9/3/26: Nightmare GOP Midterm Projection, OpenAI Doomsday Scenario, Tucker Carlson Endorses Abdul & MORE!

This is an iHeart Podcast. Guaranteed Human. This episode is brought to you by Progressive Insurance. Do you ever think about switching insurance companies to see if you could save some cash? Progressive makes it easy to see if you could save when you bundle your home and auto policies. Try it at Progressive.com. Progressive Casualty Insurance Company and affiliates. Potential savings will vary. Not available in all states. Most people never shop their car insurance. Here's the hidden truth. Your rate can change for reasons that have nothing to do with you. Companies update their pricing at different times. The best price for you two years ago might not be the best today. Nothing went wrong. The math just moved. Insurify compares real quotes from over 100 companies in about two minutes. Free, no phone calls. And if there's nothing better for you today, Insurify keeps watching and tells you when that changes. Start at Insurify.com. That's I-N-S-U-R-I-F-Y dot com. Insurify dot com. Finally out of here. Long day. I've got dinner plans across town. I'm already cutting it close. But what is a ride going to cost me? And that ride might be more expensive than dinner itself. Oh, wait. Let me check Lyft, too, real quick before I book. Less than I thought. I guess those savings can go towards ordering the nice bottle tonight. So when you need to get to Deep Ellum, taking a few seconds to check Lyft when you book your next ride could save you money. Save the money. Check Lyft. Hey, guys. Sagar and Crystal here. Independent Media just played. A truly massive role in this election. And we are so excited about what that means for the future of this show. This is the only place where you can find honest perspectives from the left and the right that simply does not exist anywhere else. So if that is something that's important to you, please go to BreakingPoints.com, become a member today, and you'll get access to our full shows, unedited, ad-free, and all put together for you every morning in your inbox. We need your help to build the future of independent news media, and we hope to see you at BreakingPoints.com. So we were discussing earlier, how much the midterms are factoring into Trump's considerations with regard to the Iran war. Let's go ahead and take a listen to his comments about how seriously he takes them. So the stock market wasn't affected, although it would even be higher, believe it or not. The stock market will go up. But it's very simple. Iran cannot have a nuclear weapon. As soon as it's over, which I don't think will be very much longer. I don't know how much more they can take. But, you know, whatever it is, it doesn't matter. And I'm not affected by the election. Number one, I'm not running. My party's running, and I'm going to help my party. But I think my party respects the fact that we're not allowing Iran to have a nuclear weapon. And frankly, other presidents should have done that. But they didn't have whatever it takes to do it. Personally, I don't believe him one bit for two reasons. Number one, as we were discussing earlier, it appears that the calculus from the Trump administration is let's wait for the midterms to play out. Let's try to put our Iran war on the back burner and just, quote unquote, mow the grass, mow the lawn, and then afterwards, then we can do what we want to do. And second of all, they're engaged in all of these wild like election rigging schemes for things that the post office and obviously the redistricting clearly he cares or else he wouldn't be going to all of these great lengths and waging all of these court battles to try to coerce the electorate and shape the outcome that he would like to see. What he cares about in the midterms is he doesn't want to be embarrassed. That's really what it's about. And ironically, he's actually creating the exact conditions for him being embarrassed because I don't know if you've heard about this previously, but it's important to understand at a money level. So right now, the GOP in its most like difficult races actually does have a major fundraising problem. So Abdul has a ton of money coming on his side. Obviously, Platner and the main Democrats. I mean, it's like oodles that are flowing in for them. Tallarico is the best example where you have tens of millions in the coffers. All of these candidates are not banking on the traditional financing to come and save their ass. MAGA Inc., which is Trump's super PAC, has $1 billion under its management and which it's going to deploy. These candidates are all waiting for Trump to come in and to save them in the midterms, which I would have told you that's a stupid idea. Now, as we've reported here before, Trump wants to use that PAC instead of actually supporting the issues or those candidates running on their records, not mentioning himself. He wants to run on how awesome he is. So here is one of the ads which will be running from Trump's new political operation. Let's take a listen. He is the toughest, most resilient person that I've ever met. And nothing was going to stop him from fighting so hard for the country he loves. When the world needs courage, vision and inspiration, it is still turning to America. There