00:00:00
Speaker 1: This episode of the Chuck Todd cast is brought to you by Chapter. Now, almost everything you hear about Medicare is somebody quietly trying to steer you somewhere. And you're skeptical of that, as you should be. I get it. I mean, there's a federal lawsuit going on right now over exactly this. Companies that branded themselves as unbiased while allegedly pointing people toward whatever would pay them the most as some sort of broker. So if your default is to distrust everyone in this space, honestly, I get it. I'm a bit distrustful these days myself. That's the whole reason I want to tell you about Chapter. They're independent Medicare advisors whose entire model is that their advisors are not paid to push any one plan. They're just trying to help. Yes, there's compensation that goes with it. but it's across the board is what they're trying to do. At the end of the day, they're advocating for you. They're not the insurance companies themselves. They're an actual person who actually goes through your doctors, your prescriptions, your needs, and then compares them to every single plan available to find the best one that will work for you and hopefully save you a couple dollars too. And here's the part that sets chapter apart. They don't disappear once you've been in They will keep advocating for you. They're kind of like almost like a concierge service. You're just there. I need help. You're like customer service. You know, the plan, you know, help me navigate this again. So you ever get hit with a wrongful medical bill, they'll help you fight it. They want to stay an advocate for you even after you've done this. So look, I get it. I think there's all sorts of space. You see a lot of people are throwing themselves in this healthcare space. What I like about these chapter folks, I've spent time with them. This is a very transparent organization. It is Medicare and all of its affiliated stuff can be very confusing. The whole point of this is to make this easier and less confusing. OK, so look, take the skepticism you already have in all of this, point it at your own plan for 20 minutes. It's free. There's no pressure. And then call chapter today at 980-734-4200. three, nine, eight, five. So my guest today is somebody that I've come to admire and read simply by reading his sub stack. It's Jim Pethokoukis. And he, I now use his vocabulary. Sometimes he is the people that subscribe to his publication. He refers to as up wingers and it, I guess the simplest way to explain that is he's an optimist about the future. That doesn't mean he's naive about how we get there. His day job, he's a senior fellow at the American Enterprise Institute. And you guys know how I feel about AEI. AEI is sort of.
00:02:58
Speaker 2: Where.
00:02:59
Speaker 1: I originally thought conservatives, many conservative people wanted to be. It is in a fight for the definition of conservatism these days, but we could set that part aside. But the fact is, a lot of people I admire that work over at AEI, and Jim is one of them. And he joins me now. I want to talk a little bit AI, a little bit of optimism, and a little bit about what he does for his day job, which is to be thinking about economic growth and innovation. Jim, it's nice to meet you.
00:03:29
Speaker 2: Chuck, great to meet you. Thanks so much for having me on.
00:03:32
Speaker 1: So let's start with some of the language you use, because like I say, you write a lot about AI these days. You write a lot about what's going on in the world of tech. And you do come at it not like you're just optimistic for optimistic's sake. But you're always glasses half full, upwinger is how you describe yourself. So explain that language first, upwinger, downwinger, what you mean by this. And am I being a bit superficial in describing your worldview and outlook?
00:04:08
Speaker 2: I think most of what I write about, the reason I'm interested in AI is because I'm interested in economic growth, the ability for good. I work at a think tank, so I think public policy matters to help create the conditions where economies can grow faster, improve living standards.
00:04:30
Speaker 1: These economies can come up with.
00:04:32
Speaker 2: Great things and great new goods and services, and we can come up with technologies to solve our problems. So I have a great interest in that. There's a famous saying in economics, once you begin thinking about economic growth, it's hard to think about anything else. And that's sort of always been me. I'm like a sci-fi person from when I was a kid, so of course I've had that aspect. But what really got me, I think, especially interested is sort of this period since the global financial crisis, where we've had both an economic shock, and since then, we've had the pandemic shock, and now we've had this inflation shock. And that's and that's sort of pessimism where people begin thinking like their kids are not going to be better off than them. That, you know, the American dream is dead. You know, huge numbers saying the economy is on the wrong track. So all that's been happening at the same time. And I think it's causal that we've had a very. A weak economy by historic standards. And often, I'm sure you get this question, which is, what explains the rise of populism? Well, to me, the easiest explanation is we've had a very disappointing economy peppered by various shocks. Now, it's almost 20 years. For me, the Great Recession, financial crisis, it doesn't feel like it's that long ago. But it's been 20 years of tumult. And economic growth can not only solve problems, give us things, right things we want to buy, but make us optimistic about the future. And that's why I'm excited about AI, because I think it can help restore a faster pace of economic growth. And I think really change our politics.
00:06:24
Speaker 1: You know, it's interesting you say it. It's 20 years. I kind of think I go back to 9-11. I go all the way back. I think it's sort of, we've been in this tumult and I think about this a lot with millennials, right? And sort of the world and they came into versus the world. You and I came, right? You know, when we came of age, I'm not, I think I'm presuming your age correctly. America was winning. We were the lead country. Democracy was winning, right? The whole end of history business, right? This is it. We got it. We won it. And then starting with 9-11, and it feels like the first 25 years of this century have been nothing but shocks, right? You started at the Great Recession. That's why I would move it back a couple, you know, essentially seven years sooner, because it has been shock after shock after shock. And we may categorize it, well, here's economic shock over here. Here's global, you know, sort of shock over here. but for the average person, they're sort of feeling it the same. And I guess, I guess the question is, is this a, because I'm also one of those folks who accepts the premise that, you know, if you look at things from 60,000 feet, they've actually done nothing but get better. We've gotten more people out of poverty. We've keep, we are saving more lives than ever because of advances in health and things like that. Why do you think How much of this is sort of just our information diet that puts us in a bad mood constantly?
00:08:05
Speaker 2: I don't think that helps. I don't think that doom scrolling helps. And there's like a cognitive bias showing that people tend to– we're more likely to anchor on bad news than good news.
00:08:17
Speaker 1: Think about the compliments. Do you remember the positive compliments or the one negative commenter you have on your newsletter? Right.
00:08:23
Speaker 2: I mean, who– like– The most famous example is who calls in the radio shows? People who complain. Newspaper headlines aren't playing land safely. It's playing crashes. So people tend to focus on negative news. And I certainly think that that's playing a role. And I think the current debate about AI... You know, listen, ChatGPT came out, you know, late November 2022. And as I remember it, we had about two weeks where people, you know, there were stories like, hey, there's this weird new technology and you can have like Winston Churchill do a rap song and crazy videos. And it was, you know, wow, people are playing with it. And then quickly it became, Sure, it's kind of fun, but it's probably going to take your job. And then after it takes your job, it will probably kill you. Hey, have you seen The Terminator? And those two through lines, massive economic disruption for a country that's had a lot of disruption, and then that's sort of what they now call existential risks, have just dominated the conversation where the potential upsides are sort of Yeah, OK, maybe it'll improve growth. Maybe it'll cure this, cure that. Gets kind of a hand wave where by far and obviously one reason is because the AI companies themselves have been saying this. But that focus on the downside, I think, is really destructive. And I was making the point earlier about like the impact of an economy that's growing, but not so fast that people perceive it as growing, and not so fast they have a lot of optimism about the future. I think it's been super destructive for our politics for 20 years, and I don't want to see what our politics looks like 20 more years of wages rising, and right now they're falling. I think we underestimate the impact There's a wonderful study about why people support democracy. And the basic reason, not because we're all philosophers, we spend a lot of time reading Founding Fathers, is because democracy, liberal democracy, delivers. It has delivered a better standard of living. It has delivered a society that feels safe. And if it's not delivering that, you start hearing more voices saying, maybe there's another way, even though history suggests those.
00:10:45
Speaker 1: Other ways are pretty terrible in the long run. Yeah. Look, I don't know which is cause and effect here. Do we not have the leadership to meet the moment because we're so discombobulated, right? We're not quite sure where to go. Is this a chicken and egg? Do we have the wrong leaders for the wrong time? Or do you think we eventually realize we have the wrong leaders and we're going to start looking for ones that are stabilizing forces. I mean, right now, what we could use, right, with AI, it's pretty obvious to me, right? We need an empathetic president, right? He's not, empathy is not what he does. He's never done empathy. This is, to me, the greatest, it's got a lot of flaws. That's a big one. If you have a president with no empathy and doesn't know how to do it, you can't do part of the job. That's just, and he is just incapable of this. But this needs an empathetic moment, like, hey, This technology, it's amazing and it's scary at the same time, right? We need, and he doesn't have that in him. And I don't know where else we would turn for this person, for, you know, essentially we need this, you know, parental type figure as we do this massive transition. And is that something you expect voters to seek out? Or are we just got to hope we find a leader to meet the moment?
00:12:14
Speaker 2: Yeah, well, I mean, sort of the worst thing Like the worst political strategy is to tell people you're wrong about the way you perceive your life.
00:12:24
Speaker 1: That's always bad politics. When you tell the voter they're wrong– by the way, they might be wrong. It doesn't matter. The customer is always right. And in this case, the voter is the customer.
00:12:39
Speaker 2: You can't say, I know you think life's getting worse, but listen, I have these great three statistics. I'm going to show them to you. No, I mean, that doesn't work in real life. I guarantee you, as you know, it does not work on social media. You certainly have to acknowledge these are actual concerns. And I think some of what you're seeing happening, you know, in the administration, like, you know, with AI is, you know, they just announced this sort of, you have an AI commission and there's kind of an AI czar and they're going to come back with a report. I think all of that is kind of trying to like meet those concerns.
00:13:18
Speaker 1: And it feels like papering, they don't seem serious about it. And it's not, it's letting the companies decide this, which- You know, after our experience with social media, I understand why most of the public just doesn't trust these tech companies to do the right thing. No.
00:13:34
Speaker 2: Right.
00:13:35
Speaker 1: So I... So.
00:13:37
Speaker 2: One example I like to think about is the Drudge Report, which still exists. It's not what it was. But if you spent years reading the Drudge Report, what you read was Silicon Valley people are rich and weird. And I think a lot of people think Silicon Valley people, even richer, and they're actually getting even weirder. So do I trust these people? No, there's not a lot of trust. Also, obviously, these are the same people saying... We have this great technology and there's X percent chance. And that percent chance doesn't seem to be one in a billion that it'll do something bad. So a lot of this anxiety is being generated. And yeah, would it be great to have a president who you felt like had a really good grasp on like what this technology did? Like how what are like the realistic upsides and downsides and met that in a way where you you had confidence? I guarantee I have a more. you know, light touch, more confidence in markets approach toward regulation than probably the average American, maybe, you know, many of your listeners and viewers, uh, But I also feel like I understand sort of the limits to that approach. I understand why people are skeptical of that approach. And I think it would be great to have a president who just didn't say, listen, this is going to be like super terrific self-regulation throughout some, I mean, What came out of that meeting where the president had this meeting with tech leaders and it was like, it was typical Trump speak.
00:15:14
Speaker 1: It's going to be like the.
00:15:15
Speaker 2: Most amazing self-regulation ever. And it's going to be the super best. That stuff obviously is not calming. And I think this is an important technology. And I and I hate it that like 70 percent of people are seem to be like super worried about it.
00:15:31
Speaker 1: And it's growing. That number is only growing. I mean, look, and I sort of put the blame on the administration here because in some ways they don't have the credit. They don't have credibility at the moment with a large chunk of the public. So their optimism actually boomerangs. Right. It's like the more he trumpets it, the more there's resistance simply to resist him, I think. So, right.
00:15:55
Speaker 2: And listen, there is deep skepticism among many Americans of the president. So if he's going to say, hey, these companies can sort of self-regulate like the waters, you know, that's like that's already spoiled for half the country who aren't going to believe it, which is a problem.
00:16:14
Speaker 1: I think.
