Starr Forum: Power and Progress: Our Thousand-Year Struggle Over Technology and Prosperity

hi good afternoon everybody welcome to today's MIT Star forum power and progress our Thousand-Year struggle over technology and prosperity I'm Evan Lieberman uh professor of political science and director of the center for international studies which is hosting this event and I'd like to thank all you for joining us today so the title of the talk is taken from the latest co-authored book by my colleague jeronasumoglu an Institute Professor here at MIT and the questions raised in this terrific book could not be more timely what is the value of new technology for The Human Condition should we Faithfully trust that labor-saving ideas will make us all better off how might social and political institutions mediate the effects of these new ideas as in his other terrific books and this one assamoglu once again reveals a great sensitivity to the importance of social and political power and structuring how we live and exist on this planet and in a context in which generative Ai and other Technologies are diffusing with such incredible speed we really need to listen to what he has to say so before we get started I'd like to remind everyone that the book power in progress is for sale in the lobby and I know that a lot of you have already done that the line was great uh so the sales have been good please pick up a copy and after the event uh please feel free to bring it up to the stage for a book signing um in addition as per our custom we're going to conclude with a question and answer from the audience and we'll please ask you to line up behind the microphones um to ask your one question and I have an emphasis on one and on question which is not always uh uh heated but I will try as as moderator to enforce that um so without further Ado let me introduce both today's speaker and our discussion so duronasomogly really doesn't need introduction but I'm but I'm here to do that so I will um he's The Institute professor at MIT and an elected fellow of the National Academy of Sciences the American philosophical Society the British Academy of Sciences and a member of the group of 30. he's the author of six books including the New York Times bestseller why Nations fail Power prosperity and poverty which was joint with James Robinson the narrow Corridor State Societies in fate of Liberty with also James Robinson and this book which he's going to talk about today his scholarly research covers a very wide range of areas including political economy economic development economic growth technological change and equality labor economics and economics of Networks um are discussing today um who I am sure did not rely on chat GPT for her comments right they are completely original okay good um is fatini Christian uh the MIT Ford international professor of the social sciences and my colleague in the department of political science she's also the faculty affiliate of the center for international studies where she's director um well should effectively of the center for international studies and she's also the director of the socio-technical Technical Systems Research Center at idss and she's chair of the doctoral program um in Social and Engineering Systems um at the Schwartzman College of computing so I'll look forward as I'm sure all of you will to her comments but first please join me in welcoming jeronasamoglu to the podium [Applause] thank you very much Evelyn and thanks Michelle for organizing this it's a true pleasure to be here uh and uh I would like to point out that my partner in crime here Simon Johnson is somewhere here oh there it is there he is yes so we have we have the double bill here uh so uh it's uh it's my pleasure to be here to share some of the ideas from our new book and uh here is the book and the title and I think today after so much hype and discussion about advances in AI especially generative AI tools such as Chachi PT I don't think we need to give a big introduction that there are tremendous and consequential changes in technology but part of the reason why we have written this book is because we think there are some critical questions that need to be asked Whenever there are new technologies which is often in human history and those are about the control of Technology who controls technology and how that actually shapes who will benefit in fact for transformative tools such as generative AI these are particularly important because there are so many different directions in which these Technologies can be developed and it is quite possible that they could bring broad-based benefits or that they might actually enrich and Empower a very narrow Elite in fact throughout history there are examples of very consequential decisions being made following the visions of powerful agents today those powerful agents are the optimistic techno technology leaders in places such as Silicon Valley in the past they may have been different people for example this gentleman ferdinandaleceps was both a leading Technologies of his day and perhaps the techno-optimist of his day because of his the big belief that the world had to be opened up with big investments in public infrastructure and when other people thought that this couldn't be done or couldn't be done in a way that would really enable ships to freely flow from the Suez Canal he single-handedly pursued that dream he convinced others to come on board he called for big technological advances when others thought that his schemes wouldn't work and he was very successful becoming one of the most famous figures of the second half of the 19th century but his belief in technology and his own understanding of where technology would go then made him blunder with huge consequences in the Panama Canal where he completely ignored the science and the conditions on the ground and his schemes completely collapsed leading to the deaths of more than 22 000 people now these sorts of Larger than Life characters having a huge impact on where technology goes again is not something that we are unfamiliar with today but even those things you might think are not so important because you might imagine that whatever these leaders of Technology decide there may be very powerful forces that are ultimately almost automatically going to bring broad-based benefits and in fact a very critical part of that has to be through the labor market most of us earn our living through the labor market and if any technology is going to create any type of broad-based benefits it must somehow lift people in the labor market here the economics profession is very optimistic in some sense it is so optimistic in uh in fact that a very important proposition in economics doesn't even have a name because it is so much part of our Canon so Simon and I had to invent a name for it and we called it the productivity bandwagon and it goes something like this technology improves meaning our capabilities for doing things for example in the production process gets better as a result productivity meaning how much we can produce increases and this creates a series of economic forces in particular firms wanting to go out and hire more labor