Jeffrey Sachs: Afraid A.I. Will Wipeout Tons Of Jobs? It Already Is.
Summary
Market Outlook: The economy is highly uneven with strength in tech-driven areas and weakness in wage growth, with East/South Asia dynamic and Europe largely stagnant.
AI Sector: AI and data center buildouts are booming, attracting significant capital and driving demand for compute and power grid expansion, but raising risks of large-scale job displacement.
AI Arms Race: The U.S.-China AI competition was likened to the nuclear arms race, with no true “winner” and a need for faster international governance to avoid destabilizing outcomes.
Platforms & Media: Big Tech platforms such as Meta (META), Alphabet/Google (GOOGL), and Microsoft (MSFT) were cited as controlling media, user data, and political influence, heightening privacy and concentration risks.
Chips & Infrastructure: References to Nvidia (NVDA) and “2nm chips” highlighted attempts to control advanced compute, while new architectures may circumvent chokepoints, underscoring rapid, global innovation.
Defense Tech: The Pentagon’s push to embed AI in weapon systems signals growing integration of AI in national security, with private firms increasingly central to defense capabilities.
Inequality & Politics: Wealth concentration among a few tech billionaires, stagnant real wages, and political dysfunction increase societal risk despite technological progress.
Investor Considerations: While AI infrastructure and defense tech are gaining momentum, investors should weigh ethical, regulatory, and geopolitical risks and consider diversification and risk management.
Transcript
If uh the dynamics of the digital economy continue to run pretty much on a market driven basis, then we're going to have large parts of society uh not only left behind but losing in absolute terms. And that's simply because the fact of uh job displacement by smart machines, by smart systems, by robots or by AI systems is not only a prospect, but it's happening already. Uh and it's happening at a pretty significant rate. [Music] Welcome to Thoughtful Money. I'm its founder and your host, Adam Tagert. Will artificial intelligence wipe out millions of the jobs done by humans today? The AI CEOs tell us it's nothing to worry about. But one of the most influential living economists, as recognized by the Economist magazine, begs to differ. He says that AI is already starting to replace human labor at scale, and we don't have a plan for the aftermath. To discuss this very important topic in depth, we've got the great privilege of speaking today with Jeffrey Saxs, professor of economics at Colombia, best-selling author, and global leader in sustainable development. Professor Saxs, thanks so much for joining us today. >> It's a pleasure to be with you. Thank you. >> Well, thank you. And look, uh, you have an incredibly busy schedule. We don't have a ton of time today. I've got a lot of questions for you. We're going to get through as many as we can and, uh, and that'll be it. Um it's been a little while since you've been on the program and so just to kind of level set I want to ask you a general question um what is your current assessment of the economy both domestically and globally right now? We can get into the details but at a high level. >> Sure. Uh basically uh things are very uneven uh within the US economy and globally. By that I mean we clearly have areas of remarkable dynamism in the US around the data centers and AI. This is a boom. Uh it's a lot of capital is being raised and a lot of capital is being invested. uh this is generating uh growth in uh uh certain parts of the economy, certain regions, certain sectors including the power grid which is needed for the big expansion of all of this compute. Globally, we also have uh many parts of the world that are very dynamic, especially in Asia. But in the United States and in the world there are also significant sectors and parts of society that are suffering. Clearly just basic wage income is struggling. The so-called affordability crisis is really a crisis of stagnant or falling real wages. uh in the world as a whole we have fast growth in East Asia uh and in South Asia. Uh we have uh economic crisis and stagnation in other regions a lot of Latin America, a lot of Africa. So the picture and Europe is basically a a rich and stagnant or even contracting economy. So the picture is unsettled. It's varied. It's a a highly disruptive scene because of so many deep changes that are simultaneously hitting our societies. And there is therefore no uh one overarching reality. It depends on who you're talking about, which part of the world you're talking about, which sectors one is talking about. What we have is uncertainty and uh great variation. >> Okay. Um I was going to do this big top down international, you know, down more into the micro US thing, but I'm going to pull up one of my questions from the end of the my list here. Um so you know mixed uneven as you said a lot of people are saying you know K-shaped economy at least you know here in the US um what is your outlook for that that chasm of of unevenness here um especially because so much of the the hopes of the future of the US economy are being pinned on the hypers scale ers uh on AI. Um I think the last time you were on we very briefly touched on both the the promise of AI but also some of the risks and and a big risk being massive jobs displacement and creation of a a large societal under/employment crisis. Um, are you feeling more hopeful at this point in time that that the prosperity that we're gunning for will kind of lift all boats here in the US or are you more concerned that it actually might leave that bottom half of the K, which which really feels more like 80, you know, 20% is the up half of the K, 80%? Is that 80% going to get left further behind here? >> Well, I I think we need to distinguish two things. If you ask will market forces solve this problem the answer is almost certainly no. If the dynamics of the digital economy continue to run pretty much on a market driven basis, then we're going to have large parts of society uh not only left behind but losing in absolute terms. And that's simply because the fact of uh job displacement by smart machines, by smart systems, by robots or by AI systems is not only a prospect, but it's happening already. Uh and it's happening at a pretty significant rate. And what we're hearing uh is uh that many sectors of the economy including uh a lot of skilled uh worker sectors college educated uh work uh is now already finding that there are job freezes and and so on. If you ask a different question uh which is uh taking into account politics and redistribution of income and reorganization of society, could this technology lift all votes in uh at at the end of the day I think the logic says yes uh this could be a kind of breakthrough that is very very broadly beneficial. Then you have to ask a third question. Are our politics on track right now to fulfill that role of underpinning a broadbased uh benefit of AI. And the answer certainly is no. Our politics are completely dysfunctional and becoming more and more dysfunctional. There's not even a rhetorical uh flourish in US politics right now uh aimed in the right direction. No one's asking the question uh in Washington. I mean uh what do we do to ensure broad-based development? Uh we're in a uh in in a scramble for wealth. We're in a completely corrupt political environment of self-deing all over the place including the president's family and everybody else involved. Um so our politics are not aimed at the broader benefits and the market dynamics which will therefore prevail in the next few years are going to be uh uh widening the income and wealth inequalities dramatically. Look, we have 10 people in the United States whose net worth right now is $2.3 trillion. This is incredible. Really? It's that's a lot of money. An average of $230 billion dollar for the 10 richest Americans. Those are all big tech uh uh super wealth holders. Um they not only control that wealth and the companies, but