 Good morning, everyone. I'm Tim Briglin, the chair of the House Energy and Technology Committee. It is, it's tax day. Well, it's actually not tax day. I guess May 14th or something is tax day now. But it is April 15th. This is our 1030 hearing. And as we've been doing in recent weeks, we've been dedicating our Thursday mornings to some work on artificial intelligence and a couple of bills that we have before our committee. We have been hearing from some members of the Artificial Intelligence Task Force, as well as some of the testimony that that task force took in doing its work, which was completed about 16 months ago, I think. And this morning, we are fortunate enough to have Professor Stephanie Seguino, who is a member of the faculty at UVM. And I have to confess that in spite of my day job, I've never taken an economics course in my life. And so this is kind of a double benefit for me today. When Professor Seguino had testified before and had done some work with the Artificial Intelligence Task Force, I became aware of that in reading through some of the notes and minutes from the work of that group. And I think it was, I'm going to say, in December of 2018 or 19. I can't remember exactly the timing. But it was very interesting to me. I thought it would be interesting as just foundational knowledge for this committee and just an interesting area of discussion. We have not, Professor Seguino, had the benefit. And I'm still kind of searching for who that resource is for our committee, but to kind of dig into some of the commercial benefits, if you will, from Vermont. I know the task force had looked at some of the areas where this can be an exciting area for job creation and job development. And we had Donna Rizzo from UVM testify a few weeks ago. And I know that she's done a lot of work at UVM, both in a teaching capacity, also as a research and a research capacity in this area. And so hopefully we will have more kind of testimony in the future kind of on some of those commercial benefits. And maybe you'll testify on some of those today, as well as some of the issues that it poses for our labor force, potentially. So I won't jump ahead. I know that you have a PowerPoint presentation that might be helpful in terms of leading us through the discussion today. And really appreciate you being here with us. So welcome. And thanks for sharing your time with us. Sure. Thank you so much. I'm not yet able to share my screen. I'm not sure what's going on there, Matthew, but. So I think you have coasting ability if you can try to click that through. If you're unable to do it, maybe Matthew can actually pull the slides. Oh, no, there it is. I can see it now. Stephanie. Success. Yep. There we go. We can see it. Thank you all for having me. I just want to just to give you my background, I am a labor economist as well as a macro economist. And so that's the lens through which I view this. And what I'd like to do is see if I have an outline here. I'd just like to give you sort of the context in which this discussion is happening. There have been significant changes in both income inequality, the way we produce goods, and globalization that have created a certain anxiety around these newer forms of technological change. So I'd like to talk about that. I'd like to talk about some of the research that has been done on what the employment effects are of AI and machine learning and a variety of these new technologies like robots, for example. And I'd like to talk to you about what I think the important role of government is in improving labor market outcomes in the context of this. So let me just give you some background with regard to this moment that we're in. I will start out by saying this. And that is that I think oftentimes people think that technology is just something that we live with. There are technological changes. And we are bound to accept those changes. I don't think that that is the case. I think technology needs to be in the service of our goals as human beings. We should not be subservient to technological change. And so I want to frame it in that particular way. So the moment that we find ourselves in is that there has been a significant growth of inequality since the 1970s. This is true for Vermont as well as the rest of the country and quite frankly globally. One of the things that has happened is that we have seen the wages for those who have less than a college education or some college education decline in real terms, right? An inflation-adjusted term since 1979. And for blue collar workers, this has amounted to like a 25% decline in real wages for that group. Well, we've also begun to, and many have argued that that was a function of globalization, that the sort of labor-intensive jobs moved overseas. But what has happened since 2000 is we've also seen the wages of college-educated workers decline. And I'll talk about why I think that's going on. And so although they're, at least prior to COVID, unemployment rates looked very low, Vermont has and does still have one of the lowest unemployment rates in the country, that really is mismeasured because of changes in the labor market. There have been, from a number of people, because they haven't been able to find jobs, especially older workers, have dropped out of the labor force. So that lowers the unemployment rate, but they theoretically would like a job. Were they able to find one? We find that many workers are discouraged workers, meaning they've given up looking, and those workers aren't counted in the official unemployment data. But more than that, I mean, those are significant, but I also think something that makes these processes less visible is that jobs are now more insecure. That is that workers are much more likely to be contingent workers on call, gig workers, but they're not, we've seen a sharp decline in year-round full-time workers with benefits. And a lot of that has been due to corporations trying to reduce the cost of benefits. So many, especially in the service sector, firms will hire workers for less than 35 hours a week to avoid paying benefits. But basically it's been a process of shifting insecure, job in sort of fluctuations in aggregate demand, shifting that burden to workers rather than corporations. And so what we find is that a lot of people are working multiple jobs due to short hours. So let me just talk a little bit about artificial intelligence and robots, which are the really two areas of technological change that I think have been of most concern, although there are other forms of technological change that have mattered as well. And the reality is, of course, the technological change has been a fact of life since the beginning of capitalist economies. That has been the nature of capitalism. Technological change is a way that firms try to improve their competitive position. So they develop a new technology that