here for many years in my graduate economics course as absolutely correct he has a model the model makes sense it fit all the data for that period of time and bing it just and he just said don't the boogeyman of automation consumes worrying capacity that should be saved for real problems real problems are problems of scarcity they're that many people in the world are poor hungry and they're not problems of robots doing all the work and us sitting around not knowing what we're doing i compare what he said with today's headlines and you you see the the headlines as you is you you must see them all the time there are every i would say every day in in the world press there's something about robots going to take our jobs robots are beating lawyers um in in in solving uh some legal problems that happened just this year in uh uh in england um uh and dot dot and i'm actually putting together a huge database where we're scraping all these stories so it will become a data source for anybody who wants to study which of these articles and fears pan out which do not what are the companies involved in making them how much is is hype and the bottom of the slide i said there are eight this is as of january of this year 18 studies projecting job losses and the job losses range hundreds of millions to don't worry it'll all be all right because the future will be like the past and in the past we always created good jobs for humans so there really is a debate about this um that we have to discuss this is a slide that comes from a paper that was just published a uh may 30th last year and when i saw this paper in in the archive uh i just was stunned it's extremely well done survey of 352 ai experts who published if there's somebody here who knows the artificial intelligence business um they told me those are the most important conferences that they have in artificial intelligence so this is real you know real pros and they ask them when will the ai outperform humans at work they did a really good job of the question asking different ways so it was a nice piece of not a cheap survey question but really well done and the first one they said in a few years folding laundry i thought i read this i said that's can't be right folding laundry requires hands it requires you have to recognize which piece of laundry is a sock which which is a shirt dot dot and it's not so easy to do and i thought that's not going to work as you'll see there is a machine that's coming out next year faster than they predicted that you will in principle will be able to buy at a reasonable price that will fold your laundry so like i quickly said oh they know more than i do clearly they truck driver well we all hear about the driverless cars coming and they go down the list the only thing i found a little weird is the last job where ai robots will be able to outperform humans is guess what it's an ai researcher so they think they're the peak they don't realize um how economists know so little about the economics world that that the robots can't possibly uh uh catch up with us until we learn a lot more to teach them but but so be it uh so i've had this debate of is this technology different i there actually are i don't want to name some of my friends who have said humbug this is just ridiculous and uh they said richard you've you read too many science fiction books when you were a kid which may or may not be true too many is a strong word but i certainly read many and star wars movies and whatever it is and i i think a lot of the futurist people be they ai people or something else do come from a science fiction orientation uh uh uh background and they say look here's this technology coming this productivity zooming up if it's going to take away jobs no we don't see any big burst of productivity in any country in the world that i know i mean there may be some small countries but in china the productivity is growing more slowly in europe it's wiggling along around the way it's wriggling around in the us do we see massive unemployment that which gives us a scare no we have relatively high employment population rates and they could add the unemployment we have could be cured with pretty normal economic policies if the governments really wanted to do something so it's not etc and then they go back to the herb simon kind of thing saying this is silly and then i put at the bottom here a machine is just a machine it's a mechanical contraption nothing special machine like any other i ask you now the question i have my answers to this you want to go to a good old-fashioned family doctor or you want to go go to a robot doctor that's knows your jihol genome has connections to every patient record in the world and uh the latest medical research and will be able through the quantum computer to in one two seconds give precise you know all the evidence there uh i i know which i choose i don't know which which you choose um but i think if you're mildly rational and the old family doctor maybe is not your father or your mother you you would choose the robot as well um do you want to be driven around by a driverless car or have mr toad uh drive you drive you around if you don't know mr toad he is from a 190 i think 08 book for children in the u.s that generated a whole bunch of movies and rides and he just drives like a madman and crashes all over the place because in the movies nobody dies and then finally you want some financial advice there's your old-fashioned stock broker mr sleezo or do you want to go to a robotic system that will give you um scientific advice that's not a normal machine if you would prefer that machine to uh um human experts in the in the in the in the air in the in the areas i admit i'm getting a biased picture mr toad is the expert driver for the humans uh and why why is this machine so different um it's because it can out-think us so i just list here all the places where the ai can out think us there was shock