they control their platforms, the media. uh they have all our data and they control the White House and the Congress to a large extent and uh they're absolutely not interested in the question of how to make sure that everybody benefits from all of this. They're scrambling in competition with each other. Uh they uh want to have trillion dollar paychecks. It's a very bizarre world where uh We don't know whether that upper part of the K is 10 people, is 1%, is 5%. It sure is not a majority of of the public. >> Okay. I I've been actually using the term a lowercase eye shape economy that I think we're in the danger of moving to where it's just this little dot at the top. Very few people doing extremely well far removed from the rest of everybody who's just getting left increasingly behind. Um okay so you are I think safe to say concerned about the current trajectory here and I want to continue tugging at the string real quick. Um I I'd just like to get your opinion on um yes there's a big market opportunity that the hyperscalers are rushing after but there also is a sovereign national interest that's been declared in AI you know sort of akin to the nuclear arms race back in the the the 30s and 40s where America is is I believe saying look we've got to win this. this is going to be the technology that that lets America remain the superpower of the future. If we lose this race, say to China, that could dislodge us. Um, A, is that a fair characterization? And B, do do you agree with that? In other words, you we'll talk for a moment in a moment about kind of the repercussions of all this, but do we need to win the AI race as as America? Well, I think the analogy to nuclear arms is correct and uh but that shows the fallacy of the claim that is made that we quote need to win this race. Who's the winner of the nuclear arms race? Nobody. We live under the nuclear sword of Damocles. We're at the edge of extinction. One wrong move, one mistake, everything ends. And believe me, if people think that's absurd hyperbole, they don't know what they're talking about. We live so close to the edge right now. We depend on not very smart people to keep the world alive. We depend on there not being a terrible accident. Now, it's interesting to go back to the uh atomic bomb and the beliefs in it. when the atomic bomb was uh uh first exploded at Los Alamos, New Mexico, and then uh and then used twice uh against Japan for no necessary reason, by the way, because the Japanese were trying to surrender at the time. But uh when it was used, American strategists believed that the US would have a monopoly on atomic bombs for at least a generation. They talked about 30 years. >> It turned out to be four years. Uh every calculation was a wrong calculation. When pushed by Teller, we moved to thermonuclear weapons. We thought that that would be a lasting advantage. The Soviet Union matched us almost instantaneously in the hydrogen bomb. Then came the uh rapid proliferation of these weapons. Then came a rather desperate uh attempt led by President John F. Kennedy uh initially uh to try to get this under control starting with the 1963 partial nuclear testban treaty and then uh the nuclear non-prololiferation treaty then uh a series of arms control agreements between the United States and the Soviet Union. Uh and in other words, the idea of American officials was we'll win this race and then they realized no there are no winners. Uh we are going to get ourselves blown up if we don't get this under some kind of uh mutual control. uh by the time uh there has been some control there are nine uh nuclear countries right now and nothing's really under control. I would say uh our situation is extraordinarily fragile. Uh the bulletin of atomic scientists puts out the uh doomsday clock which uh is the uh visual uh representation of how close we are to disaster. And the doomsday clock is set at 89 seconds to midnight, meaning the closest we've ever been in the whole atomic age or nuclear age to nuclear Armageddon. So now to come to AI, what you expressed is absolutely the view in Washington. Uh and I happened to listen a few years ago to a chillingly stupid discussion of two NATO generals at one of these, you know, strategic conferences that I attended once in a while. Um and they were saying to each other, "Yes, we must win the AI race with China. We're gonna win, we're gonna do this. Uh this is so naive. Uh China produces so many more PhDs than we do. Uh just the scale difference is vast. The talent is enormous. Uh and you get advances in the US and then you know something we never heard of before called Deep Seek comes along and says we're here. uh and then suddenly you have all over China uh you know topf flight top performing LLMs and then the United States says oh we've got a chokeold on two nanometer chips on Nvidia chips and then you get almost weekly a report of some new computer architecture that does an end run around those chips. You know, the United States has some monopoly on this. Are you kidding? This is ridiculous. Of course, it doesn't. Are we going to win the AI race against China? No. Are we going to lose the AI race? No. This is not that kind of thing. This is not, you know, this is not a race. This is a core technology in which there's going to be a lot of expertise all over the world. The idea that we're going to have some chokeold over this is absurd. So let's get it out of our heads at the beginning. And then we should ask, so given the analogy with nuclear arms, where is this dangerous? What should we do about it? And what should we do diplomatically about it? We should move more quickly to that with the nuclear arms we did not get to even the first treaty, the partial nuclear testban treaty until 18 years into the atomic age. Can we do a little bit better than that this time? That that's what I hope. >> Okay. Uh well, I I'll add my hopes to that. Um the reason why I asked the question I think you answered it is um while you and I have that hope right now it doesn't seem that that's at the forefront of the mind of the folks in Washington right so >> not at all >> right so so we have these sort of twin engines propelling the AI buildout um both the market the size of the market opportunity so the capitalists are sprinting after that and then the the state sector is saying look we're going to do everything we can to clear the playing field for you because we want to win, right? >> And and I would add one more very obvious and specific point, which is that it's not that we want to win the economic race. It is that the Pentagon is salivating. Uh we need AI and all our weapon systems. So this is becoming completely a a national security uh tool as well or what they call national security. I would call it national insecurity because everything we we are doing is raising our insecurity. But this is an extraordinary technology because it is at once everything about the uh production processes of the economy. So AI is be becoming embedded in all production processes. It's embedded in the service economy. It's embedded in our online lives. automatically uh but it's also it's embedded in our media which happens to be owned by these very uh platforms whether it's Meta or Google or Microsoft. So they own the media uh they own the platforms. Uh it is uh mortgaging our privacy because they actually own our keystrokes also uh in ways that I can't really uh understand but I know it's there quite obviously. Um and it's also they own the White House. They appoint they appointed the vice president for God's sake. They told Donald Trump who his vice president was going to be uh at a dinner in uh San Francisco. Uh it's not even subtle this stuff. Uh so they own the politics and unlike the nuclear arms by the way where basically in a way the generals just pushed away the atomic scientists after a while or hired some of them. uh but said we own this. Uh the Pentagon doesn't own uh AI. Uh this is owned by Peter Teal and Elon Musk. The Pentagon is contracting out uh to these companies which hold the technology now. So this