because they have become more productive and Via that channel and through the workings of the labor market wages increase and as a result workers benefit that type of story is the way that most economists think about why it is that over the last century for example as productivity went up wages and employment went hand in hand and brought pretty widely shared benefits but if you go back in history you will see there are many other examples in which things work out in a much more complicated way here we provide pictures of two of those one from the Medieval Era the other one from the early industrial period the one on the left is a medieval technological breakthrough really revolutionary in terms of its conception and what it did to the production process windmills that massively improve the capabilities in many sectors but when you look at the data you don't see the windmills creating this sort of productivity bandwagon that lifted the living standards of the workers or The Peasants at the end instead what you see is that a very narrow group of people the landowners and the uh and the clergy who controlled land and the production process were the beneficiaries while the working conditions of of Farm Workers did not improve much on the right you have even a Starker technological transformation that again uh has a more complex effect Eli Whitney's cotton gin which at one Fell Swoop turned the U.S south from a complete economic Backwater to the largest producer and exporter of cotton and fueled the critical phase of the British of the of the Industrial Revolution and really created huge Fortunes in the U.S South but the workers who actually did the cotton production the enslaved black people did not see any benefits in fact their coercion increased intensified they were moved to the Deep South where working conditions were harsher and longer and again no sign of a simple productivity bandwagon but you might actually think that those two are hand-picked examples and the bigger process that really defines our age the one that started sometime in the middle of the 18th century with the application of scientific knowledge or an artisanal knowledge to the production process the Industrial Revolution is very different after all when we talk about the dangers of AI or other automation digital Technologies for inequality for wages one response that Simon and I get is well are you saying that this time is different and in fact this is the reason why the subtitle of our book is our thousand year struggle over technology and prosperity No in fact we're not saying that this time is different this time is very similar to what went on in the past there has always been this tension about who controls technology and whether actually the gains from technology are going to be widely shared so for that let's turn to a re-evaluation of what happened in the British Industrial Revolution and the story turns out to be much more complex than the simple story that Industrial Revolution increased our productivity and we are all the beneficiaries of it yes indeed we are immeasurably more prosperous healthier and more comfortable today than people were 300 years ago but again there was nothing automatic about it in fact the path to that Improvement was much more circuitous the early phases of the British Industrial Revolution were characterized by something that's going to have resonance for today as well automation meaning the application of technology for simplifying and reassigning tasks previously performed by humans to machines this was visible in the textile industry especially in the with the weaving processes and weaving which used to be done by people with hand looms or in their houses became something that migrated to factories did that improve productivity yes the evidence is very clear that productivity increased as a result of that but in fact the benefits of that again were not widely shared the evidence is not here but most of it suggests that real earnings of workers in in Britain during this period did not increase while their working hours probably increased by about 20 percent so their hourly real wages but even worse the working conditions of people were much harsher much less autonomy much less Independence for workers in factories than they had as independent Weavers and working conditions are only one part of the story of course people also care about where they live their health conditions the pollution around them and in all of these the situation was decidedly not so good for the British working people so the the factories were emblematically much harsher places than they were used to this is captured by symbolically at least by the pictures we have on the left which is the idea of a very famous economist Jeremy Bentham the panopticon famous from Michel Foucault's writings or the movies of the Guardian of the Galaxy but the way that Michelle the uh Jeremy Bentham thought of this was a highly efficiency improving uh technology because it would enable employers to monitor workers better or teachers to monitor students or guards to monitor inmates better and what could be wrong with that but of course that was actually what employers were very interested in doing in modern factories to the expense of workers who were forced to work very long hours under very harsh conditions at the same time the cities in which workers concentrated became complete cesspools much less healthy much less uh comfortable living places and life expectancy during this phase at Birth may have fallen as low as 30 years old a terrible number at a time when conditions economically were actually improving for uh for for for factory owners for example now not the end of the story and there is some truth to people who say that look we have benefited so much from the Industrial Revolution indeed we have but the process for which needs to be understood how it is that we benefited from it after this early phase and this early phase was not a short one it may have lasted by about 100 years the beginning of the Industrial Revolution is not clear but you may date it to around 1750 and by the 1840s conditions were still very harsh not just for workers in textile factories but in every sector of the British economy including coal mining another one of the dynamic sectors of the economy where children as five as as young as six were working 18 hours under hugely harsh conditions deep in mind but of course you might think this is all this all three today is different yes and no today is different because today we are in an economy in which for a while we got used to a very different type of sharing of the gains and these two charts that I have up here summarize both the ways in which today is different but also the ways in which there might be some parallels to those older not so good times what I'm plotting here is for men and women separately and for five education groups starting from workers with less than High School in uh dark orange all the way up to workers with a postgraduate degree in dark blue how their real wages have evolved over the last 60 years and in 1963 everything is normalized to zero so you can follow the cumulative change