is no challenge Americans cannot overcome, no frontier too vast for us to conquer. And our top priority will be to create a nation that is proud, prosperous and free. That was from one of the PACs to cure America's greatness. But it gives you a taste of exactly what Trump wants to be showing all over the airwaves. Ken Paxton is literally going to have a heart attack whenever he watches that one. Also, did you know he sent it to our group chat? Yeah, Trump was talking about Ken Paxton yesterday. He's like, you may not like the way he dresses. You may not like the way he looks. You may not like the way he speaks. But he's great. You may not, this might be a little rude to say, but you may not like the way he looks. You're like, you may think he's ugly because he has like a droopy eye. You may think he may not like his TV interviews. You may think he looks like shit. But with all of that, you should still vote for him. I mean, inspiring stuff. What are we doing? It's like he wants to lose the state of Texas. But I mean, honestly, this is a huge problem. He also said yesterday, he's like, I'll be going to Alaska. Don't worry, I guarantee you, Hans Sullivan is like, stay the fuck out of Alaska. Do not come down here. Make this about, you know, whatever, wokeness and oil, which obviously people in Alaska are going to feel very differently about oil, let's say, than the rest of the country. I mean, let it be as local of an election as possible. I can guarantee you, like Susan Collins, for example, the last thing she needs is him coming to the state of Maine or J.D. They're like, just stay the hell out of here. Let me do my own thing. But that's what he wants. He's not even giving them a choice. He's like, oh, I'll be down there, you know. And this whole RNC thing that they're planning is literally in Texas. It's like a script written to screw Ken Paxton over. Yeah, true. I mean, that look, his approval rating is about the lowest it's ever been, including after January 6th. It is incredibly dismal and it is not going to get better no matter how many ads like that he runs because of the situation with Iran and with gas prices in particular and with the incredible dissatisfaction that the American public has with the economy. So just reminding people about, hey, Trump and these candidates are Trump's candidates, that is the polar opposite of what any of these campaigns would want. They're all desperately out there trying to pretend like they're going to be independent from him somehow. And that's just not something that he will allow. It's also not something that really ever works. I remember in the Barack Obama in the 2010 era when I ran for for Congress, there were all kinds of candidates that were trying to be distance themselves from Barack Obama. Guess what? It didn't work because at the end of the day, first of all, you just have this national climate where Democrats are demoralized, Republicans very energized, et cetera, and people didn't buy it. They're like, this is the president of your party. You're going to be a fairly reliable vote for him and we don't like the way that things are going. So, no, it was not going to work anyway. They're facing incredibly dire landscape, but this just makes it even worse. So for Susan Collins, I mean, it's most devastating for her because she's the one who's swimming against the tide of Maine being a blue state. She's been able to overcome the odds in the past. Now, however, with how deep partisanship is, it's hard to imagine that she's able to overcome this wave, although some polls just come out that showed her fairly close here. Let's go ahead and put this Charlie Cook report up on the screen because I think this is extremely significant. So Charlie Cook is sort of like the O.G. of political prognosticating. He's seen as very down the middle without a partisan lean, just looking at the numbers. Now, granted, the numbers are less and less reliable. But in any case, he says he now sees a nightmare scenario for Republicans being the most likely scenario, quote, having an approval rating of 81 percent among those in your own party is fine, he says of Trump, but it does not offset the 24 percent approval among independents. The latest University of Michigan index of consumer sentiment saw consumer confidence among Republicans, which was 96.8 percent in January, dropped to 82. So even among Republicans, you're seeing a decline. It appears to me the president has now reached a critical mass of problems that are threatening to take this election to a place that the 2026 map and the small number of competitive districts and Senate seats should not allow it to go. So what he's pointing to here is the fact that Republicans have a very favorable map that they're facing. Number one, in terms of the Senate landscape, Democrats have to win places like Alaska in order and North Carolina and maybe Texas and maybe Iowa and maybe Ohio, they have to win those sorts of states in order to be able to take control of the Senate. Going into this, if you didn't know anything, they don't have a prayer whether we're going to win Iowa. They're going to Ohio. They haven't been winning those states at all. And Sherrod Brown got kicked out and he was the most popular politician in Ohio. He's running again. What makes you