is really an extraordinary concentration of private power. Uh and uh you add in yes they're making way for they're making way for uh all of this. Um uh but they're bas basically uh turning over politics, national security, mass media, the payment system, by the way, which is going to be a digital crypto something or other payment system that maybe they'll let us know about at some point. Uh and uh all of our media. So >> pretty into the the hands of a private jet. It's almost sort of the the Tony Starkization of of the economy if you if you get the reference from the Marvel movies. >> Yes. >> Um Okay. So So again, I think you're you're proving the point I'm trying to make here, which is this is a juggernaut that is in motion and and perhaps even still building momentum. So, um, as much as you and I would like there to be more cautionary, you know, steps in here, probably not going to happen anytime soon, which gets us to, okay, so um, probably doesn't bode pretty well for that bottom 80%. Right? Like we we might get cheaper entertainment. Okay, hooray. You know, maybe we enter the Ready Player One era where we get free VR goggles and we can spend all our time in the virtual world, but we're out of jobs. Um, you know, we're we're we're out of purpose. we're we're we're out of power on a relative basis because it's concentrating in in in that that dot of the lowercase die-shaped economy that we talked about. So, um I guess a lot of concerns around that. Now, you had said um there's a way to do it differently and um I I guess first I just want to imprint on the listeners that a really smart gentleman like yourself who's very connected and studies all this stuff is quite worried about where society could go here. and and maybe you might want to I might ask you just to make a comment or two on on what you think that future looks like. Um where where the the the bottom 80% get even further dispossessed from here. And then secondly, I'd like you to talk about your vision of how we could do it differently. And and I guess where I'm going with this is what I hear from the AI fanboys is hey AI is going to create a lot of jobs as well maybe but I think on on a fractional basis it's going to be a lot fewer than it's displacing. Um so they say well even if it does we'll have universal basic income and everybody can go focus on being their best selves. I I think history is pretty clear. It's my opinion. You might feel differently that that that you know giving people a subsidy to basically do nothing creates a dependent class that that you know ends up investing more in their vices than their virtues. So uh in the in the better way we do this if technology is still doing all the work, how do we make this more equal? Is is it just redistributive policies or is there something in here that gives the bottom 80% a role? >> Well, uh first of all, I don't know the answers to this and I don't think anybody does. I haven't uh seen or heard uh anything that I find uh accurate and compelling uh that says that is how the future will play out. So I just want to uh caution on that obvious point at the beginning. Second, we've been in the midst of related changes for uh at least uh 60 years. Automation became a a thing a noticeable thing in the 1960s. uh it became a a matter of a lot of public discussion already 60 years ago and that was not wrong uh to worry about that because what we've had during a 60-year period is a bifurcation of society uh into three parts now. One is the the super uber halves, those who have more than is absolutely imaginable. uh and then the successful halves which is uh broadly speaking the professional class in America typically people who have had an advanced degree uh of some kind an MBA or law degree or medical degree or other professional degree uh and the uh broadly speaking somewhere between the high school completion and bachelor's completion depending on a lot of specifics but finding stagnant wages by and large uh and um very big difficulties of uh accumulating wealth. Uh and one of the most evident of those difficulties is homeowning for young people. uh which is deemed to be a you know pretty widespread crisis for a lot of young people. So we've already b triifificated you could say uh we already have three different uh groupings in society that live very very different lives. If you ask uh my cohort in the professional world how are things and how they been over the last 20 years. Yeah. good, very good, good incomes. Uh, stock market's been going up, lots of new opportunities, tech has been good. Uh, feels all very good. Actually, >> my house goes up in value every year. >> Yeah. If if you ask someone who basically doesn't have saving and uh if they've been very uh unlucky, but that's a lot of American uh people aged 25 to to 50 um had a high school degree uh went to work, lost their job and so forth. This is not easy. Uh health care costs are soaring. hard to save anything. What's called the affordability crisis is really again a stagnant wages crisis that comes from the fact that basic jobs have been under uh replacement dynamics for decades, not just now. It's not just AI. Uh it's robots. It's uh smart systems of many kinds that have been operating. So we're we're already in the midst of this. What is interesting is that and and it showed up in some you know everybody understands this now I hope in some very dramatic ways life expectancy gaps between having a university education and a high school education is eight or 10 years of life expectancy. Can you imagine this? Completely different lives within our society. And what is interesting, strange word, but if you're an analyst, our political system to this day hasn't reacted to this. Uh really, we have been in a kind of paralysis. Uh we had, for example, a huge fight over Obamacare. Obamacare is a half a 1% of GDP. It's a tiny little change of the national income. It was uh you know fullout war for a decade in the political system. In other words, our political system is paralyzed, unable to address big issues like the changing income distribution according to educational attainment. And now we we have a different changing income distribution. It's not really most likely uh between skilled and unskilled or high educated and lower educated. It it's actually more broadly between those who have uh financial wealth and those who don't because stock market has you know maybe there's some bubble in it but it's been booming in part because more and more national income goes to profits goes to the returns to patents goes to software goes to the monopoly or oligopoly ownership of a few digital platforms and if you happen to have enough income that you have been saving out of that income and you've been investing in your uh in in your retirement accounts and so forth, you've been a beneficiary of that wealth. But if your income has been so low that you're my god still trying to pay your student debt or uh trying to pay an unexpected uh healthc care emergency, you don't have wealth. You haven't benefited from the shift of national income from wages to profits. You've lost from that. And so we are now in an intensifying phase of this growing inequality. And I don't see the slightest evidence at all. No evidence right now that Washington's on the case. I don't hear one word about it. I don't hear even a conceptualization the Democrats. I mean, it's stupid even to say they don't represent they they don't represent the workers, but MAGA doesn't represent the workers either. The whole thing is shambalik. Both of them, Democrats, if they represent any class, they represent professionals. Uh, and MAGA represents the Trump family and his friends. Uh, and so we don't have a politics of facing up to this question. We have no serious discussion. We don't even get to your question of is there a better way than redistribution or predistribution to address this. I'm pretty sure by the way that part of the answer is actually what we would call predistribution which is that at least everybody gets education with quality and health care