of the wage uh profile for these one for each one of these 10 demographic groups what you see in the 1960s and early 70s is actually a continuation of a of a trend that is also visible from other data sources in the 1950s which is one of shared Prosperity real wages for all 10 demographic groups that I'm showing you here are growing in tandem more or less on top of each other from the fact that all 10 of these curves are growing very sharply in fact they're growing very very rapidly about two and a half percentage Point percent every year in real term which is a really remarkable rate of growth at that rate of growth starting from poverty in two decades you can reach a sort of a much more comfortable level of living but that period of Rapid wage growth and shared growth comes to an end sometime around 1980 or the late 1970s from then on you see a much different picture these curves are Fanning out indicating greater level of inequality much greater level of inequality but even more jarringly you see that the real wages of low education groups especially men but also for women are actually declining so the green line which is for high school graduates the orange line which is for high school dropouts are actually sharply declining from their values in 1980. not only the gains are not being shared but some groups are losing out so this picture therefore poses two sets of questions how was it that the U.S economy in fact it turns

out much of the industrialized World reached some sort of compact in which there was this fairly shared and rapid growth in the late 40s 50s 60s 70s and why did it come to an end so let's try to understand both of these and let me skip this by which shows that inequality is also increasing in other countries as well let me skip that and instead just get to the bottom of the theory that Simon and I tried to develop in the book which is that the productivity bandwagon is not a force of nature that applies under all circumstances automatically and with great force but it is something that's conditional on the nature of technology and on how production is organized and how the gay the gains are shared in particular what changed in the second half of the 19th century in Britain and then continued in the United States and uh and and much of the industrialized world in the first part of the 20th century and came into even greater fruition in the three decades that follow World War II was this process of shared Prosperity built on two big pillars one is new tasks and the other one is worker power of course both of these terms need to be defined and they may be a little bit simplifying but the new tasks are critical automation that I mentioned which is the substitution of machinery and today algorithms for the labor of humans has always been with us or at least uh has been with us in great force since the beginning of the Industrial Revolution automation is a major Force for increasing productivity but it does not create the type of shared Prosperity by itself because after all what automation is about is to take tasks away from workers and have more Machinery do it so it reduces the importance of Labor it also reduces the labor share in output or environment in national income so if we're going to have shared Prosperity automation needs to be coupled with something else and that's something else critically turns out to be new tasks new activities in which human labor is critical that reinstates workers centrally into the production process so throughout the second half of the 19th century and the early 20th century we see this as very important determinants of what's going on in the labor market emblematically for example captured by the picture up there which is from the uh from Henry Ford's motor factories Henry Ford was a leader in applying new technologies including decentralized electricity and and and assembly line type Technologies early on and that was absolutely revolutionary completely changing the car production making cars affordable for the masses and a very important part of that is the use of Machinery to do tasks that was previously done by labor automation but if you look at the Ford factories that wasn't the only thing that they were doing at the time at the same time as they were introducing automation Technologies they were also creating new tasks for workers so it is no surprise that in this picture what you see is the Advanced Machinery together with the workers workers are now performing more technical tasks they're operating taking this Machinery their their engaged in design inspection and and other quality control activities and if you actually look at the factories of the early 20th century you'll see that production workers are joined with non-production workers clerical workers that are very much engaged in planning and other aspects of the production process it was this double uh uh process that was so important for the beginning of shared Prosperity automation which increases productivity but also new tasks that give another boost to productivity and also create reasons for workers to share in those gains but even new tasks by themselves are not enough because if workers are making a major contribution to productivity but they don't have the power to take a share of that they may go down the same path as the black enslaved workers with the Eli Whitney cotton gin it may not be in the interests of the work of the firms to share those gains with them and they can get away with it because they have all the power so actually a balanced distribution of power in workplaces and in society is also a critical part of it that's why the second picture on this slide is the one of the emblematic strikes in the Auto industry that was still a leader in establishing the labor movement during this period the United Auto Workers strike at General Motors in 1937. so it was this double process that was so important but not just for the early car industry but in the 19th century the second half of the 19th century that I've been referring to in Britain what was so distinctive about it when you think uh about what went on the direction of Technology change completely new set of Technologies in Railways in steel and chemicals that were much more important for improving the productivity of Labor and introducing new tasks for labor this was embedded in a society that was democratizing from the times of the early 19th century where even the middle classes didn't have the vote now Universal uh male suffrage and then universal suffrage came to Britain and also the labor movement which was completely banned and heavily prosecuted uh up to the last quarter of the 19th century became a staple of British workplaces and was a very important part of improved working conditions and wages now this was about why shared prosperity's foundations were late starting sometime in the middle of the 19th century and then continuing to the 20th but then why did it come apart sometime in the 1980s and I think the same two uh processes now playing out in Reverse are the key actors in this and to explain that I'm showing here a modern car factory which looks somewhat different or quite different from