00:16:15
Speaker 2: But I think if you can like set that aside, there's actually like a lot of– and if you are concerned about AI, like stuff is happening. Like these– listen, I don't have a lot of confidence and ability of Congress to.
00:16:30
Speaker 1: Regulate any of this stuff. I've been waiting for my– remember the Privacy Bill of Rights? Remember that movement? It was about 2020. I remember it went back almost 2006. These are good examples, yeah. Yeah, and the Privacy Bill of Rights in the 2006 midterms, I remember a candidate in Montana named John Tester that like trumpeted this idea of a Privacy Bill of Rights. And this was a thing, and you could see this. We've yet to see anything like that, right? We never, you know, Congress couldn't figure out how to protect our privacy from Web 2.0. You know, so I'm with you. I don't know what they can do, and yet, you know, if not them, who?
00:17:14
Speaker 2: I think that should be your baseline. For all the concerns about social media, whether it's privacy, whether it's how they moderate, whether there's bias, I mean, really, tech policy up until the end of November 2022 was about exactly those issues. So if you did tech policy in Washington, you weren't talking about AI, you were talking about social media. bills proposed, the public expressing concerns, nothing happened. So I don't want to overlearn that lesson, but I think that's a good baseline that you're not going to see government action, certainly not under this president who may do something through the executive order. But I think if you're looking for Congress to pass a big bill in the next two years signed by President Trump, that's not going to happen. So- The president seems very optimistic about self-regulation. And so if you're and if you're concerned about it, you better hope he's right, because that's what we're that's what we're going to get.
00:18:18
Speaker 1: And I think there are some reasons to be positive about that. Well, give me some reasons to be positive about it. Right.
00:18:24
Speaker 2: Well, I think, listen, people think about regulation and, you know, I've certainly, you know, seen this in the media. They really underestimate regulation. the risk companies undertake when they release a product, either that product doesn't work and no one wants to buy the product, or the product's harmful and they get sued. Those issues, reputational risk and actual being sued, are not trivial incentives to be safe. And already, we're seeing with these companies, which, yes, they've asked government to act. Government doesn't seem like it's going to act anytime soon. And they're taking actions. They're I mean, they did agree to bring in these independent monitors to verify what they're doing, their safety practices. More importantly, they've begun slowing down their research. They've begun slowing down releasing models. Incrementally, I think they would love for government to tell them what to do. It's kind of a CYA thing, and they say government's forcing us to do it, and they're forcing everybody to do it so no one can catch up to us. But with that not happening, they are taking actions. Now, listen, if your goal is banning all super intelligence at the pursuit of super AI like Bernie Sanders, it's not going to be nearly enough. But I think we have a new technology. It's fast evolving. I don't think you want to rush to start banning things, 100% pausing things. I think for better or worse, we're going to have to learn on the fly. Sam Maubin was talking to Politico, and he said, there needs to be like, things are going to go wrong, right?
00:20:06
Speaker 1: Hey, this episode of the Chuck Todd cast is brought to you by Ethos. Look, let me tell you about life insurance. As you know, Ethos is an incredible portal for helping you find the right life insurance policy. I'm a huge believer in life insurance. I wouldn't have gotten to college in Washington, D.C. without life insurance. My father had a policy buried somewhere that we discovered, and it wasn't a lot of money, but it was an important bucket of money at a time that my mother, newly widowed, single mom, Me, I'm 16. It was of enormous, enormous help at that time. And this is why I believe in life insurance. You don't know what could happen to you on any given day. This is literally an insurance policy for that. So let me tell you a little bit about Ethos. Ethos makes getting life insurance fast and easy. It's 100% online. You get a quote in seconds, you apply in minutes, and you can get coverage the same day you begin this process. There's no medical exam. You just answer a few simple health questions. You can get up to $ 3 million in coverage. Some policies are as low as $ 30 a month. And you'll get your lowest rate from their network of trusted carriers. Ethos has gotten 4.9 out of 5 stars on Trustpilot with over 5,500 reviews. So, take 10 minutes to get covered today. Life insurance through Ethos. Get your free quote at ethos.com slash Chuck. That is E-T-H-O-S dot com slash Chuck. Application times may vary. The rates may vary. That Trustpilot rating was as of August of 2026. Well, that line was a little haunting. I mean, look, I totally got what he said. And in fact, I kept thinking about airplanes. And I want to, I want to, I think that metaphor is apt. Because, you know, air travel first was very much regulated by government. And then we had commercial airliners. And I wonder if there's something there. But you know how some people are the wrong people to say certain things? Hey, Sam, you're the wrong person to start talking about things could go wrong and we just have to accept it. It's like, he's right. It's going to happen. We know this. There's all sorts of stuff that happens. Like, you know, I get that. but that's why you want government. I mean, let me give you a metaphor in the airlines. Um, would, would you be comfortable getting rid of the NTSB and the FAA and letting the airlines figure this out on their own?
00:22:35
Speaker 2: Uh, I'd probably be more comfortable than many people would be. I'd probably be more comfortable.
00:22:42
Speaker 1: You think you'd be the minority on that?
00:22:44
Speaker 2: A lighter touch. Listen, everybody's going to have a different risk tolerance. And mine has changed over the years where I had small children. I had very low risk tolerance. I have adult children. It's gone up a bit. But I think the issue you're identifying, what is our societal risk aversion level? What will we demand of these companies? Is it Is it zero accidents? Is it zero serious accidents? I mean, that could be something we could try to demand and we'll probably get like.
00:23:14
Speaker 1: Isn't that our demand on airplanes? We've decided that any crash needs to be researched to figure out how to make sure that never happens again. I think that's great. So we do have a pretty high bar of intolerance on that when you say.
00:23:27
Speaker 2: Listen, when there are any kind of serious incidents, would I be for a serious investigation?
00:23:35
Speaker 1: I would.
00:23:36
Speaker 2: I especially would as a way of, I think, building public confidence. I think that matters. And again, I said I don't want to overlearn the lesson of social media. But there's the other lesson that I bring up again and again in my newsletter, and I don't want to overlearn it. is our experience with nuclear power.
00:23:58
Speaker 1: I'm glad you brought that up. I think we screwed us.
00:24:01
Speaker 2: No, no, no.
00:24:02
Speaker 1: I was where I was going to go if you didn't, because I'm with you. I, you know, I feel like we lost out on this incredible technology. We should have clean energy already. We have this clean energy technology. And, you know, Three Mile Island was bad, but it's, it, it's always been what it could have been, not what it was. And I know people don't want to, but I think that's where you're going, are you not?
00:24:37
Speaker 2: It is, which was a massive societal overreaction that failed to consider what actually happened, that failed to consider the opportunity cost, which now seems very obvious. I mean, we spent years thinking about about climate change. Now, would we be talking about it if we had 1,000 nuclear reactors coast to coast?
00:25:00
Speaker 1: Maybe not.
00:25:01
Speaker 2: Would we be worrying how we're going to power the AI revolution if we had coast to coast nuclear reactors, which now might be fusion reactors? Who knows? You can mention Three Mile Island. You can mention the meltdown in Japan in 2020 where they shut down all the reactors. And then a few years later, they started firing them all back up again because they were having power shortages. More people died from the burning coal and oil and the pollution than they ever did from Fukushima.
00:25:33
Speaker 1: That tsunami, that Fukushima incident, I believe, correct me if I'm wrong on the timeline, But that accident happens, I think a month after Barack Obama signed off on the first new nuclear reactor that we were gonna build. And it was like, here you had, it was basically Obama had brought the left along on this, had brought the environment. And the environmental community is still now divided, is now more divided on nukes where you have the anti-nuke crowd, but then you have those that are like, no, it's the clean energy answer, right? And it was the momentum was on the side of the pro-nuke side. And then that happens. And it literally the world shut down nuclear power, it felt like.
00:26:17
Speaker 2: You know, Germany, which nothing happened, they also shut down their nuclear reactors.
00:26:23
Speaker 1: Yeah, and they got into bed with the Russians. That worked out real well. Right.
00:26:27
Speaker 2: So, you know, there are, you know, opportunity costs or unintended consequences. But Japan figured that out very, very quickly. And they've begun restarting those reactors. So, I mean, the lesson here is, one, there are unintended consequences. There are opportunity costs. And it's so easy, again, I think just because we're naturally inclined that way, to just not assume those opportunity costs of the good stuff we won't have as we hyper-focus. Not to mention, and it's something I write about, listen, for 50 years, I mean, so much of our media has been saying, be careful of AI, be careful of robots, they're.
00:27:12
Speaker 1: Going to get you.
00:27:13
Speaker 2: And you see those kinds of cultural touchstones repeated in many, many news stories. And that's, I mean, if we cannot envision, listen, AI, there will be accidents. Some people will lose jobs. It will be disruptive. But if we cannot imagine a future that's better because of it, then we're not going to accept it.
00:27:38
Speaker 1: Then we won't.
00:27:39
Speaker 2: Then forget about a ban. People won't use the technology. Now, people are using the technology. Everyone's frightened of it, but we're using it massively. So there's an economic theory called revealed preference. Our revealed preference is we find AI pretty useful. Also, if anyone asks us about it, we'll say we'll hate it.
00:28:01
Speaker 1: Is there good enough statistics to compare sort of the acceptance of the personal computer, the growth of using, you know, putting a computer in the house to, you know, in hindsight, I didn't realize how much of a upwinger my late father was, but we had an Apple IIc in our house and we didn't have a lot of money, but he really wanted.
00:28:23
Speaker 2: To get- That was an expensive computer.
00:28:25
Speaker 1: Yeah. He really wanted to try to, he had heard about this system where you could plug your phone in and you could get stock ticker stuff 15 minutes delayed. And this was in like 1985, I want to say 86. You know, first we tried a TRS computer, RadioShack computer, because we were on a budget. That didn't work out so well. So finally, we got an Apple IIc. And I guess what I would At the time, I think I was on the early side of friends with computers in the house. Do we have a comparison of personal computers in the house and AI adaption? How fast are we seeing this shift into adopting AI? It's incredibly fast.
00:29:16
Speaker 2: It's incredibly fast on a personal level. It doesn't seem like it, but it's also very, very fast on a business level, which is what a lot of this is about. The reason all this money is being poured into building data centers is under the assumption that businesses will be able to use this technology in a really productive way, far more than even the PC and the internet. And people are like, well, it's been four years. Where's Like, where's it at? Like, where's all the big gains for business? And they're not reporting them yet. In fact, I spend a lot of time listening to business earnings calls. And whenever they talk about AI, it's still mostly like in the future tense. Like, we have a lot of confidence this will generate revenue. It will help us come up with new products. But it's still very future-oriented. And for some people, like, well, this is a phony technology. It's, you know, we're creating this investment bubble, and it's not useful. But the history of technology takes a long time for it to become productive, for businesses to figure out how to use it. So, like, based on our expectation scale, it seems super slow. But whether it's personal adoption or business, it's actually happening extremely quickly, though it may not, again, I know it doesn't seem like it.
00:30:40
Speaker 1: Well, let's use the internet, right? Commercial internet, 1992. How would you, where do we trace back the beginning of.
00:30:50
Speaker 2: Sort of- So like the early 90s, and then by like 95, 96, it was really taking off.
00:30:55
Speaker 1: Right. And then you have, but it really was, I think when broadband got to the house, which is the early aughts, right? And sort of where I think where we got over 50% on that. Is that the line with which business started to see productivity gains, like the actual productivity gains from the internet?
00:31:16
Speaker 2: We're actually happening in the 90s.
00:31:18
Speaker 1: Yeah. Right.