with quality not out of their personal consumption spending but out of their citizenship. I do think that this is absolutely right that we have as a matter of citizenship that you can get your health care without needing a MD or PhD to understand our health care system and without having to uh you know go into um uh irretrievable irreversible indebtedness to to pay your medical bills. But we're not even there at that point yet. >> Okay, Professor Zach, I have so many questions for you, but we're running out of time. So, we'll just have to reserve them for the next time to kind of land the plane here. I hear you say, look, problems big, maybe intractable. Our >> No, I don't say intractable. I just say big big big. >> Okay. But very very big. So, in the near term, you know, it might feel intractable. Um and the the our political system is not going to ride to the rescue anytime soon. >> Our political system as it's now constituted. Yes. >> As it's now constituted. Yes. >> So um uh I mean there's so many questions I know the audience would like me to ask you here, but just to make it practical for somebody watching this who is saying, "Whoa, this feels really heavy and kind of depressing." Um do you have any advice to them to at least say look here are some things you could could think about um considering so that you are less likely you're reducing your odds of being collateral damage to all of this if this juggernaut continues for the next 5 to 10 years. >> Yeah. Uh let me let me just say in in the long picture which is not a very satisfactory answer but uh it's also important to understand this is uh not a new phenomenon in some fundamental way because big technological breakthroughs have uh not only uh uh expanded the pie but have repeatedly uh reshaped the distribution of income. um usually in an unequal way at the beginning. So I just want to make the following point. When the indust like the railroads good example, right? We had the rubber barons but but then when when we had the industrial revolution in the 19th century the first half of the 19th century was very tough for those who were found themselves in uh in in grim denzian uh circumstances in the uh new factory towns. It was miserable. Uh, and that's what gave rise to the communist manifesto in 1848 and gave rise to the idea that there needs to be a proletarian revolution and so forth. 50 years on, it took a long time, but there started to be a more general rise of well-being, partly from market forces, but a lot through politics, what we called the populist movement and then the progressive movement, uh, and the reforms that came with that. And then the culmination of that in the United States was the New Deal under Franklin Roosevelt. And that brought in social security and uh many other broadbased measures to say we need to share uh this growing bounty. Uh and uh of course FDR talked about one-third of our nation ill-fed, ill-lod ill-housed and that the US was a rich country with a lot of poverty and we should do something about it. And my view is FDR was our greatest president in in our country's history for taking this on as as well as for many other things that he accomplished. Um so politics did come through in the end and the basic idea you have a larger pie if politics works everyone can be better off. I want to say that that is a logic also that applies to these new technologies. I really like them. I think they're incredible. I think that they raise productivity. I think that they do all sorts of good things. I despair of our corrupted political system, our corrupted politicians, our corrupted uh chief executive. This is not how our country should be. It's not intrinsic to the technology is what I'm saying. we can get to these reforms. At some point, our country will figure out, it really will figure out. Uh it's not right to have someone uh asking for a trillion dollar paycheck. That's just not right, Elon. Uh it's not right having 10 people who have $2.3 trillion of wealth. An interesting part of the politics is the concentration of wealth and power is so so so so extreme that actually the vote is overwhelmingly against them. You don't have to be the median voter. You have to be like you said in the 80th percentile or the 90th percentile to say the hell would this come out? And so this gives me some optimism that some political entrepreneur is going to wake up and say, "Look, we're why do we all feel poor when we're in a booming technology?" Well, it's because it's all going to those few people who think that they own the shop. And then we could actually start to get some real reform. So, I'm not I mean I I find it distressing and disturbing watching our politics. I find it really shambolic. But I'm not completely pessimistic because if it's just 10 of them, well, we're uh we're 350 million. We ought to be able to figure that one out, >> right? We we we we outnumber them. Yeah. So, so and that's good to end on. So, you're saying, look, um, while while it may feel dire, and it might be dire for for a while, maybe even a good while, you don't think it's the end of the American experiment, we've been through technological cycles like this in the past. Um, I will note the one question I didn't I didn't raise in here, which which we won't inject in the conversation, was just, yeah, provided AI doesn't kill us all, you know, at some point in here, the the the the Terminator, you know, risk here, but that's that's a discussion for a different day. Um, Professor Saxs, I very much appreciate you doing this, especially under such a tight schedule. Very enlightening, sobering, but important conversation. Really appreciate you having it just out in the open like this. Last question for you. For folks that would like to follow you and your work in between now and your next appearance on this channel, where should they go? >> Oh, thank thanks a lot. I I have a a website called jeffsachs.org. Uh, please do that. Every day there are uh deep or not even so deep fakes of me all over all over uh the websites right now because this is another part of AI. I'm hearing all sorts of things about speeches I never gave and so forth. So if you want the the the real it's at jeffsacks.org. >> Fantastic. And uh when I edit this I will put that up on the screen so folks know exactly where to go. Folks, the link will be in the description below the video as well. All right, and lastly, folks, please join me in thanking Professor Saxs for giving so much of himself and the time that he had here to us. I extend that thanks by hitting the like button and then clicking on the subscribe button below as well as that little bell icon right next to it. And um if you know, he gave us an awful lot of food for thought here. uh if you would like some professional help from a professional financial adviser on trying to think through the repercussions of of the direction of the AI juggernaut, what that might mean for your portfolio, um ways to perhaps maybe take advantage of it to, you know, if there's going to be a winning team in the near term, at least make sure you're on it. Uh and also make sure that the other parts of your investment might be better insulated from some of the risks. Uh then if you don't have a good professional financial adviser already, you know, advising you and all that, feel free to talk to the ones of the ones that thoughtful money endorses, these are the firms you see with me on this channel week in and week out. To uh schedule a free consultation with them, just fill out the very short form at thoughtfulmoney.com. Only takes you a couple seconds to fill out the form. Again, these discussions are totally free. There's no commitments involved. It's just a service they offer to be as helpful to as many people as possible. Professor Saxs, again I cannot thank you enough for uh giving so much of yourself uh to such an important topic uh talking about it so frankly and giving us so much in the limited time that you have. Really appreciate you making the time for this audience. >> Great to be with you. See you again soon.