when watson defeated uh uh uh uh assuming deep throat defeated kasparov and the and you may or may not remember that he basically said it cheated somehow he couldn't believe he couldn't accept this and in fact he might have been able to win but he actually got emotional at one point and didn't think as well as he normally did and the machine did not do that so we are a mixture of emotions and and and logic the jeopardy champions who lost to the machine were much more either gentle in their defeat uh when google's alphago in 2016 beat the our greatest korean go champion who at one point been had been the world's leading culture people were shocked and lee sidal was crying on the stage at one point he spent his whole life playing this game and it's it's it's a game where most people said it will be 50 or 100 years before the machine is going to catch up because it requires incredible intuition about moves made in the in the very beginning of a game on a huge board which will ultimately play some role you know 200 moves later it is far more difficult than chess in in in that sense then carnegie mellon and the university of alberta both put together ai machines that beat human experts in poker poker is very different from board games the board games are perfect information poker i could be playing poker with tito and he could be bluffing me and so the machine's gotta somehow rather he looks at his eyes maybe that tells he's bluffing i don't know it has to make decision with incomplete information whereas the the chess and the and the check and the um go are complete information games and and that also so i was going around saying wow wow and then this december uh the google go team did something which to me was uh absolute blow away about these machines it took it's now called alpha zero which basically took part of the learning algorithm that he used to play go and they applied over a weekend this learning algorithm they just here are the rules of some games basically the machine plays against itself in in in the you know in the in on on online and it learned in here it says in 24 hours to beat any creature on earth and what they did was they had it play against the computer machines that had beaten every hue the human champions in 24 hours i i i just was stunned and that just to me well the other people who are stunned by all this are the military and the u.s military has a great desire to maintain artificial intelligence dominance for the us the chinese when it wasn't when lee sidal lost the go game at least i've read in the chinese papers i don't read chinese but our students who do uh it was when the second person that um that the google alpha go machine which has now been completely surpassed by the alpha zero machine it beat the then champion of the world which was a chinese guy who was better than lee sedal at least at the current standing and the chinese military said oh my god this was assigned to them that they had better get to be the number one country in this if they let's hope just for the protecting the country not for doing of other other stuff so this is really uh a major thing that we have machines that can outthink us okay the other aspect uh is that the with the machines my the ai is working in digital space and it sort of it affects the real world only if we can connect it up to robots or it tells us what to do and and when the the go machine defeated lee sival they actually had somebody moving the pieces for the machine that's because the hands of robots are not that well constructed at the moment they could have built the machine it would have cost a lot of money to have done the the robot here could not have picked up the go pieces and moved it it was cheaper to hire a human to do it uh um and so i just list here story after story about what is going on we're we're taking the ai putting it in the robots and it is transforming the world of work um and just miss i i have i said thousands of stories of this and and we will we will have more than stories in in about hopefully end of summer where we'll be able to say what's happened who's buying the machines or if they not made a pass to a uh a market uh test and uh the one that to me is the most exciting for the moment is you go into a bank in shanghai you don't have any credit card anything they have a face recognition machine the face recognition you look and you're able to withdraw money you can do whatever you want it just looks at your face and it's got your face your face is different than anybody else has faced in the world and i have a harvard student who wants to replace every lock and car and in the harvard dormitories everything can be done with no keys and locks face recognition and there's evidence it will be safer and cheaper and better bang so it's it's just going to be all over the place this this kind of stuff and it's quite just a question of time now i had a short musical interview for you um unfortunately when i was five years old in kindergarten i tried to sing with the other kids and the teacher pulled me aside and said please i was destroying the music so i cannot sing to you and you'll all walk out or start throwing things or something uh but this is from a musical that'll be out in 2046 it's called annie get your ai it's based on a uh irving berlin famous musical annie get your gun in which a man and a woman are having this debate over who is better and he sings i can do anything better than you and she says no you can't yes i can so i have an ai robot and a human singing and the ai robots start saying we can do anything better than humans yes we can watson wins jeopardy yes alpha zero wins go yes liberatus scores in poker yes foldamate is the name of the machine that's going to fold the laundry geek x is the machine that beat the the a british team of london lawyers in finding flaws in uh legal documents uh spot mini is a boston robot that is a dog and supposed to carry things