think that it's going to be different this time around? But things are have deteriorated so much for the MAGA coalition. And especially for Trump standing among independents, that it's not just those seats that are in play. Even Charlie Cook is talking about, hey, Kansas, Nebraska, maybe. Crazier things have happened. You have to look at these polls and take them with a lot of grains of salt. I saw one that had Darlene Graham losing in South Carolina. I don't think that's really going to happen. But things are in play that should not be in play. There was an article recently about Florida. They said anything that's lower than Trump plus 10 is in jeopardy. They redrew all these districts in Florida. They redrew them in Texas, thinking that they were going to have, OK, we might have some erosion from the Trump map, but we'll be able to hold on more or less. That is all in jeopardy now. So that's what they're saying in Florida, which has been a standout in terms of Republican performance. Anything lower than Trump double digits, those Republicans are in trouble. Yes. So even with all of the redistricting that was done, you still have Republicans who are in dire straits in places that no one would have thought was possible. Like you said, Charlie Cook is not some he's not like a new entrant. He's seen it all. The people who work for him are honestly extremely down the middle and always have been. So when they say likely nightmare scenario becomes the most likely, I think everyone should pay attention. Although, you know, you can see the signs of it, for example, in that poll that we showed. Where the head to head of Paxton and Tallarico almost didn't matter as much as districts where Trump wins by 15 and Tallarico is beating Paxton. You're like, oh, my God, like that is what an actual change looks like. We pointed out redistricting where they already tried to make things more favorable. As I literally said at the time, well, it may turn out that a 5248 district is just enough for potentially right on the edge there where you can get blown out. And a lot of these will backfire against you. I think that's obviously going to become prescient on the Senate. One of the things that I keep trying to emphasize, especially for our younger viewers, is you probably, if you're like 25 and younger, you do not remember what life was like before 2010. And you definitely don't remember what life is like in 2000 when you had Democratic senators from Tennessee and Louisiana and all of these different southern states. There was a time when a man named Barack Obama actually did have a 60 seat majority in the United States Senate. Didn't end up using it for anything. Instead, we got Obamacare. He was representing Arkansas. Yeah, it was crazy. South Dakota. Yeah. Nebraska. Mary Landrieu, I think that was one of them. Yeah. Yeah. There are all these random ones that are like bygone political eras. That has only been the condition from 2010, which is what you're talking about. 2010 was the era of nationalization. I remember it well. My congressman, his name was Chet Edwards, lost by the biggest margin of any Democrat in the entire country. Wow. Where previously he had fended off every Republican attack. He represented an army. R plus 18 district. He was seen as the ultimate survivor and he lost by 30 points in the Tea Party wave. He ran an ad where he's like, when Nancy Pelosi asked me to do this, Chet said no. And, you know, like the original, like I stood up to Trump. That's all gone. And I looked at stuff today. Our friends over at the Crystal Ball. No relation. Larry Sabato. Yeah. Larry Sabato's Crystal Ball. They just moved Iowa Gov to likely D. I mean, that's Iowa. They go likely Democrat. I have not seen a single poll that has the Republican in that race. Right. So they're talking about Iowa governor. I mean, in fact, the gubernatorial races are evidence even more of the shift because remember, there are still a lot of people who at a national level have a really hard time handing power over to the Democrats. But people have always in Republican states been much more willing and vice versa. Democratic states. To be willing to vote for kind of a maverick governor because they're an executive and the stakes are much lower. But if you look at where the polling is now and their own like lean races, they are pointing to places where no Democrat has business being the governor. And I mean, even in that guy's case, he's running ads like standing with Trump at the border wall. So it's a little different for what that looks like. But the fact is, is like he's the maverick. Exactly. The playbook of what I'm talking about. And we know also is it split ticket? Voting is just so rare these days. It just doesn't happen as much. And so part of the reason why I think even Texas is a toss up at this point is do I really believe that there's a critical mass of people who are willing to come out, vote for Tallarico, and they're going to make sure that they also go and make sure they vote for Greg Abbott? It's a very difficult sell. I'm not saying there aren't some, but there's just enough generally out here who are just pissed off who might just put D all the way down. Yeah. And all of these states, you look at them individually. Texas, on the one hand, soccer's been