Jeffrey Sachs: Afraid A.I. Will Wipeout Tons Of Jobs? It Already Is.
Summary
Transcript
If uh the dynamics of the digital economy continue to run pretty much on a market driven basis, then we're going to have large parts of society uh not only left behind but losing in absolute terms. And that's simply because the fact of uh job displacement by smart machines, by smart systems, by robots or by AI systems is not only a prospect, but it's happening already. Uh and it's happening at a pretty significant rate. [Music] Welcome to Thoughtful Money. I'm its founder and your host, Adam Tagert. Will artificial intelligence wipe out millions of the jobs done by humans today? The AI CEOs tell us it's nothing to worry about. But one of the most influential living economists, as recognized by the Economist magazine, begs to differ. He says that AI is already starting to replace human labor at scale, and we don't have a plan for the aftermath. To discuss this very important topic in depth, we've got the great privilege of speaking today with Jeffrey Saxs, professor of economics at Colombia, best-selling author, and global leader in sustainable development. Professor Saxs, thanks so much for joining us today. >> It's a pleasure to be with you. Thank you. >> Well, thank you. And look, uh, you have an incredibly busy schedule. We don't have a ton of time today. I've got a lot of questions for you. We're going to get through as many as we can and, uh, and that'll be it. Um it's been a little while since you've been on the program and so just to kind of level set I want to ask you a general question um what is your current assessment of the economy both domestically and globally right now? We can get into the details but at a high level. >> Sure. Uh basically uh things are very uneven uh within the US economy and globally. By that I mean we clearly have areas of remarkable dynamism in the US around the data centers and AI. This is a boom. Uh it's a lot of capital is being raised and a lot of capital is being invested. uh this is generating uh growth in uh uh certain parts of the economy, certain regions, certain sectors including the power grid which is needed for the big expansion of all of this compute. Globally, we also have uh many parts of the world that are very dynamic, especially in Asia. But in the United States and in the world there are also significant sectors and parts of society that are suffering. Clearly just basic wage income is struggling. The so-called affordability crisis is really a crisis of stagnant or falling real wages. uh in the world as a whole we have fast growth in East Asia uh and in South Asia. Uh we have uh economic crisis and stagnation in other regions a lot of Latin America, a lot of Africa. So the picture and Europe is basically a a rich and stagnant or even contracting economy. So the picture is unsettled. It's varied. It's a a highly disruptive scene because of so many deep changes that are simultaneously hitting our societies. And there is therefore no uh one overarching reality. It depends on who you're talking about, which part of the world you're talking about, which sectors one is talking about. What we have is uncertainty and uh great variation. >> Okay. Um I was going to do this big top down international, you know, down more into the micro US thing, but I'm going to pull up one of my questions from the end of the my list here. Um so you know mixed uneven as you said a lot of people are saying you know K-shaped economy at least you know here in the US um what is your outlook for that that chasm of of unevenness here um especially because so much of the the hopes of the future of the US economy are being pinned on the hypers scale ers uh on AI. Um I think the last time you were on we very briefly touched on both the the promise of AI but also some of the risks and and a big risk being massive jobs displacement and creation of a a large societal under/employment crisis. Um, are you feeling more hopeful at this point in time that that the prosperity that we're gunning for will kind of lift all boats here in the US or are you more concerned that it actually might leave that bottom half of the K, which which really feels more like 80, you know, 20% is the up half of the K, 80%? Is that 80% going to get left further behind here? >> Well, I I think we need to distinguish two things. If you ask will market forces solve this problem the answer is almost certainly no. If the dynamics of the digital economy continue to run pretty much on a market driven basis, then we're going to have large parts of society uh not only left behind but losing in absolute terms. And that's simply because the fact of uh job displacement by smart machines, by smart systems, by robots or by AI systems is not only a prospect, but it's happening already. Uh and it's happening at a pretty significant rate. And what we're hearing uh is uh that many sectors of the economy including uh a lot of skilled uh worker sectors college educated uh work uh is now already finding that there are job freezes and and so on. If you ask a different question uh which is uh taking into account politics and redistribution of income and reorganization of society, could this technology lift all votes in uh at at the end of the day I think the logic says yes uh this could be a kind of breakthrough that is very very broadly beneficial. Then you have to ask a third question. Are our politics on track right now to fulfill that role of underpinning a broadbased uh benefit of AI. And the answer certainly is no. Our politics are completely dysfunctional and becoming more and more dysfunctional. There's not even a rhetorical uh flourish in US politics right now uh aimed in the right direction. No one's asking the question uh in Washington. I mean uh what do we do to ensure broad-based development? Uh we're in a uh in in a scramble for wealth. We're in a completely corrupt political environment of self-deing all over the place including the president's family and everybody else involved. Um so our politics are not aimed at the broader benefits and the market dynamics which will therefore prevail in the next few years are going to be uh uh widening the income and wealth inequalities dramatically. Look, we have 10 people in the United States whose net worth right now is $2.3 trillion. This is incredible. Really? It's that's a lot of money. An average of $230 billion dollar for the 10 richest Americans. Those are all big tech uh uh super wealth holders. Um they not only control that wealth and the companies, but they control their platforms, the media. uh they have all our data and they control the White House and the Congress to a large extent and uh they're absolutely not interested in the question of how to make sure that everybody benefits from all of this. They're scrambling in competition with each other. Uh they uh want to have trillion dollar paychecks. It's a very bizarre world where uh We don't know whether that upper part of the K is 10 people, is 1%, is 5%. It sure is not a majority of of the public. >> Okay. I I've been actually using the term a lowercase eye shape economy that I think we're in the danger of moving to where it's just this little dot at the top. Very few people doing extremely well far removed from the rest of everybody who's just getting left increasingly behind. Um okay so you are I think safe to say concerned about the current trajectory here and I want to continue tugging at the string real quick. Um I I'd just like to get your opinion on um yes there's a big market opportunity that the hyperscalers are rushing after but there also is a sovereign national interest that's been declared in AI you know sort of