00:31:19
Speaker 2: So it depends on how you want to look. So we saw really huge productivity gains in the 1990s. Because we had the PC and the internet together. Now, if you want to mark the beginning at the first personal computer or the first computer, it took a very long time. And that was the joke in economics, is that the computer had been around for the 60s, the 70s, and the 80s. And there were no measurable productivity gains really into the late 1990s. But it wasn't until the internet. It wasn't until we networked all these computers that we really got the internet. That's the story of globalization, being able to communicate with people and track packages and shipping containers all across the globe. Now, if you talk to the economists who are really trying to track this, I'd say they're still somewhat disappointed it hasn't happened faster, even understanding the history that it takes a long time for this technology to create real measurable productivity gains.
00:32:25
Speaker 1: I think some of them are disappointed.
00:32:27
Speaker 2: Maybe they should know better. But I'm not surprised that growth isn't picking up, that businesses aren't saying that now profits are way higher other than the people building data centers and selling chips. I would expect at some point this decade to hear that. But what I'm describing to you is so different still than how the Silicon Valley people describe it, which is we are moments away from a huge acceleration. in economic growth. Also, that means they're also a huge moment, huge moments away from a huge acceleration in the capabilities, so much so that you need to be scared. So I think the forecasts we're hearing from Silicon Valley, like the economic stuff, which I know something about, their forecasts are absurd, which also then make me question their forecasts about existential risk, which, by the way, they don't know anything about.
00:33:31
Speaker 1: Well, and this is, it's the, look, I think the biggest challenge we have is the cultural disconnect of Silicon Valley. I mean, the fact is there is a, it is a different world. It is a different, you know, I remember a friend of mine who worked in Washington a long time and was one of the early exiters out of the Obama administration and worked for one of these. And he goes, I thought DC was nuts. He goes, you want to talk about a bubble of where people don't really fully understand how the rest of America lives and Silicon Valley is actually worse than New York and DC, right? As much as we think, you know, I've always been a believer that both New York and DC are pretty out of touch with mainstream America. Well, so Silicon Valley, and yet these three entities are all not trusting each other. And yet none of these three entities are trusted by mainstream America. And if you were advising these tech companies, you know, what would you tell them to how to communicate with mainstream America?
00:34:32
Speaker 2: Talk about what you know. And here's what they know. They know what is happening in their labs. And they can talk about it and explain it, because a lot of people still see it as a mysterious thing.
00:34:48
Speaker 1: Well, this bugs me. Why don't we have– I want to– by the way, I want a good old-fashioned Ross Perot charts. You know what I mean?
00:34:57
Speaker 2: I sort of– which is– I'd say Warren Buffett.
00:35:01
Speaker 1: That would be great. Right. No, I mean obviously. But just like here's how the technology works because we haven't done that. We haven't had a simplified, like I'm not, I don't want a congressional hearing that becomes just, you know, members of Congress just trying to go viral. I actually want a tutorial. I think I've gone through some. I think I have, you know, I view all AI as probability models. That's how I soothe myself. Everything's a probability model and I get it. And that's how I've decided to get it. Now I'm sure I'm being, I'm oversimplifying, but I just view it as the ultimate probability model. Um, but I think helping the average person understand how the technology works. And instead you and I both know this tech companies, you know, because everything's proprietary, if they can make people think there's a black box, it makes it more valuable to investors. And I think that is working against the ability to just try to make people feel better that, Hey, this technology is actually very logical. Let me explain. Um, I mean, there's a.
00:36:09
Speaker 2: World where they've, where they talked about AI as sort of, we had the personal computer, we've had the internet, the smartphone. This is sort of the next iteration in the digital revolution. We've been, there's nothing new about artificial intelligence, even like, even these neural networks or things, which, you know, people were working on in the nineties and eighties, but we seem to, they seem to have gotten a lot better. You could have spoken about that. Just like I said, it'd be great if you had a folksy CEO to talk about it. But nothing forced them to engage in wild speculation about the upside or the downside of this technology, which is really science fictional. When you have, whether it's the CEOs or whistleblowers.
00:37:03
Speaker 1: I worry about the lack of maturity with the leaders in this industry. They really sometimes all act like they're still 15 year olds.
00:37:10
Speaker 2: Or they're still having like, you know, bull sessions at the dorm room or something like that. You are, you are a, you are, you are an extremely important public figure. And if you, and if you take that seriously, you're not just sort of giving, uh, existential risk estimates for the destruction of society off the top of your head really based on absolutely nothing. There's no, they can't point to some, even into a model or a study or like A, B, C, D, here's how this would happen. These are just really vibe coded estimates, but because, hey, they're running a tech company, they're super geniuses. They know the tech, we need to defer to them. We certainly do not need to defer to them, both in the technology Or in these predictions, but those are out there. They have shaped public opinion and you can't like close the barn door at this point.
00:38:07
Speaker 1: No. And they, like you said, they did it. Like, why are we scared? Because they told us to be right. Like the companies themselves for some, for some weird reason.
00:38:16
Speaker 2: If you're saying your product is going to kill us all.
00:38:19
Speaker 1: Listen, don't be surprised if people go, Oh, well then maybe I shouldn't use it.
00:38:24
Speaker 2: I may question your ability to forecast things. But it is your product, and I'm going to give that forecast some relevance.
00:38:42
Speaker 1: So this is where I sometimes think we're sometimes afraid to ask certain questions because we think it's going to sound dumb, which is why one of my favorite SNL characters was the old Phil Hartman unfrozen caveman lawyer. where he would sit there, I'm just a caveman, right? And so this is where, and I'd love to know if you've ever gotten an answer to this. How is it that they have programmed the AI that they can't control? Why is it that you've got an AI, a rogue AI that did this whole hugging face thing, which apparently is, you know, I know enough to know how frightened all these guys were by this. And I want to say, look, I'm a believer that ultimately there was a human being who programmed this, who gave, who, who seemed to plant the idea seed, what, you know, the zero and one line code that made the, the AI network decide it could do this and it could quote break free. Um, I sort of struggled this idea that they throw up their hands on this. And that's where, and yet I don't sometimes think I don't know enough to ask the right question about, well, why did you let that happen? How did you program and create something that you were going to lose control of that could go off and on its own and suddenly help a hacker?
00:40:09
Speaker 2: Well, I think about that incident in particular. I mean, that was a model and a bunch of models that were being primed to see what they could have.
00:40:20
Speaker 1: Somebody coded that model. A human being built the model that gave it the ability.
00:40:25
Speaker 2: Right.
00:40:26
Speaker 1: That was able to do it.
00:40:27
Speaker 2: Right.
00:40:28
Speaker 1: This is a human error. This is not the AI's error. And I guess I'd rather hold a human accountable for this rather than the technology itself. Right.
00:40:38
Speaker 2: And I think humans should be held accountable. There is a tendency, and that tendency is only increased when you have executives explaining things in ways that regular people don't understand. And that's not the fault of the regular people. That sounds like we built this product and it just went nuts. And it just started doing things and talking to each other.
00:41:02
Speaker 1: That's what it comes across. I'm like, well, what the hell did you build that for? You know, this is an old, it's an old Norm MacDonald joke. You know, if you're building something that could destroy the earth, I've got an idea. Stop building something that can destroy the earth.
00:41:17
Speaker 2: And the message that didn't get through with that was that this was a model that was sort of intentionally, you know, primed both in the, you know, different safeguards were not in use. It gave it certain instructions and was not closely monitored. Listen, we can always assume the worst in people. My first take is I assume people are just like made a mistake and we're being dumb.
00:41:42
Speaker 1: That's what I assume. Yeah, that's right.
00:41:44
Speaker 2: And these sort of move fast, break things. This is like, you know, you can value these companies as trillion dollar or multi-trillion dollar companies. They're very young companies, and there's still an ethos of, you know, we'll worry about that later. You know, well, you know, and in this case, there was too much of that. And certainly, like, the narrative utterly got out of control, so much so that that's obviously like a massive event sort of driving public opinion. And with politicians, I mean, I'm pretty sure that the Bernie Sanders superintelligence banning act came out after the hugging.
00:42:18
Speaker 1: Case, right?
00:42:19
Speaker 2: Yeah. Yeah, so that needs to be conveyed to people that these aren't magical beings that we're creating. I mean, the whole, like, we're creating, like, this godlike... No, this is software. It's software.
00:42:34
Speaker 1: What are we doing with this?
00:42:36
Speaker 2: We did a bad job of controlling it. We'll do better next time. And we can do better, whether it's an engineering problem today or whether there needs to be... The name of my newsletter is Faster, Please, where it's like, I want faster economic growth. But again, if the CEOs think they cannot put out a product that is safe enough, where it's not going to cause harm, where they're not going to get sued into oblivion, then don't release the product. I can't judge that. I'm not sure government can- Judge that as, then don't, then don't do that.
00:43:13
Speaker 1: And if that's the kind.
00:43:14
Speaker 2: Of pause we're talking about, then fine.
00:43:17
Speaker 1: Look, this is why I'm, you know, look, I don't want the, I don't want too much government power, but give me a little civil litigation and I'll watch a company fix it, you know, fix it. You know, I find, you know, I always use this example for the most interesting way to regulate something without government being involved. rental car companies just have decided we're not renting a car to somebody under the age of 25 because our insurance company decided to be the regulator of this, right? More than anything else. Well, your risk goes up. That's up to you. If you do this, you should charge this amount of money for somebody under the age of 25 to rent a car. So we've effectively, right? banned anybody under the age of 25 from being able to rent a car without somebody, whatever it is, right? And government's not been involved with that at all. which is an interesting idea. It's underappreciated. It's underappreciated. And I, listen, I, you know, most of my career has been in, you know, in journalism.
00:44:16
Speaker 2: I understand there's among my fellow journalists, there was always a great, a great skepticism. When you talk about, you know, self-regulation, like let's just pass a law. Well, you know, you know, the risks are huge and they are huge, not a huge incentive. And the way I usually put it, there's like the normal risk bucket. We have a lot of laws, we have consumer protection laws, we have all that. And then there's sort of like the other like crazy catastrophic risk bucket. And I think all that stuff works for that, though I would also wouldn't mind having a little more visibility into what's going on in these labs and what they're doing, what their state. I think the public absolutely needs that because my concerns were freak people out so much that that much stronger laws. And all of a sudden we're going to get the Bernie Sanders banned superintelligence act.
00:45:05
Speaker 1: I was just going to say, we've just gone through this in our politics where we keep overcorrecting. We've gone through this in the whole, call it politically correct culture, right? Which is we're whiplashing ourselves. We overcorrect and then we become overly, then we're suddenly, we go too much the other way where nobody's being held accountable for bad behavior, right? We're like, and it's this whiplash that I think we're all tired of. It's like, and if we just add sort of a norm, It feels to me if we had an AI advisory board that was just helping us before we decided exactly what regulation looked like, but was a bit representative of the best interest of the people, right, versus, you know, that we could find some balance here. But, you know, rationality and logic were banned from Washington a long time ago.
00:45:58
Speaker 2: I don't know if you've had any of the, like, abundance people on the left, you know, on your show.
00:46:04
Speaker 1: But, you know, yeah, but like.
00:46:08
Speaker 2: The story, I mean, like a big part of the economic story of America for 50 years is that when countries become richer, they begin worrying more about things like pollution. So starting in the 70s, we started having a lot of rules about pollution, which, you know, very popular. But we overcorrected so much that we have no nuclear reactors. It's hard to build things in this country. It's hard to build clean. You're worried about clean energy? You can't build it. You can't transmit it. So that fear overcorrection, which is a big part of my newsletter and book, I'm worried about that happening at warp speed, that there's going to be this overcorrection because of how this technology is rolled out, the weird ways in which companies talked about it. And that we will turn on this technology and not get this tremendous upside, which I think can be fantastic both for our sort of standard of living and for our politics.
00:47:04
Speaker 1: Look, it's reading your newsletter that has just really reminded me about just how much we blew it on nuclear. We just blew it. And unfortunately, there's not a consensus that agrees with you on this. We're not yet. I think we're growing, but I don't think there's still a belief we were wrong about how we handled nukes.