um but it's not cute like the robot here and i don't think it's gonna be a great success unless they make it cute because the romans got to live with the humans but it now opens doors for people the roomba cleans the floors gordon is is a barista robot that will replace all the people making the coffee for us and then there are two sex robots about which i know nothing harmony is the female and then there's also henry the male um it looks as if the sex robots are a specialty of the japanese for whatever reason so the robots end up saying working all hours ai robots victory is ours bang the human answer to this is that wait a minute each of these robots can do one thing and the i i can beat the uh the the robot that be the the machine the watson the one in jeopardy i can beat the alpha zero not even go i i can beat them in another game it knows only one thing it's specialized in that you and i can do lots of different things we're much more versatile so that's our comparative advantage and versatility and then i thought about that for a while hell and then i said wait a minute isn't the whole of financial economics about having portfolios isn't what's really going on is we're building up a massive portfolio we have a specialized robot that can beat us in poker we have a specialized robot that can beat us in this that can beat us in that that can beat us in in all kinds of things given a huge portfolio of robots even though we can beat anyone at any task except the one that it's been trained on there's always one that can uh uh convene this and so i had the the robot of the or the artificial intelligence say i can always find one better than you and i think that's going to be the truth uh you will challenge me in something and i will come out with a robot that can beat you in whatever you challenge me in regardless of of of my confidence in anything um so that's the musical interlude thingy act two this is an economics festival i better make sure there's some good economics here so um i i think there's a very simple economic principle that we should apply to the robots and the ai it's comparative advantage i do not think highly of the these studies that say 600 million jobs will disappear um 600 million jobs may disappear but there'll be other work that people can do and we have some comparative advantages the robots are not perfect biological things the way we are um and so the real question is what tasks will the robots be better and more cost effective than us and what tasks will will we have um and so it's comparative it's comparative advantage is the way to think about this and that in turn compared to the end which ultimately determines incomes uh incomes of countries and uh in in the in this case incomes of people working in in the job market there was a a uh a woman i don't i don't i never met her i just saw her she wrote it she didn't write something she was quoting the newspaper saying oh humans shouldn't worry about robots after all there's things we can do that they can't so the reporter says well what are they well just look at walking running and kicking a ball and i thought if that's our comparative advantage do you imagine how much they're going to pay us for walking kicking a ball is a little complicated because there is a robotics a league where robots play play football but as long as we don't let the robots play in the human some a small number of us can make a lot of money kicking a ball but most of us can't and i and i thought if if that's the case and the ai robots are going to take over the really complicated thought jobs that get paid a lot and we are left to being the fingers that move the pieces in the go game because it's too expensive to build a robot to do that we've got we got some real serious uh income problems uh uh to this that's the first thing in economics say then um you may be familiar with the the three laws of robotics which is from isaac asimov uh who wrote the uh irobot and a bunch of other robots books in the 1950s and uh so i decided we economists could have our own three laws not of robots but of robotics robo economics law and if you accept these these laws as correct generalizations of the world you'll end up i think the same place i i've ended up thinking about this the first one is that robots and humans are becoming better substitutes over time that's the whole point of getting of working in these robots it's to make the little robot you you saw it still can't walk um its hands are not the whole goal is to make it more like you they they learn from human biology or you know ate biology but by the biologics they're learning i think that implies to me economics higher elasticity of substitution of robots for people over time anything a human does better than a robot there will be somebody maybe at the italian institute that made the robot uh who will say oh it'll be really cool to make something a robot ai that can beat the person that just did this complicated task and the businesses will be sitting there and paying for that i just think we're going to see increasing substitutability and i'll give you some evidence for that in the minutes of these laws second what does technological change do it reduces the cost of producing things what does that mean if the costs of a substitute for us a robotic substitute gets cheaper and cheaper what's going to happen to our wage in the tasks the robots can do well our wage is going to fall we can't compete with the robot if it's getting lower and lower cost over time and again there'll be some comparative advantage areas where we could do well but it's just the spreading of the of the robots to every every corner and then finally there's which was my first thought on robots but the real thing is who owns them uh unless tito said the effect on our incomes depends on who owns the robots if i own the robot that does tito's job sorry tito you could maybe sing in the street and we'll