makes it cheaper to produce the goods that they're producing and they can compete their competitors, but eventually the competitors catch on to the new technology. And this is just a continual process of capitalism. But there's a major difference today than the technological change of 100 years ago. For many years, technological change was really simply about replacing physical labor, so brawn, if you will, and routinized tasks. So that's how we got the assembly line and a variety of other technological processes like that. But what's happening today is that not only are these technologies supplanting physical labor, they are now also supplanting mental tasks. And that is why we see the impact on the college educated, that many of those tasks are now being outsourced to machines. And I wanna just reiterate that the reason that firms take up technology is to increase their profits. That is the raison d'etre of capitalism. That's the reason that they do this. And their profit incentives will determine the kind of technologies that they adopt and develop. And it means that in many cases, new technologies may benefit firms, but not necessarily society as a whole. Just because firms have higher profits doesn't mean that society in general benefits. And I'll explain that in a little bit more detail. And so in general, I think the task of government, one of the reasons I support these two bills is for the creation of this board is that it is fundamentally important for government to adopt policies and regulations that require that businesses align their profit motives with societal wellbeing. And I will give some examples of this at the end here. So let me just talk about the debate about technological change at this stage. As I think many of you have heard, there's a lot of anxiety about that because of the potential loss of jobs. And of course that was Andrew Yang's platform when he was running for president that we are moving into a stage in which there will be job shortages and therefore we need a basic income grant. This was Yang's argument. And some of that earlier on anxiety stemmed from this research by Frey and Osborne who wanted to figure out how susceptible jobs are to computerization. In other words, job automation based on computer controlled equipment. And they did this by trying to understand what problems engineers need to solve in particular occupations. And so they looked at 700 occupations. And what they found was that 47% of US employment was at risk of elimination in the next 10 to 20 years. And so they used measures of perception, dexterity, manipulation, creative intelligence and social intelligence. So these here, the last that I mentioned, these are less susceptible to computerization because computers simply can't solve the problem of social intelligence or creativity and dexterity and manipulation. So they then created a categorization of jobs based on which are least susceptible to elimination due to computerization and which are the most. And so let me just show you that graph. There's a lot of information on here, but let me just guide you a bit. On the left here, where it says low, is 33% of jobs are not very susceptible to computerization. So these are the jobs that are likely to grow over the next 10 to 20 years. And in this, you can see reading from the top of the list here, that means management and business jobs, computer engineering and science, education jobs, legal services and healthcare. Those are jobs that are not threatened so much by computerization and AI as other jobs. And I'll skip the ones in the middle, but the ones on the right are the ones that are very susceptible according to these authors to computerization. So transportation, construction, farming and sales in particular are an office and administrative support. So according to these authors, they frame this then as almost half of US jobs being at risk of elimination, but the jobs least likely, I wanna emphasize this again because it's important, the jobs least likely are those requiring social intelligence and especially creative intelligence. And so this is just some data on what's happened in the United States with regard to low-wage, middle-wage and high-wage jobs. Many of the jobs that are susceptible to computerization are middle-wage and high-wage jobs. And what you can see here really is what we're seeing as a bifurcation of the job market over time. That is that there is a growth of low-wage jobs. There was a dip in between 2005 and 2014. You can see that gray mass below the zero percent line, but the greatest loss of jobs was middle-wage jobs and they showed the least recovery after the great recession. And high-wage jobs were little affected by the great recession and there's been a significant growth. So what this means is greater inequality because you have a much more divided workforce, those that are doing really well than those whose jobs are low-wage and probably have very few benefits. But I think it's one of the things that's important to think about with regard to the work of Frey and Osborn is that it's much more likely that specific tasks of jobs will be eliminated by computers but not the entire job. And so what we're likely to see is that job occupations are actually restructured. And I wanna just give you one example of administrative assistance because it's what I've seen at UBM since I started in 1995. Our administrative assistants used to type our syllabi out and file our work and so on and so forth. We now as professors do that work, but the job tasks have actually dramatically changed for administrative assistance. It includes budget management. They have to learn a lot of online computer programs for budgets and expense reports. They organize meetings and plan events. Similarly with radiologists, we've seen that radiologists jobs are very complex and that according to one study, there are literally 26 distinct tasks associated with being a radiologist, many of which are not suited to AI such as interpersonal skills, for example. And likewise with bank tellers, we've seen a dramatic change in this industry and this job occupation. They now handle many fewer transactions, but their tasks have changed as well. And so I think this should give us pause in assuming automatically that AI is actually going to lead to the complete elimination of occupations. So I wanna talk about where I think the issues are and actually with regard to the issue of what are the benefits to firms of adopting this and how does that affect the rest of us? So the Frey and Osborn and many people who are concerned about AI assume that the quantity of jobs is fixed and that if we eliminate certain occupations, that that means that unemployment is automatically gonna rise. And that is possible, but that's not necessarily the case. And whether the quantity of jobs is fixed or not depends upon how firms