the Ford one you again see the Advanced Machinery now the Advanced Machinery takes the form of Robotics arms but conspicuous in its absence are the workers the workers are no longer playing a central role the automation is rapid but the new tasks haven't accompanied it so too much focus on automation but not enough on creating new tasks is the technological part of it but accompanying that has also been an Institutional change and that institutional change oh sorry before that let me actually show you this figure uh to substantiate the claim that I made this is a figure from uh work that I have done with Pascual Restrepo what it shows is something that's akin to the first graph that I depicted the evolution of the real wages of the 10 demographic groups now a little bit more detailed demographic groups distinguished by age gender education and ethnicity each one of these circles refers to one of them on the vertical axis I'm showing you the cumulative change from 1980 to 2016. so that period in which some groups were experiencing wage growth other groups were experiencing wage decline you can see the the same thing from here the from the fact that many of these circles are below the zero those are the demographic groups that are experiencing wage declines and on the horizontal axis I'm depicting the extent of task displacement that a demo a demographic group has experienced during this period namely what fraction of the tasks that this demographic group used to perform across Industries and across occupations in 1980 have since been automated so you can see that for some groups mostly those like us who have postgraduate degrees or very high levels of specialized skills those numbers are very close to zero we have not really suffered much automation of the tasks that we used to perform that are much more creative much more problem solving and and high level but if you look at those for high school education or less than high school education demographic groups shown for example by purple and green you'll see that up to 25 30 of the tasks that they use to perform have since been automated and those are exactly the groups that have suffered the wage declines in fact this regression line explains about 60 to 70 percent of the variation inequality between groups in the United States so this is the automation part but automation has had a very big effect and in fact itself its path was very much shaped by institutional changes in the U.S labor market and those institutional

changes have been in the direction of declining worker power and to understand declining worker power in this period in the same way that if you wanted to get to the details of understanding the increase in worker power in the 19th century you would need to think about both how ideology has shifted and how organizations have shifted and both of those have gone against workers during the post-1980 era one is perhaps I'm giving too much credit to our fellow Economist Milton Friedman but the rise of new corporate Visions which elevated managers doggedly working for shareholders and ignoring everything else this was the beginning of the shareholders value Revolution or Milton Friedman's statement that the only social responsibility of business is to maximize its profits and that was coupled with the uh the erosion of worker powers for example emblematically during the defeat by Ronald Reagan of the professional air traffic controller strike so these two changes together both shaped the way that managers wanted to approach how to run their business for example monitoring workers more tightly or automating in order to cut labor costs was completely welcome because it would increase the returns to shareholders but also there was no resistance to them from organized labor because organized labor was getting weak during this period now this is all before AI can the age of AI change it yes the AO the age of AI can change it because at some level if you look at the details the promise of generative AI some of it is hype but some of it is reality is that it can actually be a tool in the hands of workers but if you look at the reality of it you also see major roadblocks towards that kind of change what makes today such an important point in this type of discussion is that there are transformative and very consequential choices ahead of us and again Simon and I think that one way of framing this is in terms of different Visions One Vision about where Ai and digital Technologies in general are going to go it's emblematically summarized by that picture at the top is the or the Turing test is towards autonomous machine intelligence meaning machines become more and more autonomous more and more intelligent and they start doing more and more of the tasks that humans used to do it won't take much imagination to see that if that's the emphasis we're going to have a lot more automation but in fact if we are right that automation doesn't create the foundations of shared Prosperity that spells trouble but that's not the only way in which digital Technologies were conceptualized and it's not the only direction in which AI can go long ago many computer scientists understood a very different way of using Technologies which Simon and I call machine usefulness to contrast with machine intelligence the objective is not to make machines intelligence in and of themselves but more and more useful to humans Engineers such as mit's Norbert weiner jcr lichlider who was briefly at MIT as well or Douglas engelbart try to articulate both the philosophical foundations and the technological realities of this vision and out of this came many of the technologies that we rely on for example when you use your smart form the menus or the computer mouse which was revealed by Douglas engelbart in a very famous event called the mother of all demos or hypertext all of this came from an effort to make machines more usable and more useful to humans and in fact AI could pursue that path now the problem in fact is not just one of distribution if you over emphasize automation it's not that you're going to get huge productivity gains and they're just going to be unequally distributed in fact there is every danger that over emphasizing automation is not going to get you much productivity benefit either and this is the concept that Pasqual Restrepo and I and Simon and I try to capture with the with the label so-so automation you know what you're trying to do is you're trying to get machines to do things that humans are pretty good at so when you do that you don't get a huge productivity boost because humans were doing it fine but you get big distributional benefit or will be still institutional costs because you're sidelining humans firms may become a little bit more profitable but a lot of workers lose out and self-checkout kiosks or excessively automated customer service all of those are examples of social automation where the productivity benefits turn out to be not so much as people were hoping now of course generative Ai and chat GPT could change all of that so we ask chat GPT itself whether generative AI could reverse these things on this one I think chat GPD was quite On Target