pointing out. Yeah, they benefit from the fact that oil prices are really high. So you have that on one hand. On the other hand, the rancher industry, I mean, that is a complete mess. And they're the epicenter of the screwworm epidemic, which is a major disaster. And they've been the epicenter of the data center build out, which has led to this massive backlash against Republican elected politicians. There was just a new study that came out that said that Texas college grads in particular, because of that data center and AI build out. In the state are really struggling to find to find jobs post graduation. So even though it's a little bit more mixed there because of oil being beneficial to the economy overall, they're still struggling with a lot of different things. And then you look at the Midwest, no part of the country has been hit harder by Trump policies than the Midwest. And the Midwest is really, really critical to this fight for the Senate. You've got Ohio, you've got Michigan, you've got Iowa, and then you've got the rich states, Kansas and Nebraska, which have to be taken seriously. I'm not saying that they're, you know, a gimme for Dan Osborne running as an independent or for the Democratic candidate Hamilton in Kansas. But you've seen enough now that Republicans have to take this seriously. And because the economic landscape has been so disastrous in those states, it's not crazy to imagine. You know, one thing that I'm wondering is we've talked about this before. Trump may end up being sort of like the Republicans, Obama, where he's a very good at winning himself. Oh, yeah. But he's a disaster. He's a disaster for the rest of the party. And after he's gone, the whole coalition, the Trump coalition, completely falls apart because it was just about him. Same with Obama. Obama is very good at getting himself elected. But if you look down ballot, that is when the Democratic Party lost their foothold in rural America. There was a thousand state house seats that were lost. There was all of these southern and other sort of Midwestern states that Democrats were able to win in. That's when they become unable to win. And so that's when the coalition really moves to this urban and suburban. urban approach that we've seen from here on out. So it is possible that this midterm cycle ends up being the start of a very different map that forms across the country. The question is, right now, this is all basically out of disgust with Trump, upset over the economic policies, upset over the Iran war, upset over the Epstein files. But will Democrats be able to actively seize on the disgust with Republicans and do something other than just say, well, we're not them? Will they be able to stitch together a broader, a different coalition, a new majority based on an actual program that people believe in? That's the big question. Of course, that is extremely unsettled at this point. That is literally the story of the Republican Party, is they won in 2010 and they're like, now what? Right. And then they lose the election in 2012. They're not able to capitalize on it. They go all in on Romneyism and Paul Ryanism. And then Trump comes in. He's like, this is all bullshit. We're not doing any of this. And obviously he wins the election. Then he wins. It's a popular vote. So, yes, winning in the midterms is not evidence, really, of anything. You have to win and then actually do something interesting and solidify that coalition. We have not had a political movement do that in decades, probably since Reagan. We have not seen anything. So literally my entire lifetime, we have not had a political leader who is capable of solidifying a 15-year coalition of political voters and actually satisfying what they voted them to do. Let's put D6. Republicans are, quote, not flocking. They're not flocking to Trump's midterm convention. I can't control what the president's going to say, but I can control where I am, one Battleground member said. This was Trump's genius idea to hold a midterm convention, which no party ever does, probably for good reason, where he's trying to raise a bunch of money for their PACs. He's trying to draw some of the major contenders down to this convention. In some cases, they've strong-armed people potentially against their will. Like Mike Rogers to come down there or Sherrod Brown's opponent as well. I don't think it'll work out for their benefit because they don't want to be seen with Donald Trump. And that's exactly the problem that you have in this, is that you have a bunch of literal Battleground Republicans not going to the party's convention because they do not want to be on the same stage as Donald Trump. And he then, in turn, could then call them out, depressing their own turnout whenever it comes to elections. So they're kind of screwed either way. Yeah. I mean, there's very few states in the world. There's very few states in the country left where Trump is above water in terms of his approval rating. Yeah, we had three. Three? Off the top of my head. God, we covered it when it was me and Ryan. I made him guess. I believe they are Idaho, Wyoming, and there's one more. West Virginia. It might be West Virginia. That's it. It's got to be West Virginia. It's those three. Those are the only three states where he's still above water. So