akin to the nuclear arms race back in the the the 30s and 40s where America is is I believe saying look we've got to win this. this is going to be the technology that that lets America remain the superpower of the future. If we lose this race, say to China, that could dislodge us. Um, A, is that a fair characterization? And B, do do you agree with that? In other words, you we'll talk for a moment in a moment about kind of the repercussions of all this, but do we need to win the AI race as as America? Well, I think the analogy to nuclear arms is correct and uh but that shows the fallacy of the claim that is made that we quote need to win this race. Who's the winner of the nuclear arms race? Nobody. We live under the nuclear sword of Damocles. We're at the edge of extinction. One wrong move, one mistake, everything ends. And believe me, if people think that's absurd hyperbole, they don't know what they're talking about. We live so close to the edge right now. We depend on not very smart people to keep the world alive. We depend on there not being a terrible accident. Now, it's interesting to go back to the uh atomic bomb and the beliefs in it. when the atomic bomb was uh uh first exploded at Los Alamos, New Mexico, and then uh and then used twice uh against Japan for no necessary reason, by the way, because the Japanese were trying to surrender at the time. But uh when it was used, American strategists believed that the US would have a monopoly on atomic bombs for at least a generation. They talked about 30 years. >> It turned out to be four years. Uh every calculation was a wrong calculation. When pushed by Teller, we moved to thermonuclear weapons. We thought that that would be a lasting advantage. The Soviet Union matched us almost instantaneously in the hydrogen bomb. Then came the uh rapid proliferation of these weapons. Then came a rather desperate uh attempt led by President John F. Kennedy uh initially uh to try to get this under control starting with the 1963 partial nuclear testban treaty and then uh the nuclear non-prololiferation treaty then uh a series of arms control agreements between the United States and the Soviet Union. Uh and in other words, the idea of American officials was we'll win this race and then they realized no there are no winners. Uh we are going to get ourselves blown up if we don't get this under some kind of uh mutual control. uh by the time uh there has been some control there are nine uh nuclear countries right now and nothing's really under control. I would say uh our situation is extraordinarily fragile. Uh the bulletin of atomic scientists puts out the uh doomsday clock which uh is the uh visual uh representation of how close we are to disaster. And the doomsday clock is set at 89 seconds to midnight, meaning the closest we've ever been in the whole atomic age or nuclear age to nuclear Armageddon. So now to come to AI, what you expressed is absolutely the view in Washington. Uh and I happened to listen a few years ago to a chillingly stupid discussion of two NATO generals at one of these, you know, strategic conferences that I attended once in a while. Um and they were saying to each other, "Yes, we must win the AI race with China. We're gonna win, we're gonna do this. Uh this is so naive. Uh China produces so many more PhDs than we do. Uh just the scale difference is vast. The talent is enormous. Uh and you get advances in the US and then you know something we never heard of before called Deep Seek comes along and says we're here. uh and then suddenly you have all over China uh you know topf flight top performing LLMs and then the United States says oh we've got a chokeold on two nanometer chips on Nvidia chips and then you get almost weekly a report of some new computer architecture that does an end run around those chips. You know, the United States has some monopoly on this. Are you kidding? This is ridiculous. Of course, it doesn't. Are we going to win the AI race against China? No. Are we going to lose the AI race? No. This is not that kind of thing. This is not, you know, this is not a race. This is a core technology in which there's going to be a lot of expertise all over the world. The idea that we're going to have some chokeold over this is absurd. So let's get it out of our heads at the beginning. And then we should ask, so given the analogy with nuclear arms, where is this dangerous? What should we do about it? And what should we do diplomatically about it? We should move more quickly to that with the nuclear arms we did not get to even the first treaty, the partial nuclear testban treaty until 18 years into the atomic age. Can we do a little bit better than that this time? That that's what I hope. >> Okay. Uh well, I I'll add my hopes to that. Um the reason why I asked the question I think you answered it is um while you and I have that hope right now it doesn't seem that that's at the forefront of the mind of the folks in Washington right so >> not at all >> right so so we have these sort of twin engines propelling the AI buildout um both the market the size of the market opportunity so the capitalists are sprinting after that and then the the state sector is saying look we're going to do everything we can to clear the playing field for you because we want to win, right? >> And and I would add one more very obvious and specific point, which is that it's not that we want to win the economic race. It is that the Pentagon is salivating. Uh we need AI and all our weapon systems. So this is becoming completely a a national security uh tool as well or what they call national security. I would call it national insecurity because everything we we are doing is raising our insecurity. But this is an extraordinary technology because it is at once everything about the uh production processes of the economy. So AI is be becoming embedded in all production processes. It's embedded in the service economy. It's embedded in our online lives. automatically uh but it's also it's embedded in our media which happens to be owned by these very uh platforms whether it's Meta or Google or Microsoft. So they own the media uh they own the platforms. Uh it is uh mortgaging our privacy because they actually own our keystrokes also uh in ways that I can't really uh understand but I know it's there quite obviously. Um and it's also they own the White House. They appoint they appointed the vice president for God's sake. They told Donald Trump who his vice president was going to be uh at a dinner in uh San Francisco. Uh it's not even subtle this stuff. Uh so they own the politics and unlike the nuclear arms by the way where basically in a way the generals just pushed away the atomic scientists after a while or hired some of them. uh but said we own this. Uh the Pentagon doesn't own uh AI. Uh this is owned by Peter Teal and Elon Musk. The Pentagon is contracting out uh to these companies which hold the technology now. So this is really an extraordinary concentration of private power. Uh and uh you add in yes they're making way for they're making way for uh all of this. Um uh but they're bas basically uh turning over politics, national security, mass media, the payment system, by the way, which is going to be a digital crypto something or other payment system that maybe they'll let us know about at some point. Uh and uh all of our media. So >> pretty into the the hands of a private jet. It's almost sort of the the Tony Starkization of of the economy if you if you get the reference from the Marvel movies. >> Yes. >> Um Okay. So So again, I think you're you're proving the point I'm trying to make here, which is this is a juggernaut that is in motion and and perhaps even still building