throw you some coins until we build a robot that sings better than here in the street uh but if he owns the robot he can have more leisure and the robot income will flow to him and that i think has to be the ultimate solution to uh what simon called intolerable abundance where the robots can make everything and it's it's all very nice what kind of evidence do you have for these three laws i just say in one presentation i gave him a seminar at harvard now one of my former professors and and ex-colleagues too i won't mention his name and not somebody i especially liked as a teacher or as a colleague but very tough-minded guy he said there's no evidence for this where's your estimates of elasticities of substitution and i said well well i'm going to give you some estimates in a minute of that i didn't have any i said but everything i see says that's what they're doing so here i'm i'm i'm gonna give some of these things that i regard as stronger evidence than elasticities of substitution estimated from production function data but we'll we'll have that as well and i took the hands which is a key thing that every time i i worry about the robots i always say but my hands can do so many more things than the robots can well they are working really hard at giving the robotic hands senses of touch and when i spoke to the uh the scientists the robotics person about what he's doing in the labs here in italy he said oh we're working really hard basically and maybe the hands is one of the key things and this may italian researchers at his institute i didn't know that he did this at all uh it was announced in the things they developed a lighter cheaper robotic hand the hens robotic hand so that is a really hot area of of of of work make it more like a human hand the koreans i was in korea two weeks ago and they have got some special hand that is way beyond the human hand doing things out of the human space possible so the hand could go into burning you know fire and lift something out or whatever whatever it was doing special kind of silver nanoparticle woof so they'll have hands to do everything they're making them better substitutes for us and capable of doing more things making the humans more robotic you perhaps have seen elon musk made this statement that the humans must become cyborgs to stay relevant so um and he's got a company neurolink that is designed to get our you know uh implant us with chips so we can compete with the ai um and there's just all of this work is meant to make us more robotic us and robots becoming better substitutes um i it was in the newspaper in cork ireland the uh they they they had a show where the robotics guy came and they implanted if i hadn't i would do this if you agreed and this was the medical guy said it was okay we would implant a chip in petrito and uh the kind of chip that might be very good for you uh would be i want like the one the gps you you don't have to fumble for your phone or anything you immediately know which way to go wherever you're walking and that's certainly something they will be implanting in brains some way that you know down down down the line and then i thought of one of the most iconic creatures in the u.s scarecrow um from the wizard of oz and the scarecrow had no brain had no brain he's made of straw all he wanted was a brain and you know in the movie somehow or other he ends up being smarter than everybody else and it had to have been he had a robot an implant that we didn't know about and then um i thought about president of the united states i thought he it would be very helpful for him if he had a truth-telling implant which you said to him as he was about to tweet out something totally fallacious and nonsensical if it just said no no his own brain saved him no no that's wrong and he would do what the rest of us do you know check on google check someplace to see if what we're saying is correct or not but then i found out he claims to be really smart a genius without any extra stuff so so be it um in italy you may also face a thing that i've wondered about the u.s at what point might we choose an artificial intelligence robot to be the leader of the country and and we can have a human there like moving the hand the pieces on the go board um okay now for some economics based evidence and this the the paper that i like citing for this is by daron and roberto restrapo i believe rostapo is italian from his name he's not oh yeah perhaps originally every maybe originally yeah i i've never met him i only don't um and this is a study they did where they looked at uh changes in robot intensity in labor markets in the u.s versus changes in employment pop based employment population rates and changes in wages and what you see is the downward sloping curve implying bringing in more robots you reduce the wage side and you also actually reduced employment there's some reasons to be suspicious of this finding and uh not that they just that they've done any statistically wrong they may have economists always do that but it's just the way they pose the problem so i had a a student uh harvard this last year i said look at the data completely differently don't do it by cities look at it by occupations in highly robotized industries and non-industries so it's reorganizing the data in a very different way and the student got results that are consistent with i would say uh assamoglous and vestapo's results there was a study by some german researchers that did this for you for for for for for uh uh for germany they got the same results on wages not really very much unemployment and then there's been one other study looked across countries there is growing evidence of the traditional economics elasticity of substitution that yeah you bring in more robots and wages can go down or employment goes down the falling price of robots i'm not going to go through this i'll just say robot prices are declining somewhere between five and ten percent