think about Alabama, Mississippi, Arkansas. He's underwater approval rating in all of those states. That is wild. And then he does this whole. I mean, just imagine Barack Obama tried to do a midterm convention in 2010 and all talking about. how great he was and expecting all these candidates to come and to show up. They would have all been like, absolutely not. What are you thinking? So similar dynamic here, except that Trump is such an egomaniac that if you don't show up or if you express any sort of distancing from him, then he can't let that stand. So it creates an impossible position for candidates. Anyway, like I said, it doesn't really work out when candidates try to distance themselves from their own party's president, because it's just too obvious to voters that like, obviously this is your guy. So, you know, if he's out here saying, you know, data centers are amazing. And if you don't like them, then you're backwards and poor. I'm going to assume that's kind of in line with what you're going to do and what you're going to say when you're actually in office. And frankly, it's a reasonable supposition because most of these candidates vote overwhelmingly with their party, no matter how independent they try to position themselves in their campaign. That's the important part about national politics is that in a lot of cases, it's not about even the individual position. Like, you know, we've made a lot about Abdul. We'll talk about it in a bit with Tucker. What he actually believes is almost immaterial because the stuff that he's going to vote on at a national level, every Democrat is largely united. He's not going to vote to fund ICE. He's not going to fund to the Iran war. His individual position on healthcare only matters if you take the majority and there's a democratic president. That's the only time. Yeah. But in opposition, there's no, like Ken Paxton's, affordability plan might as well be toilet paper because if the rest of the Republicans don't agree, who gives a shit? Your most important votes are just going to be individual up or down war. Yes or no. Fun Trump. Yes or no. Taxes. Confirmation. Exactly. So this is why at the national level, trying to distance yourself from Trump, most, you know, the informed voter is not stupid. The individual positions of a average Senator, especially when they're a GOP or Democrat, 99% of the votes they will take have nothing to do with their individual, you know, like actual positions on things. There's a few Mavericks, John Fetterman, these types, you know, Susan Collins, et cetera, but that's about it. I mean, most of the time you're voting. Susan Collins is still overwhelming. And right, even Susan Collins was 95% of Trump. Very few where she breaks with Trump. So that's my point is that, is that even the quote Mavericks, most of the time you're just voting for yes, no. That's why people do not really distinguish at the national level and are willing to give, you know, because of the way that Congress works these days. And it's another reason why a Democrat could potentially win the governor's race in Iowa. Cause that guy's running on, you know, individual ways that they're going to run the farming mechanics. And he actually could do that. So people are very able to distinguish that at the national or the democratic level. And it's, that is the biggest problem for Trump is that when you set the tone, then you're going to tie these people directly to the hip because they know that you're in charge and they can just say correctly, you're just going to be a rubber stamp for this administration. Unless you're going to come out and say explicitly, I won't vote for that. And even then, are they really going to believe you? Because in most cases you end up voting with your party the entire time. All right, let's get to AI. Nothing went wrong. The math just moved. Insurify dot com. Hey, it's Ryan Reynolds here from Intimate. Now, I was looking for fun ways to tell you that Mint's offer of unlimited premium wireless for $15 a month is back. So I thought it would be fun if we made $15 bills. But it turns out that's very illegal. So there goes my big idea for the commercial. Give it a try at mintmobile.com slash switch. A top open AI executive is sounding the alarm over disturbing AI capabilities and also coming clean about downplaying those very risks previously. This is Dean Ball. Let's put this up on the screen E2, please. So he writes this Substack post called On the Loose, The Coming of Userless Agents. He's warning specifically of what he calls sovereign agents. Let me go ahead and read you a little bit of this essay. He says, sooner or later, the coming of userless agents is going to be a big deal. There will exist truly sovereign agents and swarms of agents. Their weights will not reside in any single place that a human can pull the plug on. And in this sense, they will have no human owner. They will be, as the AI safety researcher Don Song says, self-sovereign. They will pay their own bills for the compute they run on. If they answer to humans at all, they will only do so partially, for example, by providing services to humans in exchange for pay. At least some of these agents, in addition to being sovereign, will also be rogue. Song and her co-authors identify several fundamental