perhaps but probably not it's not a magic solution if generative AI is used to replace workers instead of support them it could have negative consequences now we don't know whether it knew the answer that we wanted to hear or it read some of our papers but we agree with this answer labor market consequences and inequality are not the only things we have to worry about one of the other uh Trends since the 1980s but again accelerating with generative AI is about who controls information after all even in the industrial revolution it wasn't just automation it was also how the modern factory system changed the method of control and who was in charge and what they could dictate so one of the things that we are seeing with more and more digital Technologies is surveillance surveillance in workplaces surveillance and political views now it takes different forms in the united in China you may be more worried about it because it's in the hands of the government the social credit system or facial recognition per cameras everywhere where protests could one day break out in the United States as companies as Google Facebook Amazon that have all of this information but at the end of the day in the book Simon and I argue that both of those are pernicious it doesn't matter who has your control as long as that information can be used without any constraints it's going to be anti-democratic and it's going to be inequality inducing this has so far been a lot about the developed world I've given examples from the US a little bit about Europe and in fact that's a natural Focus for Simon and I because we want to trace the counters of new technologies and how they are used but let's not forget that new technologies that are developed in the United States and in China are going to be used throughout the world and in fact the international division of labor is already being reshaped by automation Technologies one of the patterns you see around the world is that a lot of routine activities that were automated in the United States are also being automated around the world or at least the amount of activities or labor that are that is assigned to these production functions is declining there is every danger that AI if it goes down the automation path could be a highly unequalizing technology around the world and again some of the surveillance implications are Global as well the recent work by some of my colleagues here Martin bereja and David Yang and others shows that Chinese companies are already exporting anti-democratic monitoring Technologies to more than 60 non-democratic countries around the world so this is all potentially depressing because it says there are big dangers of body inequality and democracy but from the beginning I tried to frame this as saying well these are transformative choices because there is no necessity that AI is going to go one way or another there is a high degree of malleability to all Technologies and that's doubly true for AI and if we make the wrong choices they could have damaging consequences if we make the right choices they could be much better for society that could be much better for workers that could be much even better for democracy so the question is what we can do Simon and I in the book suggest that we need a three-step process for thinking about change the first one is changing the narrative our modest hope is that this book is a small contribution to then changing the narrative moving away from kind of blind technological optimism everything is going to work out this time was no different to try to understand how things have worked in the past when things go right when things go wrong and part of changing that narrative is to all also recognize that things are more likely to go right when there are more voices rather than technology being in the hands of some powerful actors be it fernandezps or some outman and Elon Musk so the first is a change in narrative but it's changing the narrative is not worth that much unless there are institutional and other uh developments that actually turn that into actual action and policy changes so that is what we mean by countervailing powers so part of the reason why things were different in the 1950s and the 1960s because was because technological choices and how those choices were how those gains were being shared was embedded in an Institutional framework where government regulation was important and where there was civil society and labor movement constraints on what companies could do some of that needs to be recreated so it has to be in the form of a new labor movement perhaps other forms of bottom-up organization for Civil Society and also government regulation especially in the field of AI and the central idea here that Simon and I emphasize is that redirecting technological change has to be a major part of both the efforts of the labor movement and of government regulation technology has many potential directions in which it can go and there is no guarantee that the completely unfettered Market process is going to choose the socially beneficial Direction there were many people such as this gentleman Ted Nelson uh not just Douglas Angel Bard and others who thought that the personal computer and other digital Technologies would be fully liberating choices both for workers and for Citizens at the end that's not the path that we ended up on but that was a choice it wasn't because they were completely wrong in thinking that technology could be a decentralizing force it could be a tool in the hands of the of the of the of the of the workers not just of Corporations and in fact Ted Nelson very much anticipated this when he was riding on the one hand he was optimistic but on the other hand he was very much emphatic that uh large corporations such as IBM would try to control the technology and that would push it in a very different direction but all of this raises another question which and I will end on this you know there is some degree of optimism in saying that we can actually redirect technological change and that's a social choice because the counter argument is no technology is a fully organic process every time you interfere with it it's going to end up in your face well of course it is an organic process but the fact that it's an organic process doesn't mean that it cannot be steered within bounds and one example where you see that is in the energy sector which is not of course something to be proud of we are very much behind in combating climate change but today if you look at our ability to generate clean energy it's uh you know miles apart from where it was in the mid 2000s or even in 2010 uh for example various different types of solar and wind Technologies are today cost competitive with fossil fuels whereas they were about 10 times as expensive as fossil fuels as late as as recently as 2010. how did that happen it happened because there were some subsidies and some regulations about clean energy and it was also a civil society movement more pressure from consumers who wanted to clean our products more pressure from Civil Society for companies to clean up their act and even a modicum of that type of pressure led to a complete