respond to the adoption of technological change. So let me give you four scenarios that are possible, some that are very beneficial. So basically, firms will adopt technological change because it raises productivity, that it means it raises output per worker. Maybe it means that you can produce the same output with a lower number of workers. So this is good, it's potentially good for society that that labor could be reallocated to other things. And firms have choices of what they can do with the impact of higher productivity. So one of those possibilities is if productivity rises, it means that it's cheaper to produce goods and services. The firms could lower their prices so we would all gain the benefits of technological change. Another possibility is that firms could share the benefits of the higher productivity by giving workers higher wages. They could also, something that had been long dream to go by John Maynard Keynes and a number of economists, reduce the hours of work for the same pay for workers. And that in both of these cases, what that means is that the firms would be sharing the gains of productivity with workers. And that means generally that societally we would all benefit from technological change. One of the problems is however that there's link between productivity growth and worker wages has been broken. And I'll talk more about why that is so. And the reality is that in the last 25, 30 years productivity has risen and worker wages have stagnated. So there has been a dramatic increase in firm profits. And a lot of that has to do with the decline in worker bargaining power. And in order to restore that worker bargaining power it really requires government intervention. So as I said, I'll say more about that in a moment. The fourth scenario is the one that is the most worrisome and that is that firms could see higher profits as their costs fall. And if they don't share those productivity gains with workers, what that means is that employment will fall, inequality will rise. And when employment falls and workers have lower incomes it means that there's less demand to buy goods and services. So demand in the economy will decline. And there will in fact be higher unemployment. So this last scenario is the one that I think people worry about understandably and rightly so. The other scenarios however are possible. And so let me just give you a little bit of data here. This it shows you what's happened with productivity growth in the United States since 1947 through 2017. And the blue line here is worker compensation. And as you can see beginning in around 1975 we see a divergence of the two. It used to be that when productivity rose employers would share the gains of that productivity with workers by raising their wages or their overall compensation package. But that bifurcation began in the mid 1970s roughly beginning with the period of globalization. And I think that a lot of what this trend is about is about globalization. That is the impact of globalization on worker bargaining power. So several things that happened since 1975. One you've probably heard about a lot recently is that the minimum wage stagnated in real terms that is in inflation adjusted terms. So today the minimum wage is lower in real inflation adjusted terms than it was in 1968. That explains part of the depression work the decline in workers wages, de-unionization because of globalization have all contributed to a loss of order bargaining power. So they haven't been able to bargain for a fair share of the productivity gains that have happened. And so this is what creates a macroeconomic problem that we are in today. And that is that the redistribution of income to the highest income groups the wealthy to corporations is problematic in the sense that those groups tend to spend a smaller share of their income. They tend to save a higher share. And as a result of that it means there's less aggregate demand in the economy. So, and you know, I teach, you know, when I teach macro I often to say to my students about this issue that if there are no buyers, there are no sellers. Corporations are not gonna expand employment they're not gonna produce more goods and services if there's not enough to demand to buy back those goods and services. That is the real danger of technological change is that, that if firms retain all of those profits and don't share them then there's not gonna be enough demand generated to buy other goods and services that may be generated in the process of technological innovation. So that I think is really the macroeconomic context of this is that the technological change that results in higher productivity that raises profits but does not raise worker wages is likely to lead to a decline in employment and higher unemployment. But there are ways to raise the productivity I'm sorry, the bargaining power of workers and I think government is fundamentally important in this because the impact of globalization has been to increase the bargaining power of corporations who can now move anywhere they would like in the world to produce or source from anywhere they would like to in the world where wages are low or environmental regulations are low their bargaining power has increased but worker bargaining power has decreased. And so there are a variety of things that government can do one is to invest in robust worker training programs for displaced workers who have lost their employment due to technological change. This is something we could raise the minimum wage to raise to boost worker bargaining power and I think we're all Wisconsin that discussion right now as it's been discussed with the Biden programs that had been discussed in Congress we could also improve and equalize educational outcomes across groups especially by class and race. And I think that is to ensure access to what I would call knowledge intensive jobs that are not so likely to be threatened by AI and artificial intelligence. And finally to use tax policy to ensure a fair distribution of the benefits of technological change. You know, one proposal has been to tax robots for example. Another is higher marginal tax rates on corporations which of course is now under discussion with the Biden infrastructure plan and finally a universal basic income. I would imagine that there were other things that government can do to boost worker bargaining power and to serve as an intermediary to incentivize businesses to share their gains from technological change with workers which would avoid the macroeconomic problem of higher unemployment. So I will stop there and I'll stop sharing my screen and I'm happy to answer any questions. Great, thank you professor. And just to mention to members as well as folks who are watching online you can pull that presentation up. It's listed on our website as well as some prepared testimony that the professor has