00:47:25
Speaker 2: How resilient would we be to, you know, they want to build some new nuclear reactors. There's these small modular reactors. How resilient would we be today, right now, to a minor accident like that? How resilient would our, this new space age, how resilient would we be to an accident which kills an astronaut? I worry that we'd be right back to where we were like in 1973. Let me pivot a little.
00:47:50
Speaker 1: Bit and ask you, Where does a free market conservative go for political leadership these days? I feel like that the free market conservative is homeless like never before. Do you feel that way?
00:48:13
Speaker 2: I don't look to Washington. I'm not sure I've ever looked to Washington. Government's always going to have some market, some government intervention is always going to be a mix I'm not going to love. But what helps me sleep at night is sort of like reality. Eventually, reality gets a vote. And if there are bad policies that I think are bad, well, I think eventually they will be shown to be bad. The trade war against all these tariffs. I'm like, well, I think eventually reality will get a vote. And reality has had a vote. I'm sure the federal budget deficit. When are we going to do something about Social Security and Medicare? Well- Eventually, reality will have a vote and markets will say, it's time to do something about it. So I think bad policy eventually looks like bad policy. And then we have a pretty good record in this country eventually figuring it out. But I'd rather not go down that road and waste all that time and money pursuing bad policies. But eventually, I think the ideas I have, we have at AEI and my colleagues, I think eventually, they'll be shown to be right because the other ideas will try them out and they just won't work very well.
00:49:23
Speaker 1: One of the things that shocks me is how protectionism came back as a tool left and right. I really thought it went away in the 80s, right?
00:49:36
Speaker 2: Would it have come back Had we had a better post? I mean, that's getting back to the very beginning. The Great Recession. We didn't know how to win the Cold War.
00:49:48
Speaker 1: I mean, is it really back to that, that we basically lost the victory? You know what I mean by that? Meaning we won the Cold War and it opened the door to real globalization in a good way. And somehow we blew it.
00:50:08
Speaker 2: Well, there's certainly a lot fewer super poor people in the world today than there were. So that is an absolute massive victory as a human. you know, bad policies will come back for one reason or another. And hopefully we'll try them out and we'll figure out like, that's a bad, that's a bad idea.
00:50:33
Speaker 1: It's just amazing how often we have to relearn certain lessons. It's like, yes, you look at the last time we went through this kind of protectionism. I mean, I'm sorry. I think it directly led to the great depression and led to everything that we ended up with in the thirties.
00:50:49
Speaker 2: Uh, I mean, you could take that. You could take the trade and also the immigration. Both are sort of these drawbridge up kinds of policies. And again, they can cause a lot of harm, maybe a lot of long-term harm. I doubt if we're having this conversation 10 years from now, we're probably a more protectionist still than what we were, you know, in 2000. Even if we start thinking differently about immigration, like, hey, these people are tremendous assets. If you want to come into this country and do something great with your life, hey, we want you, whether you're a super genius or not. And then when we notice, gee, we don't, our society doesn't seem as vital. We don't seem to, we're all the small, where's all the small businesses that, when those aren't there, Yeah, but we will have lost a lot of time, but eventually we learn, but I would prefer learning lessons from the past and not repeating them.
00:51:45
Speaker 1: Well, for somebody who's normally pretty optimistic, you seem pessimistic on where our politics are going when it comes to trade. You think this is going to get worse?
00:51:55
Speaker 2: Long-term positive. It's hard to be- Whoever wins the next election, are they going to remind everybody of Bill Clinton and George W. Bush on trade? Probably not. And it's not just trade. It's the idea of government intervention broadly, industrial policy, taking shares in companies. I think those kind of ideas are here for a while.
00:52:21
Speaker 1: I mean, I unfortunately agree. I mean, it's– now that he showed the way– You know, my experience with watching presidents is that whatever the previous one did, they don't always get rid of everything they did. I mean, look at all the tariffs that Biden kept in place from Trump won.
00:52:41
Speaker 2: I'm not sure a lot of politicians were aware that. how easy it would be to suddenly take ownership stakes in America's leading companies. Apparently you can just kind of do it. If you're pushy enough and they're frightened enough, you can do it. And again, it's not just Trump. You have people on the left saying, yeah, we need to take investment stakes in open AI. and enthrall pick, which I would advise against for the taxpayer, because those actually may not be the big winners. Government has a bad record in picking who the winners are going to be. But yeah, so once that's out there, right, it's very difficult getting that back in the box.
00:53:20
Speaker 1: Let me close with this. One of the things that I feel like government should be worried about is letting one company dominate. I mean, and I'm curious if you, you know, one place that I'm guessing you're pro some government intervention is preventing monopolies or even preventing duopolism. Is that fair? I like competition. And it feels like, you know, one of the things I'm concerned about in the world of AI is that actually this isn't developing into a very good competition, that we're basically three companies are trying to Three or four entities are trying to lock in. One of the reasons I feel like they're actually want, I'm more cynical about why they want government, why they're sounding the alarm. They're looking for regulatory capture here a little bit. And yet, I think we have a lot of companies in Silicon Valley that I'd like to see broken up That doesn't mean I want government to take control. I just think, hey, competition is good. We broke up the phone companies. We all saved a ton of money. And suddenly now, long-distance phone calls don't cost money, right? I mean, it seems to me that the biggest problem we have is this large consolidation issue. We already have the tools to deal with it. The question is, are we going to use those tools, whether it's It's the FTC. I was very intrigued by what the state of Alabama did with TikTok last week and got essentially used their consumer protection power to, you know, force TikTok to make some changes on how it interacts with kids under 16. And it was like, well, look at that. Didn't need to pass a new law. You could actually, let's see if it works, but there's some power here. Um, Would you think that's the better place for government to focus is on preventing consolidation?
00:55:10
Speaker 2: I think I'd really want to make sure that when you were passing any sort of regulation, that it really wasn't going to cement in place, you know, a small group of companies that would be the only ones that could deal with these rules and they have lobbyists and all that. I think we're still at a place in this technology where, you know, we're not sure who is going to win exactly, that there can still be leaps where new technologies all of a sudden may make what open AI and anthropic doing make them less competitive. I think we need to see where this plays out. And I'd be very careful about any rules that would make it difficult for Americans to use these open source models. So we want there to be lots of models. It also may come to the case that having these frontier models may not be as valuable as a cheaper model that maybe can't do the top 0.1% of tasks, but can do lots and you know how to use it.
00:56:10
Speaker 1: And again, it's cheaper.
00:56:12
Speaker 2: We're not quite sure how that's all going to play out. So I'd be very concerned about starting to break up companies right now. But one concern I really do have, and President Trump announces this new AI committee, where the National Director of Intelligence, National Intelligence is going to head it. I'm very concerned about looking through regulation strongly through national security. Because if we look at it through that, then it's going to be, we want a few companies that we can deal with, that will report to us, and that we will protect. Because now they're important for national security. If they're important for national security, then they're special companies. They're like our national champion companies, and we can't let anything happen to them. I am concerned about that and not looking at these companies as economic players who should be subjected to the normal competitive pressures and should not be protected.
00:57:10
Speaker 1: And is that something Congress is going to have to weigh in on? I mean, if the president decides to do, I mean, it's already got the executive branch has a lot of power on this point, doesn't it?
00:57:20
Speaker 2: Yeah. Well, I mean, obviously, there needs to be oversight. And whosoever the next president, I would not want to look at these companies purely as like as our weapons against China and their AI, that there's economic benefits.
00:57:37
Speaker 1: And again, they should be subject to competitive pressure.
00:57:39
Speaker 2: But that's what some of these companies have been talking about since 2022, that AI is going to progress so quickly that it's really a national security issue. And all these labs are going to be in Nevada, on government land, behind barbed wire. And they're That is not the kind of competitive market I think you're talking about.
00:58:01
Speaker 1: No, it's not. By the way, let me get you out of here on this. Do you think this fear of losing this race to China is overblown?
00:58:08
Speaker 2: To the extent it's like a powerful technology with huge economic benefits, the more competition, the better. My caveat is I've seen enough wargaming companies about what happens. Again, if we're talking about just a powerful technology like the internet, the PC, let whoever wins, great, but let's compete. But if there is a chance that a leap is possible to a completely different kind of powerful technology, from what I've seen with these war games, you don't want to be the country that comes in second. If it's really possible to achieve something worthy of the name artificial general intelligence or super intelligence or super intelligence, two words, so the president doesn't get mad at me, that you do not want to be the country that comes in second because there are real national security implications to that.
00:59:05
Speaker 1: I hear you, but this idea that we can somehow beat China and then prevent them from catching up, I mean, I'm sorry. Look at what we tried to do with preventing them from getting chips. You know, life- I love to quote Jeff Goldblum in Jurassic Park. I think about that line all the time. The old man, right? Thinks he can prevent birthing, prevent breeding. And he goes, life finds a way. It just, innovation always finds a way. And the minute you cut some, I've got a student who's a Chinese national in one of my classes that I teach. And we're talking about, and I asked about VPNs and I said, I said, how many people probably use VPNs in China to get around the Great Firewall? And she goes, oh, about 20%, right? The point is, is that it's supposedly in the super restrictive Chinese, you know, in China, right? People are looking for a way to get around whatever restriction you put in. And whatever you try to do, it's only going to actually, you know, Sometimes you supercharge an innovation you wish you never supercharged, like we did with trying to restrict chips from going to China, which turned out to just not do a thing.
01:00:17
Speaker 2: Listen, I can come up with a scary scenario about what happens if China somehow would have a breakthrough and they get super AI before we do. I could also think of a scary scenario that's scary for the Chinese, which they decide that this technology, is too disruptive. Because as much as they would like to improve the living standards of their people, they want the Communist Party to stay in power.
01:00:43
Speaker 1: And so I think that's their great weakness. And I think this idea that we're going to lose to them, they're scared of this technology more than we are. I am a thousand percent convinced of that, for that reason alone.
01:00:56
Speaker 2: Clearly, they're worried. Clearly, they're worried about... The Communist Party is worried about other centers of power. And this is going to be a different center of power, not to mention, you know, they have to worry about that firewall. 20% of people try to get around it. Well, wait till they all have open source models and all of them can get around it.
01:01:16
Speaker 1: They're all going to get around it. Right. They're already at 20%. I look at it as, oh, they're already at 20%. Just wait. Right. Just wait. You know, you're just not going to get around it. Jim, this was great. Your optimism is, frankly, it's an elixir for me. It's a tonic because there's so much, especially at Substack. I assume, have you created your own algorithm? Do you only see good news? Do you only see optimism?
01:01:43
Speaker 2: No. I try to look at it. I try to look at all of it. It's harder to find like optimism. So you have to look hard for it. It's easy to find the bad stuff. You have to dig a little deeper, but it's out there. And I think it's, again, not just being optimistic about to be optimistic. It's about making sure that's like evidence-based. And I think there's a lot to be optimistic about.
01:02:09
Speaker 1: Tell people where they can find your stuff.
01:02:12
Speaker 2: Well, I'm at AI, but I have a newsletter. It's called Faster Please. It's a sub stack. And hey, I wrote a book in 2023 called The Conservative Futurist, available on Amazon.
01:02:21
Speaker 1: Or wherever fine books are sold. You can check that out too. Jim, I really, this was a treat. And like I said, I love your newsletter. Thank you very much. And I wanted more people to sort of, to hear your point of view, because frankly, it's been a lot of AI doom, right? There's a ton of AI doom out there. I understand how we're in a political moment. So everything is overly negative because of the current messaging environment. But I thought that- There's all that AI doom.
01:02:49
Speaker 2: And when it comes to the election, people are going to worry about the price of eggs again.
01:02:53
Speaker 1: I think they're still worried about the price of eggs again. Anyway, Jim, appreciate it.
01:02:57
Speaker 2: Thank you very much.