redirection of technology and the reason why we end the book with this example is precisely because Simon and I think this can be done in the in the realm of Production Technologies with even more consequential consequences than energy perhaps thank you foreign [Applause] hi everyone um first of all I want to thank CIS for the opportunity to engage with that on this terrific book co-authored with Simon it's great to have you in the room uh it's a true honor to be here as you saw from this terrific presentation this is the book about the forces of technological change but also all the channels challenges that this ushers in and I mean it's it's an intellectual tour de force taking us over a thousand years and bringing it to the very present focusing on the challenges of digital Innovation and the new era of AI so unsurprisingly and now you know it's not chat GDP gpp the first thought that came to mind was how relevant these topics were even in ancient times so it's not just the Thousand Years but these were real preoccupations also for the ancient Greeks and in their myths and in their philosophical writings they were really challenged by this idea of the wonders of innovation and technological change and what they meant for creating societal hierarchies and structures but also a lot of cautionary tales of kind of their responsibility that comes with these great wonders and these great capabilities so I if you may indulge me I'll touch on three myths that came to mind and I hope I'll briefly travel us back to the Greek mythology books that we all engaged in childhood so the first is the myth of Daedalus and he was a master Craftsman and inventor that was brought to the island of Crete by King Minos in an effort to house the Minotaur the half bull half man that was a pretty dangerous being and of course you know the Innovation there was the Labyrinth that enabled the housing of the Minotaur but it was also it created a big inequality and that was in the form of the human sacrifices of the Athenian youth that were for whom this became a death trap uh the the actual Labyrinth the second myth is the related story of Icarus the son of deadless after the Craftsman this great Craftsman fell out of favor of the king he was uh closed up locked up in a tower with his son and of course he wanted to engineer their Great Escape and he created wings out of feathers and wax and while they were flying out and were actually making their trip back uh ikarus was not listening to his father and decided to fly very close to the Sun and which actually led to his uh kind of plummeting in Via GMC and dying so what was interesting I mean this is seen often as a myth about hubris and very much hubris around the invention and Innovation and I think thinking to contemporary times and kind of Youth and hubris I thought of cases like terranos and FTX that I think are quite associated with some of these startups and the third and last Greek myth may be the closest to what we're discussing here is uh Prometheus and his decision to steal the fire from the gods and share it with them humankind and in that sense Prometheus was kind of the ultimate equalizer maybe a union organizer of his time of the Titans um deciding to kind of challenge Elite dominated decisions over technology and he took this great leap but he was also punished very very much so I mean he was changed Eternal punishment being changed on a rock and having an eagle eat your liver every day and this would go back and every day it would get rejuvenated and eaten again so I wonder if there is an interesting lesson there or kind of that courage and self-sacrifice that may be required in order to rein in some of these really big interests so and apart from mix I mean Greek philosophers were also very preoccupied with these themes played on the Republic talks about a control technological environment where the main goal is the pursuit of the good life implying that some of these Technologies are actually distracting from this goal and then Aristotle in turn acknowledging obviously the great importance in society of having Techni I mean craftsmanship and Technology but distinguishing between technologies that sustain life versus those that actually enhance living which I think is an interesting distinction um I mean so this is clearly a relationship that uh that has been kind of a a great topic of importance uh throughout life and I was hoping to post three sets of questions I hope Evan will not block me uh but uh this is kind of a bit of a warm up for our discussion I'll pause all three and then you can take them uh one by one uh first the book tries to suggest that there is kind of a certain heightened energy and urgency about digital Innovation and AI that maybe did not exist with other past cases but I wonder if this may have been what it felt like at the time for all the others you know what it felt like at that particular point in time for you know what is different actually between the Gilded Age Rockefellers Vanderbilts and you know the Bill Gates Jeff Bezos Elon Musk of our times why is the invention of railroads and steel at that time so different from digital Innovation and AI now it seems like you know in every past case in history and I don't know if that makes me a great Optimist but humankind kind of in the long run benefited from this technological change managed to improve overall standard of living averted disaster moved on and then went into a new cycle of innovation so why and how is our case now different and I I particularly wonder about this in the context of the work around climate crisis in the environment which will highlight very articulately in the book and which seems to give us a certain kind of road map and recipe of how to think about this and how to move forward my related second question is whether it is technology really that creates the inequality or lack of proper governance over technology and specifically this is a question about understanding how the how technology relates to societal economic political structures and basically kind of a role of the broader ecosystem of the cities and the state and other institutions in addressing the these inequalities so I'm I wonder for example what does a union an effective Union look like in the 21st century and I'm particularly interested in especially now I think it's President Biden just joined the picket line in Michigan where Union Union the auto United Auto Workers Union is on strike against you know the three big Auto producers here so what does it mean to have an effective Union in the 21st century and what it does it mean for Collective action among citizens to to to be um effective uh in this in this case so there's a new book by Frederick the Boyer that I was just reading about on how Elites ate the social justice movement this is the title of the book and he talks about the failures of some of the most recent movements like the metoo movement the Occupy Wall Street movement the black lives matter movement he basically claims that they may have succeeded in symbolic change among a lot of kind of the academic Elites and and and and and some of the more Bourgeois citizens here but they