also submitted that you can pull up as well. So representative Yan Tashka and then representative Chase. Yeah, so you've hit on a lot of good points there that I think argue for raising the minimum wage, raising the marginal tax rate and a lot of other things that I agree with. I'm gonna date myself but I remember in 1965 I went to the World's Fair in New York City and I went through an exhibit, very futuristic exhibit by General Motors which touted the technological developments that would give us all more leisure time and our standard of living would go up and all that. As the future really developed though it seems like all the problems that you presented there as well as even exempt employees did not wind up with more leisure time. Regular workers have lower number of hours and therefore decreased income as a result but exempt workers also are working more than 40 hours a week in a lot of cases. Exempt workers that are on salary are often expected to work 45 to 60 hours a week, 45 to 55 hours a week on a regular basis otherwise they fall behind their colleagues. So there's a lot of change that has happened in the labor market that is even beyond what you presented and I just wanted to bring that out. So we do need changes in I think government regulation and putting more emphasis on our working society, thank you. Representative Chase. Thank you, professor. I think I remember you from class back at UVM. You were looking for me. Good to see you again. Good to see you. I was wondering if you had any data, like there's always been the kind of concept of the self-made business or whatever where somebody claims to develop an entirely new groundbreaking technology that only they themselves have done and there hasn't been any contribution from others. I was wondering if you had any data on that sort of thing like of these new technologies that are changing the way the labor market works that are so drastically increasing productivity, the time and money cost of developing these technologies is often related to universities and so forth which is a huge social sharing of cost to develop but then the profits and the use of these is often captured by the very small group that you mentioned earlier. Do you have any information on the ratio of how those compare the proverbial socialized costs versus privatized profits? Yeah, so let me say this. I just typed in a book that I encourage you all to read by Marina Matsukato who is a UK economist and it's called The Entrepreneurial State. So I don't know that I can answer Seth and the specificity that you like but one thing that we do understand and Marina's work really exemplifies this that a lot of the technological changes that have happened are actually the result of government investment. And so she, Matsukato has a really interesting graphic in which she looks at the iPhone, for example and looks at the GPS system which was developed by government entity, I believe in the United States. The facial recognition, so many of these technologies, the government has taken the risk of inventing and once the risk is gone, yes, exactly, once the risk is eliminated, firms will jump in and so we've sort of socialized the risk but privatized the profits of many of these technologies. So I think there's a legitimate argument to engage business in a discussion about that those benefits of higher productivity need to be equitably shared because society has actually paid for them. It is not really the firm that has adopted or invented the technology that has actually taken on the risk. And so I think that's can be broadly said about many of the new technologies in AI and machine learning and robots that have been are emerging right now. Thank you. Professor, I've got a few questions. The first one I have, I wanna make sure is not just a semantic question and that is kind of the reference to profits and those more accruing to the organization or the corporation as opposed to kind of the labor force, if you will. My sense is that the AI issue that we have to struggle with in society and how it affects labor markets goes way beyond kind of private corporate interests. I mean, this is gonna affect workforces in the nonprofit sector, I would think, nonprofit sector, government workers and there are the risks and potential rewards there as well. And I guess the semantic question is that, the quote unquote profits that you're referring to could be characterized as increased efficiency in a nonprofit organization that might ultimately result in less work or less workers. But I wanna be clear that these risks and opportunities are really spread across any employer, I would think. Well, we think of the nonprofit sector differently than we do the private sector, right? Nonprofits are there to achieve certain goals, not necessarily profits. So I don't know, I think about them they're in a different category, but yes, ultimately when these technologies spread they'll spread to those sectors as well, they'll spread to government as well, right? Government organizations may adopt some of these technologies as well. But I think the bigger issue is what happens in the private sector, I think that's driving what's going on. Yeah, I mean, kind of related to Seth's question, I was reading a report in the last week, I can't remember where it was, but about the accelerated development of vaccines in the last year and how a lot of that work really kind of came out of the academic world. But ultimately kind of seeped into the private sector where I think fortunately for society that we have ended up with a handful of vaccines which is great, but there was a lot of AI involved in how the acceleration of that work occurred. It started in the academic world and it seeped into the private sector again with good results, but it definitely is related to some of the questions that you've raised. I'm curious, what's the debate that's going on in the world of labor economists these days? I'm sure that there is not full agreement and what the possible outcomes of the expansion of AI in our for-profit world is, but can you give us a sense as to kind of how broad, how wide the dispersion is amongst labor economists as to potential outcomes and benefits? Where is the robust debate in this world right now in the labor economist community? Well, I think it is between the kind of two polls that I identified. Some who argue that AI and robots are gonna lead to a sharp decline in employment, but I think emerging from this is more recently in understanding that it's a bit more complex than that, that job structures, if you will, job descriptions will change and tasks will change. So I think that's the poll. We're trying to predict the future and many economists, including David Autor, who I cited, look at the past, that we have had technological change for a couple hundred years, so this is not a new phenomenon. But