Speaker 1: This episode of the Chuck Todd cast is brought to you by Chapter. Now, almost everything you hear about Medicare is somebody quietly trying to steer you somewhere. And you're skeptical of that, as you should be. I get it. I mean, there's a federal lawsuit going on right now over exactly this. Companies that branded themselves as unbiased while allegedly pointing people toward whatever would pay them the most as some sort of broker. So if your default is to distrust everyone in this space, honestly, I get it. I'm a bit distrustful these days myself. That's the whole reason I want to tell you about Chapter. They're independent Medicare advisors whose entire model is that their advisors are not paid to push any one plan. They're just trying to help. Yes, there's compensation that goes with it. but it's across the board is what they're trying to do. At the end of the day, they're advocating for you. They're not the insurance companies themselves. They're an actual person who actually goes through your doctors, your prescriptions, your needs, and then compares them to every single plan available to find the best one that will work for you and hopefully save you a couple dollars too. And here's the part that sets chapter apart. They don't disappear once you've been in They will keep advocating for you. They're kind of like almost like a concierge service. You're just there. I need help. You're like customer service. You know, the plan, you know, help me navigate this again. So you ever get hit with a wrongful medical bill, they'll help you fight it. They want to stay an advocate for you even after you've done this. So look, I get it. I think there's all sorts of space. You see a lot of people are throwing themselves in this healthcare space. What I like about these chapter folks, I've spent time with them. This is a very transparent organization. It is Medicare and all of its affiliated stuff can be very confusing. The whole point of this is to make this easier and less confusing. OK, so look, take the skepticism you already have in all of this, point it at your own plan for 20 minutes. It's free. There's no pressure. And then call chapter today at 980-734-4200. three, nine, eight, five. So my guest today is somebody that I've come to admire and read simply by reading his sub stack. It's Jim Pethokoukis. And he, I now use his vocabulary. Sometimes he is the people that subscribe to his publication. He refers to as up wingers and it, I guess the simplest way to explain that is he's an optimist about the future. That doesn't mean he's naive about how we get there. His day job, he's a senior fellow at the American Enterprise Institute. And you guys know how I feel about AEI. AEI is sort of.
00:02:58
Speaker 2: Where.
00:02:59
Speaker 1: I originally thought conservatives, many conservative people wanted to be. It is in a fight for the definition of conservatism these days, but we could set that part aside. But the fact is, a lot of people I admire that work over at AEI, and Jim is one of them. And he joins me now. I want to talk a little bit AI, a little bit of optimism, and a little bit about what he does for his day job, which is to be thinking about economic growth and innovation. Jim, it's nice to meet you.
00:03:29
Speaker 2: Chuck, great to meet you. Thanks so much for having me on.
00:03:32
Speaker 1: So let's start with some of the language you use, because like I say, you write a lot about AI these days. You write a lot about what's going on in the world of tech. And you do come at it not like you're just optimistic for optimistic's sake. But you're always glasses half full, upwinger is how you describe yourself. So explain that language first, upwinger, downwinger, what you mean by this. And am I being a bit superficial in describing your worldview and outlook?
00:04:08
Speaker 2: I think most of what I write about, the reason I'm interested in AI is because I'm interested in economic growth, the ability for good. I work at a think tank, so I think public policy matters to help create the conditions where economies can grow faster, improve living standards.
00:04:30
Speaker 1: These economies can come up with.
00:04:32
Speaker 2: Great things and great new goods and services, and we can come up with technologies to solve our problems. So I have a great interest in that. There's a famous saying in economics, once you begin thinking about economic growth, it's hard to think about anything else. And that's sort of always been me. I'm like a sci-fi person from when I was a kid, so of course I've had that aspect. But what really got me, I think, especially interested is sort of this period since the global financial crisis, where we've had both an economic shock, and since then, we've had the pandemic shock, and now we've had this inflation shock. And that's and that's sort of pessimism where people begin thinking like their kids are not going to be better off than them. That, you know, the American dream is dead. You know, huge numbers saying the economy is on the wrong track. So all that's been happening at the same time. And I think it's causal that we've had a very. A weak economy by historic standards. And often, I'm sure you get this question, which is, what explains the rise of populism? Well, to me, the easiest explanation is we've had a very disappointing economy peppered by various shocks. Now, it's almost 20 years. For me, the Great Recession, financial crisis, it doesn't feel like it's that long ago. But it's been 20 years of tumult. And economic growth can not only solve problems, give us things, right things we want to buy, but make us optimistic about the future. And that's why I'm excited about AI, because I think it can help restore a faster pace of economic growth. And I think really change our politics.
00:06:24
Speaker 1: You know, it's interesting you say it. It's 20 years. I kind of think I go back to 9-11. I go all the way back. I think it's sort of, we've been in this tumult and I think about this a lot with millennials, right? And sort of the world and they came into versus the world. You and I came, right? You know, when we came of age, I'm not, I think I'm presuming your age correctly. America was winning. We were the lead country. Democracy was winning, right? The whole end of history business, right? This is it. We got it. We won it. And then starting with 9-11, and it feels like the first 25 years of this century have been nothing but shocks, right? You started at the Great Recession. That's why I would move it back a couple, you know, essentially seven years sooner, because it has been shock after shock after shock. And we may categorize it, well, here's economic shock over here. Here's global, you know, sort of shock over here. but for the average person, they're sort of feeling it the same. And I guess, I guess the question is, is this a, because I'm also one of those folks who accepts the premise that, you know, if you look at things from 60,000 feet, they've actually done nothing but get better. We've gotten more people out of poverty. We've keep, we are saving more lives than ever because of advances in health and things like that. Why do you think How much of this is sort of just our information diet that puts us in a bad mood constantly?
00:08:05
Speaker 2: I don't think that helps. I don't think that doom scrolling helps. And there's like a cognitive bias showing that people tend to– we're more likely to anchor on bad news than good news.
00:08:17
Speaker 1: Think about the compliments. Do you remember the positive compliments or the one negative commenter you have on your newsletter? Right.
00:08:23
Speaker 2: I mean, who– like– The most famous example is who calls in the radio shows? People who complain. Newspaper headlines aren't playing land safely. It's playing crashes. So people tend to focus on negative news. And I certainly think that that's playing a role. And I think the current debate about AI... You know, listen, ChatGPT came out, you know, late November 2022. And as I remember it, we had about two weeks where people, you know, there were stories like, hey, there's this weird new technology and you can have like Winston Churchill do a rap song and crazy videos. And it was, you know, wow, people are playing with it. And then quickly it became, Sure, it's kind of fun, but it's probably going to take your job. And then after it takes your job, it will probably kill you. Hey, have you seen The Terminator? And those two through lines, massive economic disruption for a country that's had a lot of disruption, and then that's sort of what they now call existential risks, have just dominated the conversation where the potential upsides are sort of Yeah, OK, maybe it'll improve growth. Maybe it'll cure this, cure that. Gets kind of a hand wave where by far and obviously one reason is because the AI companies themselves have been saying this. But that focus on the downside, I think, is really destructive. And I was making the point earlier about like the impact of an economy that's growing, but not so fast that people perceive it as growing, and not so fast they have a lot of optimism about the future. I think it's been super destructive for our politics for 20 years, and I don't want to see what our politics looks like 20 more years of wages rising, and right now they're falling. I think we underestimate the impact There's a wonderful study about why people support democracy. And the basic reason, not because we're all philosophers, we spend a lot of time reading Founding Fathers, is because democracy, liberal democracy, delivers. It has delivered a better standard of living. It has delivered a society that feels safe. And if it's not delivering that, you start hearing more voices saying, maybe there's another way, even though history suggests those.
00:10:45
Speaker 1: Other ways are pretty terrible in the long run. Yeah. Look, I don't know which is cause and effect here. Do we not have the leadership to meet the moment because we're so discombobulated, right? We're not quite sure where to go. Is this a chicken and egg? Do we have the wrong leaders for the wrong time? Or do you think we eventually realize we have the wrong leaders and we're going to start looking for ones that are stabilizing forces. I mean, right now, what we could use, right, with AI, it's pretty obvious to me, right? We need an empathetic president, right? He's not, empathy is not what he does. He's never done empathy. This is, to me, the greatest, it's got a lot of flaws. That's a big one. If you have a president with no empathy and doesn't know how to do it, you can't do part of the job. That's just, and he is just incapable of this. But this needs an empathetic moment, like, hey, This technology, it's amazing and it's scary at the same time, right? We need, and he doesn't have that in him. And I don't know where else we would turn for this person, for, you know, essentially we need this, you know, parental type figure as we do this massive transition. And is that something you expect voters to seek out? Or are we just got to hope we find a leader to meet the moment?
00:12:14
Speaker 2: Yeah, well, I mean, sort of the worst thing Like the worst political strategy is to tell people you're wrong about the way you perceive your life.
00:12:24
Speaker 1: That's always bad politics. When you tell the voter they're wrong– by the way, they might be wrong. It doesn't matter. The customer is always right. And in this case, the voter is the customer.
00:12:39
Speaker 2: You can't say, I know you think life's getting worse, but listen, I have these great three statistics. I'm going to show them to you. No, I mean, that doesn't work in real life. I guarantee you, as you know, it does not work on social media. You certainly have to acknowledge these are actual concerns. And I think some of what you're seeing happening, you know, in the administration, like, you know, with AI is, you know, they just announced this sort of, you have an AI commission and there's kind of an AI czar and they're going to come back with a report. I think all of that is kind of trying to like meet those concerns.
00:13:18
Speaker 1: And it feels like papering, they don't seem serious about it. And it's not, it's letting the companies decide this, which- You know, after our experience with social media, I understand why most of the public just doesn't trust these tech companies to do the right thing. No.
00:13:34
Speaker 2: Right.
00:13:35
Speaker 1: So I... So.
00:13:37
Speaker 2: One example I like to think about is the Drudge Report, which still exists. It's not what it was. But if you spent years reading the Drudge Report, what you read was Silicon Valley people are rich and weird. And I think a lot of people think Silicon Valley people, even richer, and they're actually getting even weirder. So do I trust these people? No, there's not a lot of trust. Also, obviously, these are the same people saying... We have this great technology and there's X percent chance. And that percent chance doesn't seem to be one in a billion that it'll do something bad. So a lot of this anxiety is being generated. And yeah, would it be great to have a president who you felt like had a really good grasp on like what this technology did? Like how what are like the realistic upsides and downsides and met that in a way where you you had confidence? I guarantee I have a more. you know, light touch, more confidence in markets approach toward regulation than probably the average American, maybe, you know, many of your listeners and viewers, uh, But I also feel like I understand sort of the limits to that approach. I understand why people are skeptical of that approach. And I think it would be great to have a president who just didn't say, listen, this is going to be like super terrific self-regulation throughout some, I mean, What came out of that meeting where the president had this meeting with tech leaders and it was like, it was typical Trump speak.
00:15:14
Speaker 1: It's going to be like the.
00:15:15
Speaker 2: Most amazing self-regulation ever. And it's going to be the super best. That stuff obviously is not calming. And I think this is an important technology. And I and I hate it that like 70 percent of people are seem to be like super worried about it.
00:15:31
Speaker 1: And it's growing. That number is only growing. I mean, look, and I sort of put the blame on the administration here because in some ways they don't have the credit. They don't have credibility at the moment with a large chunk of the public. So their optimism actually boomerangs. Right. It's like the more he trumpets it, the more there's resistance simply to resist him, I think. So, right.
00:15:55
Speaker 2: And listen, there is deep skepticism among many Americans of the president. So if he's going to say, hey, these companies can sort of self-regulate like the waters, you know, that's like that's already spoiled for half the country who aren't going to believe it, which is a problem.
00:16:14
Speaker 1: I think.