a year and it depends a little bit upon how many corrections you you're able to make for the new quality robot because like a computer they often have the same price but the robot now does things much much better so i just went through all the data on that everything says going down the most important robot in the world currently is the automobile welding robot and when you go to if you go to a welding factory welding was a really skilled human task very high paid and the traditional machines could not do this and now it's all done by robots and the people just run around do little little tasks around what the robot is doing and this is just for that particular machine they took it not only did they look at the cost of the robot it was the cost of having the robot operating the cost of the management the it's the full cost of the system these are going down i don't i think that's unquestionable we're going to see more and more pressure declining our costs of robot substitutes for us then this i'm not sure about but we have seen over the last 15 20 years in every advanced country in most developing countries we have seen labor share declining shift of income from workers to capital and this at least in the u.s statistics this is even though executives get a large fraction of their pay but in the form of capital income it's a compensation in terms of the you know you have to pay a tax inc it's labor compensation but they don't want it in salary they want it is part of stock options or in stocks which is capital so these numbers would be much bigger i believe if the other countries i'm not sure i just don't know quite how they deal with that problem for the u.s it would be much larger fall um but this these changes occurred before the robots at ai have really hit us so uh this isn't real evidence for the law it just says we're having this going on the world robotics market is increasing incredibly rapidly when darron did his study the the the robots have grown since he ended his project four times there are four times as many it's going to double over the next couple of years the it's it's just it is an accelerant for the moment an accelerating uh uh curve um artificial intelligence the same thing so whatever evidence we can generate from economics type data is going to be just the beginning and it's at a period of time when the companies are struggling to how do i use this to make profits what do i do with it etc etc um but i believe there's going to be a lot of interesting phd theses and studies that will follow this and we'll see how important it is um so i see a future where the ai robots are doing the work making the money and the owners of the people who are cleaning up so first i said the upper one percent but no it's not the upper one percent the 0.1 percent and then there's the 0.01 and probably now it's the 0.001 i mean the inequality increase at the top of the income distribution is astounding uh that's all i can say it's just amazing how much money is going to so few people i'll i'm not saying anything bad about those people they may do good things but it is something you're wrong about so now is the last act talking heads uh what shall we do first i thought about what do you do as an individual you're a student a young person what would i do uh the u.s bureau of labor statistics predicts that the basically the only place we're going to see growth of jobs are jobs where your computer and information jobs these are for instance for university graduate type people here but the same thing is is true of other jobs it's it's it's places where you will be working with the machines and hopefully not being replaced by the machines what do you do i think you have to major in uh or minor in computer science you have to take courses i think that every university should as part of a requirement for its degree for the people doing poetry they have to take three or four computer science courses that will help them get a job and help them communicate with the machines that are going to be in their lives for for the rest of their lives i think that's a natural there's a danger here one of the uh national commissions that i actually wasn't on that i testified before there was a fear of the of the computer scientists that we would see these cobweb cycles and in the u.s we've had these big growth of people majoring in computer sciences the market breaks then there are lots of people running around and nobody major so it goes through very traditional cycles and the question is will we see a cycle like that coming out because right now the american computer science degrees has zipped up past the peak that i have in this graph i'll give you those numbers at the end it's been a 50 percent increase in the last like three four years everybody wants to major in computer science uh or minor in it with a major somewhere else and i don't think that we're gonna see a real bust um because i think this is the land going to be the language of business and the language of lots of other things so you you need it um the other thing that we see going on and this is uh is that we're seeing more of the robot producers saying well the robots are going to work with people so they worry about building robots called cobots it can't be a giant robot that's huge and clunky that will if it tries to pick up your your grandmother who out of her bed you know sort of picks up the whole bed and it's got to have all the feelings and the sensors to be able to gently pick an elderly person up and companies sitting there saying we're going to have a lot of robots doing the work we need the people to cooperate with the robots the the robot person here said uh that they've done some experiments that if the robot does something and smiles people feel really good so the robot will do your job and it will go up to you and say smile very nicely and you'll feel good rather than say damn robot did my job better than