I think that the jury is still out in terms of understanding it, but we are trying. So I think here's the issue, that I think some technological change is good and we don't want people to have to do scrutinized, boring, mind-numbing jobs. We would like to free up that time to do things that are more interesting, that are more fulfilling and so forth. So it's not like we wanna, I think there's an agreement amongst economists, we don't wanna stop technological change. But we also understand that if all of the benefits accrue to corporations, then there actually will be a negative impact on workers. And the biggest problem I see in this debate is that people will often say, well, productivity rises and consumers will benefit from lower prices, but consumers are also workers. And so if we don't look at the impact on the labor force of this, we're really missing an assessment of what the complete impact is of technological change. So I would just say, Representative Rigglin, that I think that the jury is still out, but I would say that it is, I'm not sure a lot of people are buying the Andrew Yang argument. And I don't get me wrong, I'm not really critical of him. I think he's raised an interesting debate, but I don't think many people see it as black and white as he does. And the question is, what are the policies that we can adopt to capture the benefits and avoid the negative impacts? Yeah. And I suppose when you're running for president, it's much easier to present things in a black and white way. So he's probably a little more nuanced than that is a scratch below the surface. I've got a few more questions, more related to government policy. And I'm gonna ask one of them and then defer because there's a couple of other hands up, but this kind of relates to your last slide about things that government can do. One of the things that I'm really worried about is kind of looking ahead on this issue. Government is often quite reactive to these issues. It's slow, it's filled with compromise. And in many ways those things are good in terms of the deliberation, but the pace of change with these issues right now, I think is an issue of concern because as the pace of technological change, and we're talking about AI here, accelerates. Yet we are dealing with government, which frankly hasn't really changed a whole lot in the last century, probably, in terms of its ability to deal with change in the pace of that. There's kind of a mismatch there in terms of the tool we have. How does government react, if at all, relative to how quickly things are moving in the private sector and the pace of technological change? So maybe this is a comment as much of a question, but I'm curious if you'd have a thought on this. How can we react as policy makers to this extreme acceleration? Right, so I have a couple of thoughts. Most of my research in the mid-90s and early 2000s was on East Asian economies, in particular South Korea. There's no reason that government can't be a robust player in this. And South Korean government, its role in the industrialization process in South Korea really demonstrates the potential for government to be a leader, to be a strategic leader in economic development. And that requires that it's not impossible. I hear what you're saying about the wheels turn slowly sometimes, but I think that if we commit to this area and monitoring it, bringing in the expertise that's needed, we can anticipate these changes and we won't be being reactive, right? So as I said, I highly admire what I have seen in East Asian economies. They have democracies as well that they struggle with, but they have been very clear about the role of government to guide the economy in the process of industrialization. And again, I encourage you to read the book. It's actually Mariana Matsukato, my apologies, The Entrepreneurial State. And I think that we can do that. That's why I think I'm such a fan of this legislation is I think that if you have a group that is monitoring this, that is bringing in expertise, you can anticipate these changes rather than being reactive to them. So that's my optimistic view of the world. Great, thank you. Representative Sims and then Representative Yantatchka. Yeah, thank you so much for this testimony, Professor Sabino, sobering about the current state of our economy and what the future looks like without more sort of active engagement. And you've talked a big picture about some of the things government can do, including raising the minimum wage and thinking, rounding us in sort of the jurisdiction of our committee here, energy and technology. I'm curious, I don't know if you've had a chance to look at H2-63 or H4-10, the bills on the table for us right now around AI and automation or more generally, what do you see as key levers that here in our committee, we can be focusing on to ensure the future prosperity of all firm owners? Yeah, thank you, that was a hard question. I don't know that I can answer it to your satisfaction. I've looked at the bill. I mean, I think that, for me, there are two issues. One are the employment effects of AI. The other is the abuse of those technologies. And in this, I'm thinking about the racially discriminatory effects of AI for example and protecting Vermont citizens against technologies that are discriminatory in their functioning with facial recognition, for example, using it for in policing, all of those things. So I think those are the two broad categories. There's probably a consumer category that you might wanna think about and I don't have any specifics on that right now, but I think that might be useful to look into. And so I would just put it into those three buckets and I have to say that I think this is new enough that I think, to be honest with you, I think that it would be really great if you would convene or if somebody would convene a group of experts on this, economists like myself, but others as well to really look into what the specific issues are in Vermont with regard to these technologies. So I don't think I know, I don't know that right off hand, I don't know that we have an assessment of to what extent AI has infiltrated both our private sector as well as government operations, such as I said, using AI with policing. So I think that assessment needs to happen first and then I think you can plot your way through what the policies are. And I would just say my guiding principle in doing this work, looking at these issues is that we have shifted from a world in which the saying that Representative Yanchaqa mentioned that what's good for corporations is good for America. There used to be a saying, what's good for GM is good for America. And so whatever benefited corporations, we just, we adopted and embraced and the world has changed. And so our challenge both as academics as well as government is to develop the incentives for corporations to align