momentum. So, um, as much as you and I would like there to be more cautionary, you know, steps in here, probably not going to happen anytime soon, which gets us to, okay, so um, probably doesn't bode pretty well for that bottom 80%. Right? Like we we might get cheaper entertainment. Okay, hooray. You know, maybe we enter the Ready Player One era where we get free VR goggles and we can spend all our time in the virtual world, but we're out of jobs. Um, you know, we're we're we're out of purpose. we're we're we're out of power on a relative basis because it's concentrating in in in that that dot of the lowercase die-shaped economy that we talked about. So, um I guess a lot of concerns around that. Now, you had said um there's a way to do it differently and um I I guess first I just want to imprint on the listeners that a really smart gentleman like yourself who's very connected and studies all this stuff is quite worried about where society could go here. and and maybe you might want to I might ask you just to make a comment or two on on what you think that future looks like. Um where where the the the bottom 80% get even further dispossessed from here. And then secondly, I'd like you to talk about your vision of how we could do it differently. And and I guess where I'm going with this is what I hear from the AI fanboys is hey AI is going to create a lot of jobs as well maybe but I think on on a fractional basis it's going to be a lot fewer than it's displacing. Um so they say well even if it does we'll have universal basic income and everybody can go focus on being their best selves. I I think history is pretty clear. It's my opinion. You might feel differently that that that you know giving people a subsidy to basically do nothing creates a dependent class that that you know ends up investing more in their vices than their virtues. So uh in the in the better way we do this if technology is still doing all the work, how do we make this more equal? Is is it just redistributive policies or is there something in here that gives the bottom 80% a role? >> Well, uh first of all, I don't know the answers to this and I don't think anybody does. I haven't uh seen or heard uh anything that I find uh accurate and compelling uh that says that is how the future will play out. So I just want to uh caution on that obvious point at the beginning. Second, we've been in the midst of related changes for uh at least uh 60 years. Automation became a a thing a noticeable thing in the 1960s. uh it became a a matter of a lot of public discussion already 60 years ago and that was not wrong uh to worry about that because what we've had during a 60-year period is a bifurcation of society uh into three parts now. One is the the super uber halves, those who have more than is absolutely imaginable. uh and then the successful halves which is uh broadly speaking the professional class in America typically people who have had an advanced degree uh of some kind an MBA or law degree or medical degree or other professional degree uh and the uh broadly speaking somewhere between the high school completion and bachelor's completion depending on a lot of specifics but finding stagnant wages by and large uh and um very big difficulties of uh accumulating wealth. Uh and one of the most evident of those difficulties is homeowning for young people. uh which is deemed to be a you know pretty widespread crisis for a lot of young people. So we've already b triifificated you could say uh we already have three different uh groupings in society that live very very different lives. If you ask uh my cohort in the professional world how are things and how they been over the last 20 years. Yeah. good, very good, good incomes. Uh, stock market's been going up, lots of new opportunities, tech has been good. Uh, feels all very good. Actually, >> my house goes up in value every year. >> Yeah. If if you ask someone who basically doesn't have saving and uh if they've been very uh unlucky, but that's a lot of American uh people aged 25 to to 50 um had a high school degree uh went to work, lost their job and so forth. This is not easy. Uh health care costs are soaring. hard to save anything. What's called the affordability crisis is really again a stagnant wages crisis that comes from the fact that basic jobs have been under uh replacement dynamics for decades, not just now. It's not just AI. Uh it's robots. It's uh smart systems of many kinds that have been operating. So we're we're already in the midst of this. What is interesting is that and and it showed up in some you know everybody understands this now I hope in some very dramatic ways life expectancy gaps between having a university education and a high school education is eight or 10 years of life expectancy. Can you imagine this? Completely different lives within our society. And what is interesting, strange word, but if you're an analyst, our political system to this day hasn't reacted to this. Uh really, we have been in a kind of paralysis. Uh we had, for example, a huge fight over Obamacare. Obamacare is a half a 1% of GDP. It's a tiny little change of the national income. It was uh you know fullout war for a decade in the political system. In other words, our political system is paralyzed, unable to address big issues like the changing income distribution according to educational attainment. And now we we have a different changing income distribution. It's not really most likely uh between skilled and unskilled or high educated and lower educated. It it's actually more broadly between those who have uh financial wealth and those who don't because stock market has you know maybe there's some bubble in it but it's been booming in part because more and more national income goes to profits goes to the returns to patents goes to software goes to the monopoly or oligopoly ownership of a few digital platforms and if you happen to have enough income that you have been saving out of that income and you've been investing in your uh in in your retirement accounts and so forth, you've been a beneficiary of that wealth. But if your income has been so low that you're my god still trying to pay your student debt or uh trying to pay an unexpected uh healthc care emergency, you don't have wealth. You haven't benefited from the shift of national income from wages to profits. You've lost from that. And so we are now in an intensifying phase of this growing inequality. And I don't see the slightest evidence at all. No evidence right now that Washington's on the case. I don't hear one word about it. I don't hear even a conceptualization the Democrats. I mean, it's stupid even to say they don't represent they they don't represent the workers, but MAGA doesn't represent the workers either. The whole thing is shambalik. Both of them, Democrats, if they represent any class, they represent professionals. Uh, and MAGA represents the Trump family and his friends. Uh, and so we don't have a politics of facing up to this question. We have no serious discussion. We don't even get to your question of is there a better way than redistribution or predistribution to address this. I'm pretty sure by the way that part of the answer is actually what we would call predistribution which is that at least everybody gets education with quality and health care with quality not out of their personal consumption spending but out of their citizenship. I do think that this is absolutely right that we have as a matter of citizenship that you can get your health care without needing a MD or PhD to understand our health care system and without having to uh you know go into um uh irretrievable irreversible indebtedness to to pay your medical