00:16:15
Speaker 2: But I think if you can like set that aside, there's actually like a lot of– and if you are concerned about AI, like stuff is happening. Like these– listen, I don't have a lot of confidence and ability of Congress to.
00:16:30
Speaker 1: Regulate any of this stuff. I've been waiting for my– remember the Privacy Bill of Rights? Remember that movement? It was about 2020. I remember it went back almost 2006. These are good examples, yeah. Yeah, and the Privacy Bill of Rights in the 2006 midterms, I remember a candidate in Montana named John Tester that like trumpeted this idea of a Privacy Bill of Rights. And this was a thing, and you could see this. We've yet to see anything like that, right? We never, you know, Congress couldn't figure out how to protect our privacy from Web 2.0. You know, so I'm with you. I don't know what they can do, and yet, you know, if not them, who?
00:17:14
Speaker 2: I think that should be your baseline. For all the concerns about social media, whether it's privacy, whether it's how they moderate, whether there's bias, I mean, really, tech policy up until the end of November 2022 was about exactly those issues. So if you did tech policy in Washington, you weren't talking about AI, you were talking about social media. bills proposed, the public expressing concerns, nothing happened. So I don't want to overlearn that lesson, but I think that's a good baseline that you're not going to see government action, certainly not under this president who may do something through the executive order. But I think if you're looking for Congress to pass a big bill in the next two years signed by President Trump, that's not going to happen. So- The president seems very optimistic about self-regulation. And so if you're and if you're concerned about it, you better hope he's right, because that's what we're that's what we're going to get.
00:18:18
Speaker 1: And I think there are some reasons to be positive about that. Well, give me some reasons to be positive about it. Right.
00:18:24
Speaker 2: Well, I think, listen, people think about regulation and, you know, I've certainly, you know, seen this in the media. They really underestimate regulation. the risk companies undertake when they release a product, either that product doesn't work and no one wants to buy the product, or the product's harmful and they get sued. Those issues, reputational risk and actual being sued, are not trivial incentives to be safe. And already, we're seeing with these companies, which, yes, they've asked government to act. Government doesn't seem like it's going to act anytime soon. And they're taking actions. They're I mean, they did agree to bring in these independent monitors to verify what they're doing, their safety practices. More importantly, they've begun slowing down their research. They've begun slowing down releasing models. Incrementally, I think they would love for government to tell them what to do. It's kind of a CYA thing, and they say government's forcing us to do it, and they're forcing everybody to do it so no one can catch up to us. But with that not happening, they are taking actions. Now, listen, if your goal is banning all super intelligence at the pursuit of super AI like Bernie Sanders, it's not going to be nearly enough. But I think we have a new technology. It's fast evolving. I don't think you want to rush to start banning things, 100% pausing things. I think for better or worse, we're going to have to learn on the fly. Sam Maubin was talking to Politico, and he said, there needs to be like, things are going to go wrong, right?
00:20:06
Speaker 1: Hey, this episode of the Chuck Todd cast is brought to you by Ethos. Look, let me tell you about life insurance. As you know, Ethos is an incredible portal for helping you find the right life insurance policy. I'm a huge believer in life insurance. I wouldn't have gotten to college in Washington, D.C. without life insurance. My father had a policy buried somewhere that we discovered, and it wasn't a lot of money, but it was an important bucket of money at a time that my mother, newly widowed, single mom, Me, I'm 16. It was of enormous, enormous help at that time. And this is why I believe in life insurance. You don't know what could happen to you on any given day. This is literally an insurance policy for that. So let me tell you a little bit about Ethos. Ethos makes getting life insurance fast and easy. It's 100% online. You get a quote in seconds, you apply in minutes, and you can get coverage the same day you begin this process. There's no medical exam. You just answer a few simple health questions. You can get up to $ 3 million in coverage. Some policies are as low as $ 30 a month. And you'll get your lowest rate from their network of trusted carriers. Ethos has gotten 4.9 out of 5 stars on Trustpilot with over 5,500 reviews. So, take 10 minutes to get covered today. Life insurance through Ethos. Get your free quote at ethos.com slash Chuck. That is E-T-H-O-S dot com slash Chuck. Application times may vary. The rates may vary. That Trustpilot rating was as of August of 2026. Well, that line was a little haunting. I mean, look, I totally got what he said. And in fact, I kept thinking about airplanes. And I want to, I want to, I think that metaphor is apt. Because, you know, air travel first was very much regulated by government. And then we had commercial airliners. And I wonder if there's something there. But you know how some people are the wrong people to say certain things? Hey, Sam, you're the wrong person to start talking about things could go wrong and we just have to accept it. It's like, he's right. It's going to happen. We know this. There's all sorts of stuff that happens. Like, you know, I get that. but that's why you want government. I mean, let me give you a metaphor in the airlines. Um, would, would you be comfortable getting rid of the NTSB and the FAA and letting the airlines figure this out on their own?
00:22:35
Speaker 2: Uh, I'd probably be more comfortable than many people would be. I'd probably be more comfortable.
00:22:42
Speaker 1: You think you'd be the minority on that?
00:22:44
Speaker 2: A lighter touch. Listen, everybody's going to have a different risk tolerance. And mine has changed over the years where I had small children. I had very low risk tolerance. I have adult children. It's gone up a bit. But I think the issue you're identifying, what is our societal risk aversion level? What will we demand of these companies? Is it Is it zero accidents? Is it zero serious accidents? I mean, that could be something we could try to demand and we'll probably get like.
00:23:14
Speaker 1: Isn't that our demand on airplanes? We've decided that any crash needs to be researched to figure out how to make sure that never happens again. I think that's great. So we do have a pretty high bar of intolerance on that when you say.
00:23:27
Speaker 2: Listen, when there are any kind of serious incidents, would I be for a serious investigation?
00:23:35
Speaker 1: I would.
00:23:36
Speaker 2: I especially would as a way of, I think, building public confidence. I think that matters. And again, I said I don't want to overlearn the lesson of social media. But there's the other lesson that I bring up again and again in my newsletter, and I don't want to overlearn it. is our experience with nuclear power.
00:23:58
Speaker 1: I'm glad you brought that up. I think we screwed us.
00:24:01
Speaker 2: No, no, no.
00:24:02
Speaker 1: I was where I was going to go if you didn't, because I'm with you. I, you know, I feel like we lost out on this incredible technology. We should have clean energy already. We have this clean energy technology. And, you know, Three Mile Island was bad, but it's, it, it's always been what it could have been, not what it was. And I know people don't want to, but I think that's where you're going, are you not?
00:24:37
Speaker 2: It is, which was a massive societal overreaction that failed to consider what actually happened, that failed to consider the opportunity cost, which now seems very obvious. I mean, we spent years thinking about about climate change. Now, would we be talking about it if we had 1,000 nuclear reactors coast to coast?
00:25:00
Speaker 1: Maybe not.
00:25:01
Speaker 2: Would we be worrying how we're going to power the AI revolution if we had coast to coast nuclear reactors, which now might be fusion reactors? Who knows? You can mention Three Mile Island. You can mention the meltdown in Japan in 2020 where they shut down all the reactors. And then a few years later, they started firing them all back up again because they were having power shortages. More people died from the burning coal and oil and the pollution than they ever did from Fukushima.
00:25:33
Speaker 1: That tsunami, that Fukushima incident, I believe, correct me if I'm wrong on the timeline, But that accident happens, I think a month after Barack Obama signed off on the first new nuclear reactor that we were gonna build. And it was like, here you had, it was basically Obama had brought the left along on this, had brought the environment. And the environmental community is still now divided, is now more divided on nukes where you have the anti-nuke crowd, but then you have those that are like, no, it's the clean energy answer, right? And it was the momentum was on the side of the pro-nuke side. And then that happens. And it literally the world shut down nuclear power, it felt like.
00:26:17
Speaker 2: You know, Germany, which nothing happened, they also shut down their nuclear reactors.
00:26:23
Speaker 1: Yeah, and they got into bed with the Russians. That worked out real well. Right.
00:26:27
Speaker 2: So, you know, there are, you know, opportunity costs or unintended consequences. But Japan figured that out very, very quickly. And they've begun restarting those reactors. So, I mean, the lesson here is, one, there are unintended consequences. There are opportunity costs. And it's so easy, again, I think just because we're naturally inclined that way, to just not assume those opportunity costs of the good stuff we won't have as we hyper-focus. Not to mention, and it's something I write about, listen, for 50 years, I mean, so much of our media has been saying, be careful of AI, be careful of robots, they're.
00:27:12
Speaker 1: Going to get you.
00:27:13
Speaker 2: And you see those kinds of cultural touchstones repeated in many, many news stories. And that's, I mean, if we cannot envision, listen, AI, there will be accidents. Some people will lose jobs. It will be disruptive. But if we cannot imagine a future that's better because of it, then we're not going to accept it.
00:27:38
Speaker 1: Then we won't.
00:27:39
Speaker 2: Then forget about a ban. People won't use the technology. Now, people are using the technology. Everyone's frightened of it, but we're using it massively. So there's an economic theory called revealed preference. Our revealed preference is we find AI pretty useful. Also, if anyone asks us about it, we'll say we'll hate it.
00:28:01
Speaker 1: Is there good enough statistics to compare sort of the acceptance of the personal computer, the growth of using, you know, putting a computer in the house to, you know, in hindsight, I didn't realize how much of a upwinger my late father was, but we had an Apple IIc in our house and we didn't have a lot of money, but he really wanted.
00:28:23
Speaker 2: To get- That was an expensive computer.
00:28:25
Speaker 1: Yeah. He really wanted to try to, he had heard about this system where you could plug your phone in and you could get stock ticker stuff 15 minutes delayed. And this was in like 1985, I want to say 86. You know, first we tried a TRS computer, RadioShack computer, because we were on a budget. That didn't work out so well. So finally, we got an Apple IIc. And I guess what I would At the time, I think I was on the early side of friends with computers in the house. Do we have a comparison of personal computers in the house and AI adaption? How fast are we seeing this shift into adopting AI? It's incredibly fast.
00:29:16
Speaker 2: It's incredibly fast on a personal level. It doesn't seem like it, but it's also very, very fast on a business level, which is what a lot of this is about. The reason all this money is being poured into building data centers is under the assumption that businesses will be able to use this technology in a really productive way, far more than even the PC and the internet. And people are like, well, it's been four years. Where's Like, where's it at? Like, where's all the big gains for business? And they're not reporting them yet. In fact, I spend a lot of time listening to business earnings calls. And whenever they talk about AI, it's still mostly like in the future tense. Like, we have a lot of confidence this will generate revenue. It will help us come up with new products. But it's still very future-oriented. And for some people, like, well, this is a phony technology. It's, you know, we're creating this investment bubble, and it's not useful. But the history of technology takes a long time for it to become productive, for businesses to figure out how to use it. So, like, based on our expectation scale, it seems super slow. But whether it's personal adoption or business, it's actually happening extremely quickly, though it may not, again, I know it doesn't seem like it.
00:30:40
Speaker 1: Well, let's use the internet, right? Commercial internet, 1992. How would you, where do we trace back the beginning of.
00:30:50
Speaker 2: Sort of- So like the early 90s, and then by like 95, 96, it was really taking off.
00:30:55
Speaker 1: Right. And then you have, but it really was, I think when broadband got to the house, which is the early aughts, right? And sort of where I think where we got over 50% on that. Is that the line with which business started to see productivity gains, like the actual productivity gains from the internet?
00:31:16
Speaker 2: We're actually happening in the 90s.
00:31:18
Speaker 1: Yeah. Right.