their profit motives with societal wellbeing. Well, you then have to decide what are those indicators of societal wellbeing that you're concerned with? Is it wages? Is it work hours? Is it work stability? There are a variety of other things. And so I think that's the work that you need to do to begin with is to say, what are those indicators of wellbeing that you want to protect? Assess where we are with AI and robots in Vermont and then figure out what are the mechanisms that you can use to incentivize businesses to align their profit motives with societal wellbeing? I think that's the general challenge and you have to get into the nitty gritty of digging deeper in that frame to figure out what the specifics are. And as Representative Briglin said, this is a dynamic field. And so things are changing, but I think we certainly have enough time if we move on this to catch up and develop the mechanisms to stay one step ahead of what's happening. Thank you so much. Representative Rogers and then Representative Yantachka. Thanks. I am going to ask you the same question that I've asked most of our witnesses and to see if you have a thought on it. So I guess I'm interested in kind of this question of responsibility and where responsibility lies. So you spoke about two specific components, the labor piece and the bias piece like racial or gender bias or other forms of biases. I'm just wondering if you have a perspective on where you believe that that responsibility should lie if it's with developers of technology, users of technology, buyers, funders. If it can't go with any of those and it has to be with the government as a regulator. So just any comments on that and then also within state government itself we're kind of debating or considering whether it makes the most sense to house some sort of an ongoing commission in the agency of digital services in commerce and community development or somewhere else within state government. So if you have any comments on either of those I would be interested to hear them. Thank you. You ask a really interesting question with regard to the issue of responsibility. I think there are corporate responsibility is high but we live in a capitalist system and so firms are driven by the profit motive, right? We have ethical firms. We have firms like Ben and Jerry's that have broader goals than just profits but that's not necessarily widespread. And so while I think that corporations have that responsibility, if I were the owner of a business I certainly would it's not the nature of our society and economy. You certainly government can incentivize firms to be responsible but I think ultimately it has to be government that is the responsible entity that is overseeing the well-being of society as a whole. Many of the governments that I've worked with and places that I've studied there's often a tripartite relationship between representatives of business, representatives of labor and representatives of government and government has typically been the convener who has brought these groups together to work out an arrangement that is beneficial for all groups. And so I think fundamentally that responsibility lies with government. Last part of your question about where to house this I have to think about that. I don't think digital services is the place for it. It is probably, I would say probably it is with commerce because what you're trying to do is to shape the decision-making of businesses. But certainly the issues around race I think are with AI are very severe and that needs to be housed that there needs to be some relationship with whether it's Susanna Davis and her department or other places that would be monitoring that. So there needs to be some connection also probably with labor as well because the impact is on workers, a lot of it, not all of it, but and they need to be, you need to be aware of what the effects are so that you can design policies to address those problems. Thank you. Representative Yann Tachka. Yes, thanks again for the opportunity. I have a question about AI and promotion of AI development but I wanna start with another observation I spent when I was working for IBM I was assigned to a plant and single thing in Germany for a year and I took my family over and one thing I was really surprised is that the German government gives every family something called Kindergeld and that is a payment for each child in the family that's given to just basically sent to the family as a deposit into their account every month. And that's very similar to, it's not quite the same thing as a guaranteed income but it certainly is something that I think is in President Biden's proposal. Anyway, that's an aside. My question on AI is we, we probably could train a lot of people in the skills that are required for AI type jobs but there's no way to actually keep them in Vermont unless we can promote the types of businesses or the types of production that would employ those types of jobs. So the question is, how do we, do we need to give incentives for companies that develop AI to locate in Vermont or is there another way to encourage that type of business in Vermont? So really good questions. Let me just say briefly with your comment about Germany and Kindergelt. This was in the Biden package, it's child allowances or actually child tax credits and I don't know how familiar you are with this but it's one of the best things I've seen happen in all of my years as an economist. The, if this pass, I guess it did pass that it will reduce child poverty in Vermont by 33% and in the country as a whole by 50%. You cannot have workers who are in the knowledge intensive world who have not been well educated, have not been living in families in which there is sufficient finances for the family to be able to support the child. So that in and of itself promotes worker wellbeing, if you will, and their ability to take on some of these new jobs. So it's much more than just AI, there are many components to making the economy work for everybody and that is certainly one of them. With regard to, you know, I'm not a great fan of tax credits to incentivize a corporation to move to Vermont. We find that actually those tax incentives are minimal in terms of their impact. They might have a temporary effect but because of how mobile firms are today, there's no saying that they will stay and I just will remind you of years ago in Detroit with General Motors that they had asked for tax reductions in order to stay in Detroit. They got millions of tax cuts and then they left five years later. So that's not a recipe. I think what matters for business is a qualified workforce, good public schools, a healthy environment and a variety of other things about sort of the ease of doing business but tax incentives themselves are not sufficient. So in many ways, Vermont has a lot of good things that do attract