bills. But we're not even there at that point yet. >> Okay, Professor Zach, I have so many questions for you, but we're running out of time. So, we'll just have to reserve them for the next time to kind of land the plane here. I hear you say, look, problems big, maybe intractable. Our >> No, I don't say intractable. I just say big big big. >> Okay. But very very big. So, in the near term, you know, it might feel intractable. Um and the the our political system is not going to ride to the rescue anytime soon. >> Our political system as it's now constituted. Yes. >> As it's now constituted. Yes. >> So um uh I mean there's so many questions I know the audience would like me to ask you here, but just to make it practical for somebody watching this who is saying, "Whoa, this feels really heavy and kind of depressing." Um do you have any advice to them to at least say look here are some things you could could think about um considering so that you are less likely you're reducing your odds of being collateral damage to all of this if this juggernaut continues for the next 5 to 10 years. >> Yeah. Uh let me let me just say in in the long picture which is not a very satisfactory answer but uh it's also important to understand this is uh not a new phenomenon in some fundamental way because big technological breakthroughs have uh not only uh uh expanded the pie but have repeatedly uh reshaped the distribution of income. um usually in an unequal way at the beginning. So I just want to make the following point. When the indust like the railroads good example, right? We had the rubber barons but but then when when we had the industrial revolution in the 19th century the first half of the 19th century was very tough for those who were found themselves in uh in in grim denzian uh circumstances in the uh new factory towns. It was miserable. Uh, and that's what gave rise to the communist manifesto in 1848 and gave rise to the idea that there needs to be a proletarian revolution and so forth. 50 years on, it took a long time, but there started to be a more general rise of well-being, partly from market forces, but a lot through politics, what we called the populist movement and then the progressive movement, uh, and the reforms that came with that. And then the culmination of that in the United States was the New Deal under Franklin Roosevelt. And that brought in social security and uh many other broadbased measures to say we need to share uh this growing bounty. Uh and uh of course FDR talked about one-third of our nation ill-fed, ill-lod ill-housed and that the US was a rich country with a lot of poverty and we should do something about it. And my view is FDR was our greatest president in in our country's history for taking this on as as well as for many other things that he accomplished. Um so politics did come through in the end and the basic idea you have a larger pie if politics works everyone can be better off. I want to say that that is a logic also that applies to these new technologies. I really like them. I think they're incredible. I think that they raise productivity. I think that they do all sorts of good things. I despair of our corrupted political system, our corrupted politicians, our corrupted uh chief executive. This is not how our country should be. It's not intrinsic to the technology is what I'm saying. we can get to these reforms. At some point, our country will figure out, it really will figure out. Uh it's not right to have someone uh asking for a trillion dollar paycheck. That's just not right, Elon. Uh it's not right having 10 people who have $2.3 trillion of wealth. An interesting part of the politics is the concentration of wealth and power is so so so so extreme that actually the vote is overwhelmingly against them. You don't have to be the median voter. You have to be like you said in the 80th percentile or the 90th percentile to say the hell would this come out? And so this gives me some optimism that some political entrepreneur is going to wake up and say, "Look, we're why do we all feel poor when we're in a booming technology?" Well, it's because it's all going to those few people who think that they own the shop. And then we could actually start to get some real reform. So, I'm not I mean I I find it distressing and disturbing watching our politics. I find it really shambolic. But I'm not completely pessimistic because if it's just 10 of them, well, we're uh we're 350 million. We ought to be able to figure that one out, >> right? We we we we outnumber them. Yeah. So, so and that's good to end on. So, you're saying, look, um, while while it may feel dire, and it might be dire for for a while, maybe even a good while, you don't think it's the end of the American experiment, we've been through technological cycles like this in the past. Um, I will note the one question I didn't I didn't raise in here, which which we won't inject in the conversation, was just, yeah, provided AI doesn't kill us all, you know, at some point in here, the the the the Terminator, you know, risk here, but that's that's a discussion for a different day. Um, Professor Saxs, I very much appreciate you doing this, especially under such a tight schedule. Very enlightening, sobering, but important conversation. Really appreciate you having it just out in the open like this. Last question for you. For folks that would like to follow you and your work in between now and your next appearance on this channel, where should they go? >> Oh, thank thanks a lot. I I have a a website called jeffsachs.org. Uh, please do that. Every day there are uh deep or not even so deep fakes of me all over all over uh the websites right now because this is another part of AI. I'm hearing all sorts of things about speeches I never gave and so forth. So if you want the the the real it's at jeffsacks.org. >> Fantastic. And uh when I edit this I will put that up on the screen so folks know exactly where to go. Folks, the link will be in the description below the video as well. All right, and lastly, folks, please join me in thanking Professor Saxs for giving so much of himself and the time that he had here to us. I extend that thanks by hitting the like button and then clicking on the subscribe button below as well as that little bell icon right next to it. And um if you know, he gave us an awful lot of food for thought here. uh if you would like some professional help from a professional financial adviser on trying to think through the repercussions of of the direction of the AI juggernaut, what that might mean for your portfolio, um ways to perhaps maybe take advantage of it to, you know, if there's going to be a winning team in the near term, at least make sure you're on it. Uh and also make sure that the other parts of your investment might be better insulated from some of the risks. Uh then if you don't have a good professional financial adviser already, you know, advising you and all that, feel free to talk to the ones of the ones that thoughtful money endorses, these are the firms you see with me on this channel week in and week out. To uh schedule a free consultation with them, just fill out the very short form at thoughtfulmoney.com. Only takes you a couple seconds to fill out the form. Again, these discussions are totally free. There's no commitments involved. It's just a service they offer to be as helpful to as many people as possible. Professor Saxs, again I cannot thank you enough for uh giving so much of yourself uh to such an important topic uh talking about it so frankly and giving us so much in the limited time that you have. Really appreciate you making the time for this audience. >> Great to be with you. See you again soon.