me etc so the prediction that's just the prediction for the moment is that the next generation of robots are really going to be oriented towards working with people so the advice the people is you better be ready to work with the robots etc then there's just implanting stuff i would gladly implant chips into myself the rfid chip a lot of people have done that already so you um the uh the for the facial recognition you don't need the chips or some things can still be outside and it can do it they're 3d printing an ear so chop my ear off the uh we will have built through 3d is is a the the 3d printing is very much an artificial intelligence robotics operation people have almost no uh role in it and the fears of the chinese china is manipulating dna to treat a new class of superhumans well that's not putting the chip in that's just doing the dna because we have a lot of concern over just biological changing us um etc so eventually i'm sure we're gonna evolve to be cyborgs uh and i gave a talk to a group of union people in the u.s about maybe a year and a half ago and one of the union guys raises a woman raises her hand she says professor would you rather be a human or a robot and i hesitated i started to think and there's this famous comic jack benny who was known for being stingy and when they asked when the thief holds a gun at his and says jack benny your life or your money jack's thinking his money is very important to him and i i i i i think about that and i asked the question to some students and uh one of the students said human because they have feelings and robots don't make sense we're working hard at giving robots consciousness and feelings and there's a lot of interesting you know neuroscience stuff about what is consciousness and how we can do that then another student you said she was an a a robotics major at northeastern she said machines are smarter but they're not yet conscious and she was thinking we'll make them conscious and then clearly and then i thought well the right answer of course is cyborg how could you not want to have a heart that beats better than your heart when it starts to fail how could you not want to have all of those stuff but i ask at the end you it's a choice you will face i think in in the near future so so the second and last act of this drama is what can society do so you can do your things and i can do my things to deal with this change and i think that there are two kinds of policies society have i call them before and after a before policy will pat will try to level the playing field for when the ai robots really dominate economies and that means that we people and every person in society has to have some ownership stake so we would not rely upon marginally trustworthy politicians to pay welfare state benefits when some other people are upset with their taxes and dot dot dot ownership is a stronger in in in western capitalist economies certainly private property is a stronger moral basis than um using a tax system um maybe that's more so in the us than elsewhere so the other is after policies and i think of uh uh tom picketty who's gonna have a global capital tax bill gates proposed a a a um a a robot tax and then a lot of people like the universal basic income tax but i think if people have ownership it's stronger than having a relying on politicians any case i'm going to talk about these policies the first thing is there's a very strong case for people having ownership of their own company it's not complete ownership but certainly partial ownership in the u.s we have 10 percent of our workers work in companies that are employee stock ownership companies those companies do marginally better than other companies they are much more stable they don't lay people off that quickly in recessions the workers are happier it just goes they have lower better retention rates clearly works for the 10 of our companies that have us they're given a tax break if you do this so we we encourage this um so we know it works a lot of big companies around the world you mentioned the the geo group that's a group of largely it's powerful in europe more so than in the u.s and it's a group of i would think pretty positive minded companies that really want to have workers have bigger stakes in companies and spread this as the right model of capitalism rather than have a few people own everything so that's and i've always been endorsed that etc but if you start doing calculations you quickly realize if the robots and ii are going to lower my income a lot ownership in my company increasingly low wage job is not going to give me a good living standard you have to have people own a share of business capital writ large and so how can we do that i mean marx said something about workers owning the means of production which i think probably is a very correct and good statement but he didn't take it full length that that means the workers have got to be like capitalists and we have to have ways of distributing it and we have to do it uh in a way that doesn't have class warfare and etc etc well we have one outstanding example of this and um with that's and i'll talk about it in a minute the sovereign wealth funds we have a couple of examples where we do have wide ownership particularly in the u.s but it's other places pension funds if you have a private pension funds it does not invest in your company it invests broadly the united states the pension funds owned a lot of our capital market a sweden introduced a special tax where the money goes to a pension fund that that that owns parts of the swedish stock market or global stock market people have investments in in this that's worked out reasonably well so we we have pension funds but that money only goes for your retirement you're not allowed to tap it beforehand i put individual ownership here because and i was talking about this in brazil some very wealthy brazilian guy said why they just buy shares like i