businesses and we're seeing this with COVID and the relocation of a lot of people to Vermont to work remotely, for example. So this is one of those areas where I can talk generally because I have studied a lot of different countries but I think we would need to sit down for Vermont. We do not have an economic development strategy for Vermont. We don't have a rural development economic strategy for Vermont. We need to figure that out and we can then think about what kind of strategies to use to promote certain kinds of industries. We need to figure out what are strategic industries to promote in Vermont. This is how the countries that have grown the most rapidly in the last 50 years are countries in which government has taken a lead in doing that. They have said these are the strategic industries we wanna support. They have developed mechanisms for doing that. They have developed mechanisms for holding those companies accountable for reaching certain goals if they are given subsidies. So it requires a bigger economic development strategy and sadly in all of my time in Vermont I have not seen that thought go into it. So I think that's the work of your committee should be embedded in that broader goal of where do we go in this global context with these technological changes with the mobility of capital and so on and so forth. Yeah, and the last term of the Shumlin administration IBM was considering some, well, they did self in global foundries and there were incentives given to I believe global foundries with the guarantee that they would stay in Vermont for at least eight years. Thankfully they're still here but there's no guaranteed that they'll continue staying here. I guess you said that we have to be able to promote the types of businesses that would do the kind of development we want. And I guess I still have the question of how do we promote that? What do we do to promote that? To promote certain kinds of industries or jobs? To promote the development of companies that'll locate and stay in Vermont. That's a good question. I just want to go back to the research on this and the research tells us that tax incentives have very little to do with where companies locate. And so it is specific to the industry or the business. It's specific to what their needs are and what would incentivize them to come here and stay here. So it's really hard to just generalize. I mean, if we want to promote, for example, organic farming in Vermont, in the farm to table food industry, that's a very different set of incentives and efforts of government in terms of regulation and so forth than if you're trying to promote the development of financial firms, for example. So I wish I could be more specific, but it really depends upon the industry. And I think what it speaks to is that it requires, I mean, in response to representative Big Briglin's comment, that's why government is reactive here and why it hasn't been reactive in other places because we have not planned ahead and we don't know where we're going. We have simply left it to the private sector, which may or may not generate benefits for the state as a whole, right? So just to say representative Rinchaka, I mean, one of the things that's coming, that's happening is I think that hopefully the Biden infrastructure bill will pass. That's gonna create a lot of green energy jobs. And the question is then, if we want those firms to produce here in Vermont, what kind of training do workers need that they may not have for those industries? So might there be a partnership between businesses that locate here and the state providing some kind of worker training for those types of jobs? And I wanted to say that when we talk about those things, it is fundamentally important to think about gender because many of these jobs, including AI, have excluded women and discriminated against women. So you also have to be sensitive to creating incentives for firms to treat workers fairly. So for example, one example is that firms that do get subsidies or some kind of incentive that is conditioned upon ensuring that women have a proportionate share of the jobs that are created. So again, you have to think about what are the goals we want to achieve for the state in terms of equity and then tailoring the policies to attract certain industries to meet those goals and they will be specific to the various industries. Hey, thank you. I appreciate your... So Seth, I'm gonna butt in line because my question, I think is related to Mike's, which is the challenges, some of the things that you've mentioned, Professor, in terms of the big issues that have driven the separation of improvement in productivity relative to improvement in wages in the last, call it 40 years, were related to things like globalization, starting as early as 50 years ago, maybe even earlier than that. And then more recently, technological change, which has accelerated. Both of those types of things would seem really challenging to get our arms around or to address at a state level. And I'm just thinking about some of the constraints of policymakers in a state as small as Vermont, relative to maybe that needle can be moved in a California or a Texas or maybe even a New York. But the challenges of what we do in Vermont from an economic policy standpoint to address things like globalization and technological change and the acceleration of that and how that affects the labor force. One of the things that I'm hearing you say is fundamentally, if you prepare Vermonters from a workforce perspective, you're getting ahead of the game there simply by having a better trained, more ready to work in the current economic environment, you're ahead of the game there. But I don't know if that's necessarily, maybe that's economic policy, but I'm struggling with how is a small state to we deal with these truly global issues and AI isn't gonna be considered an industry. It's any more than the internet is considered an industry. It's ubiquitous, but I'm concerned with our ability to really affect some of these issues. Not that we shouldn't try, but as a very small state with a very small economy. Well, I mean, I think you're absolutely right. I always struggle with there are things that can be done at the national level that are much harder to do at the state level, especially for a small state. So you're not going to change the trajectory of AI in Vermont only. But what you can do, as I said, is there are certain ways you can incentivize businesses that use these technologies in a way that achieves the goals that we want them to achieve. Whether it's race bias in the use of technologies, whether it is gender bias in hiring or even in research, for example. I mean, Silicon Valley is notorious for its treatment of women. And so to the extent that we have those kinds of firms here that are in any way getting contracts with state government, that's