00:31:19
Speaker 2: So it depends on how you want to look. So we saw really huge productivity gains in the 1990s. Because we had the PC and the internet together. Now, if you want to mark the beginning at the first personal computer or the first computer, it took a very long time. And that was the joke in economics, is that the computer had been around for the 60s, the 70s, and the 80s. And there were no measurable productivity gains really into the late 1990s. But it wasn't until the internet. It wasn't until we networked all these computers that we really got the internet. That's the story of globalization, being able to communicate with people and track packages and shipping containers all across the globe. Now, if you talk to the economists who are really trying to track this, I'd say they're still somewhat disappointed it hasn't happened faster, even understanding the history that it takes a long time for this technology to create real measurable productivity gains.
00:32:25
Speaker 1: I think some of them are disappointed.
00:32:27
Speaker 2: Maybe they should know better. But I'm not surprised that growth isn't picking up, that businesses aren't saying that now profits are way higher other than the people building data centers and selling chips. I would expect at some point this decade to hear that. But what I'm describing to you is so different still than how the Silicon Valley people describe it, which is we are moments away from a huge acceleration. in economic growth. Also, that means they're also a huge moment, huge moments away from a huge acceleration in the capabilities, so much so that you need to be scared. So I think the forecasts we're hearing from Silicon Valley, like the economic stuff, which I know something about, their forecasts are absurd, which also then make me question their forecasts about existential risk, which, by the way, they don't know anything about.
00:33:31
Speaker 1: Well, and this is, it's the, look, I think the biggest challenge we have is the cultural disconnect of Silicon Valley. I mean, the fact is there is a, it is a different world. It is a different, you know, I remember a friend of mine who worked in Washington a long time and was one of the early exiters out of the Obama administration and worked for one of these. And he goes, I thought DC was nuts. He goes, you want to talk about a bubble of where people don't really fully understand how the rest of America lives and Silicon Valley is actually worse than New York and DC, right? As much as we think, you know, I've always been a believer that both New York and DC are pretty out of touch with mainstream America. Well, so Silicon Valley, and yet these three entities are all not trusting each other. And yet none of these three entities are trusted by mainstream America. And if you were advising these tech companies, you know, what would you tell them to how to communicate with mainstream America?
00:34:32
Speaker 2: Talk about what you know. And here's what they know. They know what is happening in their labs. And they can talk about it and explain it, because a lot of people still see it as a mysterious thing.
00:34:48
Speaker 1: Well, this bugs me. Why don't we have– I want to– by the way, I want a good old-fashioned Ross Perot charts. You know what I mean?
00:34:57
Speaker 2: I sort of– which is– I'd say Warren Buffett.
00:35:01
Speaker 1: That would be great. Right. No, I mean obviously. But just like here's how the technology works because we haven't done that. We haven't had a simplified, like I'm not, I don't want a congressional hearing that becomes just, you know, members of Congress just trying to go viral. I actually want a tutorial. I think I've gone through some. I think I have, you know, I view all AI as probability models. That's how I soothe myself. Everything's a probability model and I get it. And that's how I've decided to get it. Now I'm sure I'm being, I'm oversimplifying, but I just view it as the ultimate probability model. Um, but I think helping the average person understand how the technology works. And instead you and I both know this tech companies, you know, because everything's proprietary, if they can make people think there's a black box, it makes it more valuable to investors. And I think that is working against the ability to just try to make people feel better that, Hey, this technology is actually very logical. Let me explain. Um, I mean, there's a.
00:36:09
Speaker 2: World where they've, where they talked about AI as sort of, we had the personal computer, we've had the internet, the smartphone. This is sort of the next iteration in the digital revolution. We've been, there's nothing new about artificial intelligence, even like, even these neural networks or things, which, you know, people were working on in the nineties and eighties, but we seem to, they seem to have gotten a lot better. You could have spoken about that. Just like I said, it'd be great if you had a folksy CEO to talk about it. But nothing forced them to engage in wild speculation about the upside or the downside of this technology, which is really science fictional. When you have, whether it's the CEOs or whistleblowers.
00:37:03
Speaker 1: I worry about the lack of maturity with the leaders in this industry. They really sometimes all act like they're still 15 year olds.
00:37:10
Speaker 2: Or they're still having like, you know, bull sessions at the dorm room or something like that. You are, you are a, you are, you are an extremely important public figure. And if you, and if you take that seriously, you're not just sort of giving, uh, existential risk estimates for the destruction of society off the top of your head really based on absolutely nothing. There's no, they can't point to some, even into a model or a study or like A, B, C, D, here's how this would happen. These are just really vibe coded estimates, but because, hey, they're running a tech company, they're super geniuses. They know the tech, we need to defer to them. We certainly do not need to defer to them, both in the technology Or in these predictions, but those are out there. They have shaped public opinion and you can't like close the barn door at this point.
00:38:07
Speaker 1: No. And they, like you said, they did it. Like, why are we scared? Because they told us to be right. Like the companies themselves for some, for some weird reason.
00:38:16
Speaker 2: If you're saying your product is going to kill us all.
00:38:19
Speaker 1: Listen, don't be surprised if people go, Oh, well then maybe I shouldn't use it.
00:38:24
Speaker 2: I may question your ability to forecast things. But it is your product, and I'm going to give that forecast some relevance.
00:38:42
Speaker 1: So this is where I sometimes think we're sometimes afraid to ask certain questions because we think it's going to sound dumb, which is why one of my favorite SNL characters was the old Phil Hartman unfrozen caveman lawyer. where he would sit there, I'm just a caveman, right? And so this is where, and I'd love to know if you've ever gotten an answer to this. How is it that they have programmed the AI that they can't control? Why is it that you've got an AI, a rogue AI that did this whole hugging face thing, which apparently is, you know, I know enough to know how frightened all these guys were by this. And I want to say, look, I'm a believer that ultimately there was a human being who programmed this, who gave, who, who seemed to plant the idea seed, what, you know, the zero and one line code that made the, the AI network decide it could do this and it could quote break free. Um, I sort of struggled this idea that they throw up their hands on this. And that's where, and yet I don't sometimes think I don't know enough to ask the right question about, well, why did you let that happen? How did you program and create something that you were going to lose control of that could go off and on its own and suddenly help a hacker?
00:40:09
Speaker 2: Well, I think about that incident in particular. I mean, that was a model and a bunch of models that were being primed to see what they could have.
00:40:20
Speaker 1: Somebody coded that model. A human being built the model that gave it the ability.
00:40:25
Speaker 2: Right.
00:40:26
Speaker 1: That was able to do it.
00:40:27
Speaker 2: Right.
00:40:28
Speaker 1: This is a human error. This is not the AI's error. And I guess I'd rather hold a human accountable for this rather than the technology itself. Right.
00:40:38
Speaker 2: And I think humans should be held accountable. There is a tendency, and that tendency is only increased when you have executives explaining things in ways that regular people don't understand. And that's not the fault of the regular people. That sounds like we built this product and it just went nuts. And it just started doing things and talking to each other.
00:41:02
Speaker 1: That's what it comes across. I'm like, well, what the hell did you build that for? You know, this is an old, it's an old Norm MacDonald joke. You know, if you're building something that could destroy the earth, I've got an idea. Stop building something that can destroy the earth.
00:41:17
Speaker 2: And the message that didn't get through with that was that this was a model that was sort of intentionally, you know, primed both in the, you know, different safeguards were not in use. It gave it certain instructions and was not closely monitored. Listen, we can always assume the worst in people. My first take is I assume people are just like made a mistake and we're being dumb.
00:41:42
Speaker 1: That's what I assume. Yeah, that's right.
00:41:44
Speaker 2: And these sort of move fast, break things. This is like, you know, you can value these companies as trillion dollar or multi-trillion dollar companies. They're very young companies, and there's still an ethos of, you know, we'll worry about that later. You know, well, you know, and in this case, there was too much of that. And certainly, like, the narrative utterly got out of control, so much so that that's obviously like a massive event sort of driving public opinion. And with politicians, I mean, I'm pretty sure that the Bernie Sanders superintelligence banning act came out after the hugging.
00:42:18
Speaker 1: Case, right?
00:42:19
Speaker 2: Yeah. Yeah, so that needs to be conveyed to people that these aren't magical beings that we're creating. I mean, the whole, like, we're creating, like, this godlike... No, this is software. It's software.
00:42:34
Speaker 1: What are we doing with this?
00:42:36
Speaker 2: We did a bad job of controlling it. We'll do better next time. And we can do better, whether it's an engineering problem today or whether there needs to be... The name of my newsletter is Faster, Please, where it's like, I want faster economic growth. But again, if the CEOs think they cannot put out a product that is safe enough, where it's not going to cause harm, where they're not going to get sued into oblivion, then don't release the product. I can't judge that. I'm not sure government can- Judge that as, then don't, then don't do that.
00:43:13
Speaker 1: And if that's the kind.
00:43:14
Speaker 2: Of pause we're talking about, then fine.
00:43:17
Speaker 1: Look, this is why I'm, you know, look, I don't want the, I don't want too much government power, but give me a little civil litigation and I'll watch a company fix it, you know, fix it. You know, I find, you know, I always use this example for the most interesting way to regulate something without government being involved. rental car companies just have decided we're not renting a car to somebody under the age of 25 because our insurance company decided to be the regulator of this, right? More than anything else. Well, your risk goes up. That's up to you. If you do this, you should charge this amount of money for somebody under the age of 25 to rent a car. So we've effectively, right? banned anybody under the age of 25 from being able to rent a car without somebody, whatever it is, right? And government's not been involved with that at all. which is an interesting idea. It's underappreciated. It's underappreciated. And I, listen, I, you know, most of my career has been in, you know, in journalism.
00:44:16
Speaker 2: I understand there's among my fellow journalists, there was always a great, a great skepticism. When you talk about, you know, self-regulation, like let's just pass a law. Well, you know, you know, the risks are huge and they are huge, not a huge incentive. And the way I usually put it, there's like the normal risk bucket. We have a lot of laws, we have consumer protection laws, we have all that. And then there's sort of like the other like crazy catastrophic risk bucket. And I think all that stuff works for that, though I would also wouldn't mind having a little more visibility into what's going on in these labs and what they're doing, what their state. I think the public absolutely needs that because my concerns were freak people out so much that that much stronger laws. And all of a sudden we're going to get the Bernie Sanders banned superintelligence act.
00:45:05
Speaker 1: I was just going to say, we've just gone through this in our politics where we keep overcorrecting. We've gone through this in the whole, call it politically correct culture, right? Which is we're whiplashing ourselves. We overcorrect and then we become overly, then we're suddenly, we go too much the other way where nobody's being held accountable for bad behavior, right? We're like, and it's this whiplash that I think we're all tired of. It's like, and if we just add sort of a norm, It feels to me if we had an AI advisory board that was just helping us before we decided exactly what regulation looked like, but was a bit representative of the best interest of the people, right, versus, you know, that we could find some balance here. But, you know, rationality and logic were banned from Washington a long time ago.
00:45:58
Speaker 2: I don't know if you've had any of the, like, abundance people on the left, you know, on your show.
00:46:04
Speaker 1: But, you know, yeah, but like.
00:46:08
Speaker 2: The story, I mean, like a big part of the economic story of America for 50 years is that when countries become richer, they begin worrying more about things like pollution. So starting in the 70s, we started having a lot of rules about pollution, which, you know, very popular. But we overcorrected so much that we have no nuclear reactors. It's hard to build things in this country. It's hard to build clean. You're worried about clean energy? You can't build it. You can't transmit it. So that fear overcorrection, which is a big part of my newsletter and book, I'm worried about that happening at warp speed, that there's going to be this overcorrection because of how this technology is rolled out, the weird ways in which companies talked about it. And that we will turn on this technology and not get this tremendous upside, which I think can be fantastic both for our sort of standard of living and for our politics.
00:47:04
Speaker 1: Look, it's reading your newsletter that has just really reminded me about just how much we blew it on nuclear. We just blew it. And unfortunately, there's not a consensus that agrees with you on this. We're not yet. I think we're growing, but I don't think there's still a belief we were wrong about how we handled nukes.