do and and uh well in the the pension funds have tax advantages so no one in their right mind is going to buy too much individual shares when your pension fund holds it and and you get all these tax breaks for it um and so we've seen a huge drop of ownership by individuals it's just it makes no no sense and then there is the sovereign wealth funds which are funds set up by countries or states could be set up by an italian district uh province and uh uses some public monies to establish a fund that then invests widely in capital markets and gives dividends or some other payment to individual citizens and these are growing very rapidly but they're controlled by the same characters that control finance ministries and the imf and so on okay so the one i want to talk about is the u.s alaska permanent fund that's the only one that gives dividends to citizens and the dividends have been between one and two thousand dollars depending on how much the the fund is doing uh a year to every person in the state everybody so the newborn baby gets a thousand dollars the elderly grandma gets a thousand dollars the husband and the wife each get the thousand dollars or the two thousand dollars when they were giving the two thousand dollars um that is a huge increase in the income of people in alaska it goes to everybody you might say the rich people don't need it but it's just it just it's establishes as a universal right and it doesn't matter the rich people are getting it the key thing it's going to poor people um the uk tried something like this and there's some discussion of it again um but not as the level that i would think we have to do with the robots and the et cetera and now the question becomes how do you what do we have equivalent to oil monies what we have is all this knowledge that we have created as societies a lot of it funded by public investments that companies are used and some way we have to think about taxing companies for their use of knowledge based things putting it into a fund when president trump gave the they estimated over 10 years will be 2.03 trillion dollar tax cut for rich people in the u.s a lot of through the corporate tax if he had only listen to me but he doesn't listen to uh to me he hasn't got a good robot chip to give many advice uh he could have just said this is for come only this tax cut comes only to companies that have workers as part owners of the companies would have transformed the american capitalist system uh in an incredible way his people would not talk to anybody but with ideas like that so bang and then should we worry and plan now or later and so i have an american philosopher of our famous philosopher alfred newman i'm sure you all know and he never worries about anything and he says it's the babies the baby's gonna worry about it the baby would look at that robot face we had here and they're trying to make the robot faces look incredibly babyish so the baby's a little bit scared and i look at every time i see a kid i keep thinking it's not going to be the alpha zero that's gonna compete with it for jobs it's gonna be alpha 20 assuming they improve it every year kid's gonna come out and it's gonna be creating it's gonna be connected to the cloud and there'll be a quantum computer now maybe the kid will have an implant and but i think they have but then the kid also says that this robot doesn't cry it doesn't make dirty diapers um and they'll it'll be such baby-ish but who knows when maybe some people will think of substituting it for babies so i think we have to plan now act sooner rather than later and we should start seriously thinking about establishing particularly funds for citizens that will build up its capital slowly and slowly move with the help of some businesses that are favorable to this uh in in a direction that five or ten years down the road every person has a fund of money um and you might say oh this is terrible the the the the the poor people just be idle well rich people have been idle for years and they're our most distinguished and powerful citizens so we can have idle poor people too and people will find something to do i'm not sure what my son would like to just play computer games and he said that would be a good life for him but but that that'll be a problem to face better that than face the problem of uh him working for me because i own the robot that does your work thank you so thank you so much richard for this very inspiring and also entertaining i would say presentation we have not very many minutes i'm afraid but for a few questions from the floor let me see if anybody's willing to ask okay there is a person down there no it's better if you use a mic okay thank you very much professor freeman for your lecture that was absolutely fascinating i'm not an economist so this is just a question of personal interest i was wondering if there is any research being done on how much work robots would create in the sense that we still need people to project them and to well build them up so i was wondering if the research is also looking at that okay we have to answer immediately or we take a few i can answer okay as of now there is no no real research on that you put your finger on something very important because all previous machines have created jobs for people that we didn't think were there robots will do that as well and i just think the key issue is whether those will be good high-paying jobs or moving the uh the pieces on the board that will be a low-paying job but that that's clearly an area for for future research of a significant kind very good more questions well while people think let me ask you a couple one is who runs these sovereign funds will be a person or a robot and the second is do you think that what are the implications for unions strategies to some extent of all this because maybe you know robots may be also a kind of new forms of slavery and uh i doubt that robots will get unionized while