an opportunity to do that. So I think what we deal with in Vermont is not so much changing the trajectory of the technologies, but the impact of the use of those technologies by firms here. So there are things that can be done to incentivize firms to move in a particular direction. And as I said, it's industry-specific, it's occupation-specific. But it depends on how you wanna focus on this. So I focused a lot on the gap between the increases in productivity from these types of technological change to wages. And there are ways to bolster the bargaining power of workers, but there may be other things that can be done. I think, I wish I had more specific suggestions for you, but I feel like we're at early stages of actually trying to grapple with this as a governmental level. But I also wanna, well, I understand your concern. There's so much that government can do, even a small state like Vermont, that can do, that can change the discussion. So frequently, as an economist and in the private sector, there's very little discussion of ethics around the way that we do things. And government can be a voice for making ethical choices around production. And that is an important bully pulpit, if you will, to create an environment in which we are all thinking about these things, including business owners and so forth. So I think there's a lot that government can do by simply illuminating these issues and advocating for our goals, which our goals are not to be the most modern, technologically advanced country in the world, our goal is to ensure the wellbeing of our citizens, that people can provide for themselves with dignity and economic security. And technology needs to be subservient to that. We may not have a lot of power in a state of Vermont, but we can articulate that goal and we can advocate for that goal and we can incentivize people to study that further and to come up with solutions that will move us in that direction. Thank you. Go ahead, Seth. Thank you. And thank you, Professor, for that last bit. I want to highlight that. That was very well said, and I appreciate that. I wanted to ask one of the patterns that I think I've observed and I'd like to double check is that a lot of these new industries are not kind of popping up on their own. They're popping up around education and research centers. So like Silicon Valley, we've got Berkeley and Stanford. Boston is huge. They've got like MIT and Harvard, that sort of thing. So I guess the first question is, does that pattern hold globally or is that more of a U.S. phenomenon? And as a follow-up, what can we do with Vermont universities and schools to help build these new cutting edge, bleeding edge sort of technology, research and development sort of things that will create these new industries that will be ubiquitous in 50 years? I would say when I look at some countries, yes, the government is involved in funding research in particular areas. And I'm just thinking of a paper that I just wrote on this with regard to the United Kingdom. They often make their grants or even their research grants contingent upon things like having equitable gender representation amongst researchers or racial representation, for example. I think that you're in some sense, you're competing with the money that corporations have to incentivize research with government. And I think we could do a better job in Vermont of utilizing researchers in the state or inviting them to participate in research on these areas. But we have to have a clear direction or at least a sense of a direction on where we're going to ask them to do this work. I mean, as a scholar at the University of Vermont, I do a lot of work for the public sector and racial profiling and a number of other areas that I do research on, but I'm well aware of the fact that I'm one of the few faculty that do that. And I think we have enormous intellectual resources in Vermont to move us in that direction. And it doesn't always need to be funded research. I mean, I think that it helps a lot if you can fund that to some extent, but I think simply building that connection with the university. I'm a fellow at the Gund Institute for the Environment. And one of the things that we talked about last year, for example, was working with the legislature, was working on climate change to have us as partners of the table. I mean, we would do this just to share our research and to learn from what government needs for research to incentivize, we give grants out, for example, at the Gund Institute, capitalist grants. And we could shape some of that to fit some of the needs of the state of Vermont. But if there's not that collaboration between the legislature and the universities, then that's not gonna get done, but that is a vehicle. And yes, money would be nice, but even simply convening those resources would be tremendous. And I think there would be a lot of willingness to research on topics related to Vermont in this regard. Thank you. It's, I want to say it's a huge untapped resource for the state that I think the state could really benefit from and the legislature could benefit from in terms of some of these issues. Yes, we have dozens of staff at our disposal. Actually, other than some attorneys and Matthew, we are kind of a self-supportive citizen legislature, which has its merits too. I don't mean to be on that. Professor, thank you very much for joining us. I don't see any hands up right now, which is the queue for if other folks have questions to please get in the queue. But I really want to thank you for your testimony. This is not an area that we have dug into. And to some extent, it's a little bit beyond the purview of our committee. At the same time, I think it demonstrates how the issues that we're gonna be wrestling with in these two bills go far beyond this committee. They're in the Commerce Committee, they're in the Judiciary Committee, which deals more with the civil liberties type stuff and other areas as well. And actually, I think that is really, you've shown a light on that, I think with this discussion this morning. So really appreciate you being here with us. So thank you. Members, we're gonna be on the floor, I think at 1.15 this afternoon. As I previewed yesterday, I'm hopeful that we're gonna have some more discussion. I don't think it'll be lengthy, but there's a few things that we wanna get through on S1 tomorrow. I wanna do that discussion after we're done with the floor. I don't have a sense yet as to whether that's gonna be pre-lunch or after lunch, so we'll be flexible with how that works. But that's our next committee meeting just to give folks a heads up. So, and the other thing, actually I wanted to check in with Lucy as we adjourn. I wanted to ask you about some of the NCSL stuff and if I can be helpful with a path forward on that. So we are adjourned. Thank you all and see you this afternoon.