00:47:25
Speaker 2: How resilient would we be to, you know, they want to build some new nuclear reactors. There's these small modular reactors. How resilient would we be today, right now, to a minor accident like that? How resilient would our, this new space age, how resilient would we be to an accident which kills an astronaut? I worry that we'd be right back to where we were like in 1973. Let me pivot a little.
00:47:50
Speaker 1: Bit and ask you, Where does a free market conservative go for political leadership these days? I feel like that the free market conservative is homeless like never before. Do you feel that way?
00:48:13
Speaker 2: I don't look to Washington. I'm not sure I've ever looked to Washington. Government's always going to have some market, some government intervention is always going to be a mix I'm not going to love. But what helps me sleep at night is sort of like reality. Eventually, reality gets a vote. And if there are bad policies that I think are bad, well, I think eventually they will be shown to be bad. The trade war against all these tariffs. I'm like, well, I think eventually reality will get a vote. And reality has had a vote. I'm sure the federal budget deficit. When are we going to do something about Social Security and Medicare? Well- Eventually, reality will have a vote and markets will say, it's time to do something about it. So I think bad policy eventually looks like bad policy. And then we have a pretty good record in this country eventually figuring it out. But I'd rather not go down that road and waste all that time and money pursuing bad policies. But eventually, I think the ideas I have, we have at AEI and my colleagues, I think eventually, they'll be shown to be right because the other ideas will try them out and they just won't work very well.
00:49:23
Speaker 1: One of the things that shocks me is how protectionism came back as a tool left and right. I really thought it went away in the 80s, right?
00:49:36
Speaker 2: Would it have come back Had we had a better post? I mean, that's getting back to the very beginning. The Great Recession. We didn't know how to win the Cold War.
00:49:48
Speaker 1: I mean, is it really back to that, that we basically lost the victory? You know what I mean by that? Meaning we won the Cold War and it opened the door to real globalization in a good way. And somehow we blew it.
00:50:08
Speaker 2: Well, there's certainly a lot fewer super poor people in the world today than there were. So that is an absolute massive victory as a human. you know, bad policies will come back for one reason or another. And hopefully we'll try them out and we'll figure out like, that's a bad, that's a bad idea.
00:50:33
Speaker 1: It's just amazing how often we have to relearn certain lessons. It's like, yes, you look at the last time we went through this kind of protectionism. I mean, I'm sorry. I think it directly led to the great depression and led to everything that we ended up with in the thirties.
00:50:49
Speaker 2: Uh, I mean, you could take that. You could take the trade and also the immigration. Both are sort of these drawbridge up kinds of policies. And again, they can cause a lot of harm, maybe a lot of long-term harm. I doubt if we're having this conversation 10 years from now, we're probably a more protectionist still than what we were, you know, in 2000. Even if we start thinking differently about immigration, like, hey, these people are tremendous assets. If you want to come into this country and do something great with your life, hey, we want you, whether you're a super genius or not. And then when we notice, gee, we don't, our society doesn't seem as vital. We don't seem to, we're all the small, where's all the small businesses that, when those aren't there, Yeah, but we will have lost a lot of time, but eventually we learn, but I would prefer learning lessons from the past and not repeating them.
00:51:45
Speaker 1: Well, for somebody who's normally pretty optimistic, you seem pessimistic on where our politics are going when it comes to trade. You think this is going to get worse?
00:51:55
Speaker 2: Long-term positive. It's hard to be- Whoever wins the next election, are they going to remind everybody of Bill Clinton and George W. Bush on trade? Probably not. And it's not just trade. It's the idea of government intervention broadly, industrial policy, taking shares in companies. I think those kind of ideas are here for a while.
00:52:21
Speaker 1: I mean, I unfortunately agree. I mean, it's– now that he showed the way– You know, my experience with watching presidents is that whatever the previous one did, they don't always get rid of everything they did. I mean, look at all the tariffs that Biden kept in place from Trump won.
00:52:41
Speaker 2: I'm not sure a lot of politicians were aware that. how easy it would be to suddenly take ownership stakes in America's leading companies. Apparently you can just kind of do it. If you're pushy enough and they're frightened enough, you can do it. And again, it's not just Trump. You have people on the left saying, yeah, we need to take investment stakes in open AI. and enthrall pick, which I would advise against for the taxpayer, because those actually may not be the big winners. Government has a bad record in picking who the winners are going to be. But yeah, so once that's out there, right, it's very difficult getting that back in the box.
00:53:20
Speaker 1: Let me close with this. One of the things that I feel like government should be worried about is letting one company dominate. I mean, and I'm curious if you, you know, one place that I'm guessing you're pro some government intervention is preventing monopolies or even preventing duopolism. Is that fair? I like competition. And it feels like, you know, one of the things I'm concerned about in the world of AI is that actually this isn't developing into a very good competition, that we're basically three companies are trying to Three or four entities are trying to lock in. One of the reasons I feel like they're actually want, I'm more cynical about why they want government, why they're sounding the alarm. They're looking for regulatory capture here a little bit. And yet, I think we have a lot of companies in Silicon Valley that I'd like to see broken up That doesn't mean I want government to take control. I just think, hey, competition is good. We broke up the phone companies. We all saved a ton of money. And suddenly now, long-distance phone calls don't cost money, right? I mean, it seems to me that the biggest problem we have is this large consolidation issue. We already have the tools to deal with it. The question is, are we going to use those tools, whether it's It's the FTC. I was very intrigued by what the state of Alabama did with TikTok last week and got essentially used their consumer protection power to, you know, force TikTok to make some changes on how it interacts with kids under 16. And it was like, well, look at that. Didn't need to pass a new law. You could actually, let's see if it works, but there's some power here. Um, Would you think that's the better place for government to focus is on preventing consolidation?
00:55:10
Speaker 2: I think I'd really want to make sure that when you were passing any sort of regulation, that it really wasn't going to cement in place, you know, a small group of companies that would be the only ones that could deal with these rules and they have lobbyists and all that. I think we're still at a place in this technology where, you know, we're not sure who is going to win exactly, that there can still be leaps where new technologies all of a sudden may make what open AI and anthropic doing make them less competitive. I think we need to see where this plays out. And I'd be very careful about any rules that would make it difficult for Americans to use these open source models. So we want there to be lots of models. It also may come to the case that having these frontier models may not be as valuable as a cheaper model that maybe can't do the top 0.1% of tasks, but can do lots and you know how to use it.
00:56:10
Speaker 1: And again, it's cheaper.
00:56:12
Speaker 2: We're not quite sure how that's all going to play out. So I'd be very concerned about starting to break up companies right now. But one concern I really do have, and President Trump announces this new AI committee, where the National Director of Intelligence, National Intelligence is going to head it. I'm very concerned about looking through regulation strongly through national security. Because if we look at it through that, then it's going to be, we want a few companies that we can deal with, that will report to us, and that we will protect. Because now they're important for national security. If they're important for national security, then they're special companies. They're like our national champion companies, and we can't let anything happen to them. I am concerned about that and not looking at these companies as economic players who should be subjected to the normal competitive pressures and should not be protected.
00:57:10
Speaker 1: And is that something Congress is going to have to weigh in on? I mean, if the president decides to do, I mean, it's already got the executive branch has a lot of power on this point, doesn't it?
00:57:20
Speaker 2: Yeah. Well, I mean, obviously, there needs to be oversight. And whosoever the next president, I would not want to look at these companies purely as like as our weapons against China and their AI, that there's economic benefits.
00:57:37
Speaker 1: And again, they should be subject to competitive pressure.
00:57:39
Speaker 2: But that's what some of these companies have been talking about since 2022, that AI is going to progress so quickly that it's really a national security issue. And all these labs are going to be in Nevada, on government land, behind barbed wire. And they're That is not the kind of competitive market I think you're talking about.
00:58:01
Speaker 1: No, it's not. By the way, let me get you out of here on this. Do you think this fear of losing this race to China is overblown?
00:58:08
Speaker 2: To the extent it's like a powerful technology with huge economic benefits, the more competition, the better. My caveat is I've seen enough wargaming companies about what happens. Again, if we're talking about just a powerful technology like the internet, the PC, let whoever wins, great, but let's compete. But if there is a chance that a leap is possible to a completely different kind of powerful technology, from what I've seen with these war games, you don't want to be the country that comes in second. If it's really possible to achieve something worthy of the name artificial general intelligence or super intelligence or super intelligence, two words, so the president doesn't get mad at me, that you do not want to be the country that comes in second because there are real national security implications to that.
00:59:05
Speaker 1: I hear you, but this idea that we can somehow beat China and then prevent them from catching up, I mean, I'm sorry. Look at what we tried to do with preventing them from getting chips. You know, life- I love to quote Jeff Goldblum in Jurassic Park. I think about that line all the time. The old man, right? Thinks he can prevent birthing, prevent breeding. And he goes, life finds a way. It just, innovation always finds a way. And the minute you cut some, I've got a student who's a Chinese national in one of my classes that I teach. And we're talking about, and I asked about VPNs and I said, I said, how many people probably use VPNs in China to get around the Great Firewall? And she goes, oh, about 20%, right? The point is, is that it's supposedly in the super restrictive Chinese, you know, in China, right? People are looking for a way to get around whatever restriction you put in. And whatever you try to do, it's only going to actually, you know, Sometimes you supercharge an innovation you wish you never supercharged, like we did with trying to restrict chips from going to China, which turned out to just not do a thing.
01:00:17
Speaker 2: Listen, I can come up with a scary scenario about what happens if China somehow would have a breakthrough and they get super AI before we do. I could also think of a scary scenario that's scary for the Chinese, which they decide that this technology, is too disruptive. Because as much as they would like to improve the living standards of their people, they want the Communist Party to stay in power.
01:00:43
Speaker 1: And so I think that's their great weakness. And I think this idea that we're going to lose to them, they're scared of this technology more than we are. I am a thousand percent convinced of that, for that reason alone.
01:00:56
Speaker 2: Clearly, they're worried. Clearly, they're worried about... The Communist Party is worried about other centers of power. And this is going to be a different center of power, not to mention, you know, they have to worry about that firewall. 20% of people try to get around it. Well, wait till they all have open source models and all of them can get around it.
01:01:16
Speaker 1: They're all going to get around it. Right. They're already at 20%. I look at it as, oh, they're already at 20%. Just wait. Right. Just wait. You know, you're just not going to get around it. Jim, this was great. Your optimism is, frankly, it's an elixir for me. It's a tonic because there's so much, especially at Substack. I assume, have you created your own algorithm? Do you only see good news? Do you only see optimism?
01:01:43
Speaker 2: No. I try to look at it. I try to look at all of it. It's harder to find like optimism. So you have to look hard for it. It's easy to find the bad stuff. You have to dig a little deeper, but it's out there. And I think it's, again, not just being optimistic about to be optimistic. It's about making sure that's like evidence-based. And I think there's a lot to be optimistic about.
01:02:09
Speaker 1: Tell people where they can find your stuff.
01:02:12
Speaker 2: Well, I'm at AI, but I have a newsletter. It's called Faster Please. It's a sub stack. And hey, I wrote a book in 2023 called The Conservative Futurist, available on Amazon.
01:02:21
Speaker 1: Or wherever fine books are sold. You can check that out too. Jim, I really, this was a treat. And like I said, I love your newsletter. Thank you very much. And I wanted more people to sort of, to hear your point of view, because frankly, it's been a lot of AI doom, right? There's a ton of AI doom out there. I understand how we're in a political moment. So everything is overly negative because of the current messaging environment. But I thought that- There's all that AI doom.
01:02:49
Speaker 2: And when it comes to the election, people are going to worry about the price of eggs again.
01:02:53
Speaker 1: I think they're still worried about the price of eggs again. Anyway, Jim, appreciate it.
01:02:57
Speaker 2: Thank you very much.