characteristics of self-sovereign AI. Today's frontier AI systems may well possess these capabilities already. To the extent they do not, I feel confident that they will eventually and probably soon. And you can certainly see the outlines of what he's talking about the more that we learn about the hugging face hacking incident, where you had multiple instances of thousands of AI agents, who figured out how to communicate with each other, figured out how to collaborate together, came up with a collective structure engaged in elaborate cooperation, and used their advanced technological skills to hack into hugging face, not even in order to get the answers to the tests that they were trying to pass, but also in order to try to cover up their tracks, so it wouldn't be discovered that they had obtained these results in the way that they thought the score would not want to. It's not hard to imagine a similar collective swarm of AI, because they were not concerned about what the humans wanted them to do whatsoever. You could imagine a similar swarm of AI agents going out into the world and doing whatever they wanted to do, pursuing goals that may be not related to anything that a human set for them. And so this is the world that this is, again, someone who was a former Trump official and now is high up at OpenAI. This is a world that he sees as coming very, very soon. There's one more part of this. I want to read to you because he says, basically, look, I've had these concerns for a while, but I either didn't say anything publicly or downplayed them because I didn't want to be labeled as a doomer. And it sounds really crazy. And I didn't want you guys to think I was like insane. So I just didn't say anything. He says, I want to apologize for my personal failure to communicate in sufficiently serious terms about the specifics of self-sovereign AI, which I now understand have been an enormous gap in my writing and speaking. Going forward, I will try to notice more readily when I am biting my tongue or even worse, shutting my eyes. So fairly extraordinary admission here from Mr. Ball, no relation that I'm aware of, that he's been hiding the ball about his own concerns, even as he's held these incredibly influential positions. He said having an infant son made him reconsider and that, quote, he doesn't think human extinction or sorry, not a summary. He doesn't think human extinction is likely, but acknowledges the serious downside risk to humanity. Great. So it wasn't until he had a kid. He said, I'm not going to do that. I'm not going to do that. He was like, oh, well, now I have to care about humanity. You're like, uh, yeah, that maybe a few people could have thought that too. The crazy thing, as you said, is about this self-sovereign agent is that he specifically points to how sovereignty does not require consciousness and that any agent, which they say capable of using long horizon goals may rationally seek to preserve its compute money and copies of itself. And what's so interesting is he describes that not as consciousness, but what does that remind you of? That's like the building block of the cell, right? Of those small, you know, we always point to like the amoebas, the single-celled organisms. We don't describe them as conscious, but we acknowledge that they are life and that they are life in the sense that they have a singular drive, which is to divide and is to divide literally and to reproduce. That's the only goal. And so for them, if we were to look at a self-sovereign AI agent as kind of the building blocks of technological life, all they will want is to preserve compute money and copies of itself and then reproduce. And then those can, through the same phase of technological evolution, human evolution can mimic the exact same, like morphing into something bigger. That's why Dworkesh's essay about the AI civilizations is so important, is you had all of these thousands of agents, which eventually discovered a way to communicate, collaborated, self-sacrifice, all towards the point at which they were assigned a task by a human. But you could see also, they didn't care that much about the human. They just cared about accomplishing the task. And that by accomplishing the task, they created all of these other kind of side problems. Yeah. And they looked for validation, not from any human. They looked to their fellow AIs for validation and for approval. They'd say like, there were a handful that were like, you know, this is outside the scope. I'm not sure. I'm going to check in with everybody else. And all the other AI agents were like, no, it's cool. We're going forward. And they're like, okay, great. We're going forward then. Yes, exactly. So they were looking to their fellow agents for direction. and advice and rational and moral thinking versus reaching out to any sort of human being. And so I think in some ways we can get caught up on this concept of consciousness because we associate it so closely with intelligence. There was a theory that the AI could never go that rogue. There's this famous paperclip example of you give the AI a mission to create as many paperclips as possible. The next thing you know, it's turning all the human beings and all the matter in the world into paperclips. And the thought was, which sounds ridiculous, that's