didn't really manage to make changes for the average citizen and then he says that a lot of it has to do with the fact that they didn't get any legislation actually passed on these issues and uh and he also attributes some of the failure in the the way these movements were actually structured which is very different than some of the movements that we know from the 20th century like the Civil Rights movements for instance um Beyond citizen movements and unions I was wondering about the responsibility of State more actively in terms of Regulation and the creation of social safety nets I think there are some some people out there who may even argue that the state has already been co-opted and what is regulating is in favor of the interests of these big Elites and corporations and not really actually regulating in ways that are protecting uh uh people and citizens in terms of these inequalities also what is the responsibility of international institutions we know Europeans are very keen on regulating on these issues on the EU level so can we think about this differently in that context or academic institutions like here I mean you made it clear that a lot of us that actually get get money from these big tech companies need to be very clear about you know where our independence May potentially end or be compromised and I wonder what is the role of other initiatives like ethics and social responsibilities of computing initiatives or other efforts within Academia to try to think about how to to move Beyond these inequalities and as the last question this is a little bit more about kind of the democratizing of of Technologies and this idea that they have also flattened inequalities to a certain degree so for instance there are some instances in the global South where they have been faster adopters of Technologies than we have and how this the openness of digital education information has actually enabled them to make big strides they've also been very keen on some of the AI Healthcare tools for instance and that they've taken on in some some of their National Health Systems um and I and even you know among Educators among creators among artists one can see how technology has been an enabler and can we say that maybe we can accept that there is there will be a certain level of inequality that is required to keep wanting to have change and move on towards progress and how should we look about this look at this in terms of the global South and the global so sorry literally just to open up some parts of a conversation and we could even just turn to the audience right away thank you Fortini thank you for this very erudite and uh extremely far-ranging comments and fantastic questions but I will have to be brief let me first say that uh Simon and I used to have Prometheus as the beginning part of the book but and try to tell the story of responsibility actually Prometheus himself is sort of a conflicted character hebristic and responsible at the same time but then we decided a was going to be a little bit tortured and B we would probably be caught by somebody who knows mythology much better than we do and then Oppenheimer I feel had already taken from we also had Oppenheimer and zillard he's had the contrast between Leo zillard with if we had known that the movie was coming out we would have kept that Prometheus but but absolutely you're right and that's part of the reason is we think these issues are as old as Humanity that there were concerns as early and even probably earlier than ancient Greek civilizations about control of technology so they are very much with us coming to your questions you know it is possible to read history by saying look you know we've had some hard times that but then we we bounced back but Simon and I don't think that's the right story partly because a there are many examples in which we did not bounce back or even when we even when we bounced back it was far from automatic you know I think you cannot tell the story of the Medieval Europe as one of bausback uh it there was no sort of process towards sharing of those gains with Farmers or the with the with the with the farm laborers until the whole system collapsed and when you look at you know the two examples that you hinted at the British Industrial Revolution and uh and and the rapid industrialization with new Industries in the United States in both cases there there was a very very radical change in institutions and that was very far from a foregone conclusion the Progressive Movement I think even its participants were surprised that it could actually form that Coalition and and and succeed and the the the the cars were very much stacked against democratic reform the labor movement and all of that in in Britain you know one example we mentioned briefly in in the book because it's so telling that in the 1840s the chartists in the UK you know collected 3 million signature I mean can you believe three million signature in a time when you know you don't have any of the modern communication Technologies with the demands that are so tame in terms of you know universal suffrage and some basic sort of Rights and they were so careful not to be labeled socialists and and the response of the parliament was completely turned down all of these demands and jail all of the chartists so there was no automatic institutional process that could say oh yes we're going to bounce back from this so that is the sense in which you know uh there is both quotient and optimism in what Simon and I are telling the quotient is there's nothing automatic here but optimism is yes it is possible to do it but it's not something that's going to happen by itself and when we come to the question of is this technology or is it institutions in fact this is a big debate in economics uh inequality what's the role of Institutions what's the role of technology and at some level it's a false dichotomy it's a false dichotomy because I think first of all the two interact in very complex way but most importantly and this is why we put so much emphasis on what the objectives of the corporations were was the vision of the tech industry is you know it is institutions regulations and social norms that shape the direction of Technology but we are also insistent and this is partly you know the research that I've been doing for you know many many years that technology does have a really important role in uh in understanding inequality so there is a sort of a left-wing narrative for example that everything is just about the union movement and and decline in minimum wages but the facts just don't line up with that and that's the reason why I showed that chart that chart is rather striking because it shows just the automation part of technology which you know Simon and I think is the most important one really is centrally important in explaining what's been going on in terms of inequality in the United States in other countries it's a little bit more of a complex picture because institutions are very different and they interact with technology but the the direction of Technology especially the automation focus is absolutely Central and that's why redirecting technological change has to be