unions can you know to some extent have some saying about the way robots should be used in in in work organization right in a way that makes them more in a way cobalts perhaps rather than robots yes for the first thing that yeah the sovereign funds are run by um you know investment banker type people uh hopefully they majored in economics and evil powers so i'm a bit worried about yes but it's certainly the wall street is moving much more to using the artificial intelligence and if you're doing what the um uh alaska plan kind of thing would largely do you you you you just diversify very widely and to have a diversified fund you don't need too much human uh uh you know as special expertise the and the unions first there is a uh a a union robot what's uh ibm is created at behest of a small american union our walmart that was trying to organize the one and a half million americans who work for walmart they realized that there are about four people in the union they cannot service workers they cannot help them so they went to ibm to create an artificial intelligence robot that knows the company handbook and is now being given more u.s labor law so when a worker calls up and says um watson union watson what should we do they'll be able to give advice and then they're having chat rooms and the robot is listening and learning from the chat rooms as well um so i actually think the using robotics by unions can give them a lot of intelligence using ai to help them do their job of defending workers and and i agree they should be representing workers on uh both bringing in the robots and uh and how they would interact with workers uh and what workers would like and prefer because for a long time that'll be the cobots the the eg matal in the in in germany has a pretty uh forward-looking uh view of this because they they know that they can't keep their wages high unless they bring in the robots so the union clearly is not fighting them but you do want to make sure that they have used yeah that use them see i mean owning the robots is if you if you have a union and union can negotiate that's the equivalent of an ownership stake if you don't have the union you want to have our shares yourself as a worker from your point of view are governments aware of this uh fast development of uh i i ai if the governments are already aware of this the militaries are ah i mean the the civilian governments of this the american one certainly is very well for the moment they may be aware of nothing in the real world um but the the the government's always a pretty short-sighted so you have to look for let's say some permanent civil servant type people who may have a longer perspective uh military in my experiences the longest perspective because they really are troubled by what the next generation of machines will be for defending uh countries so i would say no generally no um yeah that's probably the right answer obama did establish a group to to to to think about this he was reasonably forward-looking in this but he staffed his group with many of my friends who are traditional economists and they basically said don't worry uh in the past we we dealt with this with good you know social they probably didn't use social democratic but social democratic policies and we can do that in the in the next phase there is a very last question down there please i would like to ask a very practical question i'm a handicap i'm on a wheelchair can hear so as i was saying i'm on a wheelchair and i will have a very practical question as of today i could be standing on foot with the nasoskeleton which is 150 million euro worth but doesn't give it to me and the misspelling doesn't give it to me well i really would like to walk on my food but on my feet but i don't have the money that that would require in the lottery or something but that's not not a a serious answer i i mean that's going to be a problem every place for all kinds of people the medicines some of the medicines come out in there and they're very expensive and this technology the price the price will decline i think the the price would decline very very rapidly so you just hope that they push this technology hard in the u.s what people with particular ailments have done is lobby the federal government we have a law that's for orphan diseases and an orphan disease was a disease by a small small number of people a hundred thousand people in the world from some genetic defect and there was nobody researching it to be done um and they lobbied the government to put special deals tax breaks for companies the money to buy the medicines that for these people to to make sure that they would be safe so the other thing would be through some political uh mechanism obviously it couldn't be you you yourself lobbying to lobby for in this case would be a speeding up of research in this area with special attention to getting the costs down and i think that way citizens can indeed move the governments who then will incentivize the firms to do things in this in this area i think the question was also addressed to myself well the question was also as to me uh of course technology brings about also reduction in cost obviously for not for the most advanced and most sophisticated certainly welfare should be reorganized and different weight should be ascribed to different policies of policies for people with handicaps for disabilities or for families with problems this can only happen when different priori priorities are recognized as a matter of fact in italy for example there's a strong trying to have more money being devoted to pensions for people that can actually work so this is what i could say that thanking again richard for this very enlightening presentation which is not so only scary personally i'm very much looking forward for this fold laundry machine and and also the ironing i would i would be very much interested in an ironing machine robo so thank you so much you