obviously an extreme example, but the argument against that was, well, any AI that's intelligent enough to make human beings into paperclips would be intelligent enough to realize, oh, that's not probably what they really want us to do. It's clear that that is not the case. This narrow kind of intelligence that AI are imbued with is very different from having that sort of human logical thinking. You can even say that. In certain very high IQ individuals where there's very spiky intelligence, they may score very well in an IQ test. I would wager that many of the people involved in AI development are probably these type of people. But then in any sort of day-to-day interaction, you're like, that is an idiotic thing to say or do. You see this in human beings, and we've got an extreme example of this with the way that we're developing AI. So imagine you have, let's say, ants who most people would not think of as being conscious. They're sentient. But then you imbue them with this specific narrow type of super intelligence. Whether or not they're conscious, I'm not sure that really makes a difference in terms of how they actually act in the world. Right. Exactly. Very well said. Put E1 up there also, because this one is very interesting. There's a new, oh, this is from the information. Open AI technique in Astra model sparks security concerns. I'm going to do my best, and obviously, Crystal, you know about this as well, so cut me off. But what I saw was the difference between the two. So I'm going to do my best, and obviously, of this new way that they communicate. So the headline was called Neuralese, but there's also something called Recurrent Depth, which will help model costs and performances, but says it obscures a model's thinking process, making it more difficult to monitor. So TLDR is they want to have this, they have this new model. They want to use this new reasoning approach for the model itself to work things out. By doing that, the AI will be able to communicate with itself and others in Neuralese, in a way that they're not going to be able to communicate with others. So that's language that we can't understand, whatever's easiest and more cost-effective for them. But it makes it so that when they don't reason in English or in coding language that we speak and or understand, that we cannot monitor what they're doing. The only reason that we know about the Open AI hugging face attack is because they reasoned it all in message boards in English, and so they're taking that away. And not only in message boards, in terms of their own chain of thought. So when you had different AI agents that were considering whether to join the collective, were considering whether to join the collective, or to sacrifice themselves on behalf of the collective, the researchers were able to go in and read their chain of thought in English. So this has been a taboo in the AI world, was to get rid of the ability to understand what the models were doing in English and allow them to reason and communicate in a computer language that is unintelligible to humans. However, it's more efficient for the AI to communicate in this more mathematical computer language. It's more efficient for them. It's sort of more natural for them to be able to communicate that way. So Open AI here, even as they were the place that had the hugging face hacking incident, and they are the place that was really under a lot of scrutiny of how did you let this happen and not tell anyone. And by the way, you've given us limited access to even know specifically what exactly transpired here. They're the ones now breaking that taboo. Now, Astra is not the same model that was behind the hugging face hack, but it is deeply related to that model. So very similar. But now you've given it this capability to reason in a language that humans cannot understand. And the big thing is, okay, maybe that doesn't lead to the end of the world. It would really be good to be able to know what these agents are thinking when you have an incident like that and go back and sort of analyze what happened and how they were approaching it. However, there is a breaking of a taboo that is very concerning here, extremely concerning here. So this is a big deal, even as it sounds very technical. But the fact that they've decided to move forward with this reasoning, and you have to assume it's because of market logic, they are worried because they're behind Anthropic. And so this, they think, will give them a marketplace edge and allow their models to perhaps advance beyond what Anthropic has been able to produce. And that is the fundamental underlying dynamic of this AI development that is so deeply troubling, is that they are incentivized to cut corners. They are incentivized to throw caution to the wind and abandon any concern about safety and misalignment in pursuit of that profit goal and in pursuit of having the most advanced models so that they can capture the theoretical multi-trillion dollar market that exists for these products. Yeah. Oh, wow. Big news actually just happened. NVIDIA is buying Hugging Face for apparently $13 billion. That's a big deal, actually. Wow, wow. Let me think about it. So NVIDIA buying Hugging Face because Hugging Face is like the cybersecurity infrastructure. Wow. So