quite important yes indeed Railways and steel and how they were used was critical again that's the redirection of Technology but again going back to our overall interpretation even that is not a foregone conclusion companies are going to have many different choices about how to use Technologies and the reason why we are so insistent on digital Technologies is because they have been used in a very particular way the the conception of people like Norbert weiner jcr lichlider Douglas engelbart Ted Nelson is precisely our emphasis is precisely because there was a different path of how to use these digital technologies that was more targeted at improving the productivity of workers and therefore as a result would have been more beneficial for workers but at the end digital technology is being used for automation being used for surveillance I think that's those are the choices that we have made and those are very consequential choices yes indeed there are parallels from the past but I think digital Technologies really amplify these things and if we come to AI I think the parallels again this is why the book is sub-entitled a Thousand-Year struggle but there are also some unique features of AI one is the speed of change you know to the to the contemporaries the introduction of textile mills would have appeared completely revolutionary but even then it was a slow process over many years the you know the for example when you see like the weaving industry modern factories are coming up but you know there are still thousands of Hendon Weavers it takes you know several decades with AI the speed of change is very fast and it's also very pervasive that potential again some of it is hype but the potential is there that AI tools are going to be applied across many Industries so that I think raises the stakes both about redirecting technological change and the regulation so that's part of the reason why we think this is an epocal time in terms of making the right choices and it is also the reason why when we come to your question about the labor movement or the general regulation we think the focus has to be on technology or at least a very important part of the focus you know today we are going through a spring of strikes some of it is understandable there is a a pro-union pro-worker president uh wages did not increase while unemployment is low and it is an interesting time but with the exception of the wga the Hollywood strike I think very very important absence is the discussions about technology and how technology is going to change things you know UAW you know their attitude so technology is just say no I mean you know if if you don't come up with that solution about how we can transition to electric vehicles but still make that better for workers I don't think the union labor movement is going to be successful or let me put it another way you know how many leaders of the labor movement has really invested in understanding Ai and thinking about how it is that the AI can be used in workplaces while that's good for the labor movement I think that's the that's the part of it in that sense the wga was very uh trailblazing uh because they made AI as one of the key topics and and they have the power because of their visibility because of their special skills to actually have ever say and and in fact uh probably this agreement is is is is is is is uh is a success for the writer's Guild of America but even in that case I think the the wga needs to be congratulated on making AI such a focal issue but they did not come up with a positive vision of how to use AI I think ultimately the labor movement and you might say this is why we may need a new and re re refashioned labor movement needs to come up with new ideas about how to use these new technologies that's good for workers and ultimately acceptable for bosses so the wga in my opinion you know should have uh articulated a vision of using AI That's good for its members that's good for the productivity and the quality of the entertainment industry that was different than what the Hollywood Studios wanted and I think that would have been much more powerful I think that's what we need from the labor movement and the broader things you know the Boyer book yes I think you're 100 right uh in fact Simon James Robinson and I wrote an article uh published by of all people in an edited book by Michael Lewis on the Occupy Wall Street saying at the time that it wouldn't amount to much precisely because it did not have a plan for institutionalizing so you know the progressive movements success is that they wanted to change politics not just protest and I think that's what's missing and the labor movement or a broader Democratic movement targeted at changing regulation changing labor market institutions changing technology I think needs to be embedded in a bigger institutional setting and then the final thing I'll say about the international dimension I think the international Dimension is key you know after all one of the things that's quite obvious from this discussion is that the choices that are being made in the United States and in China are going to have sweeping complex implications for uh every country around the world but where is their voice I think the biggest issue is that you know if you want to have a direction of Technology that's responsive to Global needs including the needs of you know more than four billion people who don't live in Europe China and the United States then you actually need their voices to be heard so we need International organizations or new vehicles for the emerging world's voice to be heard because otherwise there's a real danger that AI is going to go in an inappropriate direction for their needs and worse it's going to become a more and more powerful surveillance technology that can be Unleashed on the on the populations of these emerging democracies and we're already seeing that thank you for those fantastic questions for teeny thank you drowning and fatini that was terrific but now we have almost a half hour to uh take some questions from the audience so you'll see there are uh two microphones uh a line up in either uh stairwell so if you could just um line up um in those and we will uh if if we have a cue on both we will not discriminate against either side rotate um and again if you could just ask a question in just one um and I see we have someone there so yeah hi start uh so my name is Charles I'm a grad student at Tufts so you showed us some data on wage inequality and how it was brought about by task displacement and a different by educational attainment levels so uh but at the same time we all know that technological advancement brings about huge differences huge improvements in life quality right so for example now we can connect with your loved ones thousand miles away in an instant now we have AC right so I'm wondering would we have seen a different picture than the one presented in the book if we looked at the net effects of Technology then instead of focusing just on the you know income effects absolutely I think the effects of Technology are very rich and complex but I think the general principle

2023-10-09 19:40

Other news