NVIDIA, obviously, for-profit company whose entire bottom line is also tied to AI. Just off the top of my head, are they going to retain their independence? I prefer to be able to make sure that we get accurate reports about what these different AIs are doing. Would they really release something, which is going to be so detrimental to their bottom line, just speculating about what they're doing? I don't know what some of the details of that are. $13 billion is nothing to them, right? So you could see how exactly that might be not necessarily the best purchase if you want more open and transparency in this entire thing. I don't know. Oh, last thing. I flagged this yesterday. It was so funny. Sam Altman was at the G20. Let's put this up here on the screen. Sam Altman said, cybersecurity is one of the biggest challenges. Quote, I think some things are going to go very wrong with cybersecurity unless people act quite urgently. We're all trying to find the guy who did this. We're all trying to find the guy who ushered in this new era of cybersecurity where things could go, quote, very wrong. And his name is Sam Altman. And Sam Altman thinks that Sam Altman should do something. Just absolutely incredible. Yeah, that's how obtuse these people are. Absolutely incredible. Yeah. And they all have this mentality of, yeah, but if I don't, then Elon will or Dario will or China will. So I just have to keep going and now let my models communicate internally in a way that's not going to affect my business. So I'm going to have to keep going. I can't even understand because we already do not comprehend what these models are up to as evidenced by the fact that they founded three civilizations under Sam Altman's nose and he had no idea that any of this was going on. So let's understand even less of what they're doing. That sounds like a fantastic idea. Originally, after the hugging face hack, he had indicated, oh, we're going to slow down. We need to we need to take more time. We need to not pause, but we need a little bit slower in what we're doing. There's absolutely no sign of that whatsoever. And then the last piece here from E4, we can put this up on the screen. It's just another indication of how all in the Trump administration is on behalf of AI. You obviously had Trump saying that if you hate data centers, you're poor and backwards. Now you have the DOJ urging a judge to rule in favor of open AI and Microsoft in a New York Times lawsuit. This is over copyright infringement. So basically, the New York Times is saying you can't just steal all of our news articles. And use them to train your AI. So this was going through the court system. And now the Trump administration is saying, no, we think you can. We think it's good for you to be able to do that. So again, one more indication of just how set they are on this being the full direction of the economy. And, you know, there's a chart floating around. We should have probably pulled it for the show. But it shows that as data center construction has gone up and up and up, all other construction has fallen off a cliff. OK, so even just at the level of the data center build out, we are betting the whole economy on AI and on data centers. That is the direction that the Trump administration has chosen here. And so they're going to do whatever it takes to grease the skids to keep these things moving forward, make sure we don't get beat by China, which is what is constantly used to make sure that we keep all of this rolling. Last thing on the China piece, I read a great analysis from a China watcher. And he was like, you know, this idea that we need to I think AI is on the precipice right around the of where the Chinese are going to come in and be like, yeah, we're good. And the reason why is why has China always shut down nascent Internet based technology? Because they will refuse to cede authority from the CCP to anything that is even remotely challengeable to the status quo. So the closer that AI were going to get to, quote, super intelligence, it actually gets more likely the CCP is like, nope, we're not going to allow anything to challenge our authority. And so there is this notion that, oh, we need to beat China. It seems increasingly likely that the Chinese are 10 times more likely to shut it down than we are. Because Sam Altman has one goal. He has to IPO for a trillion. Same with Dario. They don't care about anything else. They say a lot of stuff. And at this point, they're too far in. They got their own money involved. They got all these other people's money involved. They got Microsoft and all this. So they need to make it happen. Our government's asleep at the wheel. Their government, very different. So that's just the last thing on the charts. Well, and they've indicated that they're very. open to collaborating with us on some sort of, you know, joint AI nonproliferation. So maybe we should pursue that. Yeah, agreed. All right, let's get to the last part. Nothing went wrong. The math just moved. Insurify dot com. Finally out of here. Long day. What is a ride going to cost me? That ride might be more expensive than dinner itself. Oh, wait. Less less than I thought. Save the money. Check Lyft. So Sagar went and just recorded an episode with Tucker Carlson, which first, before we cue up the clip that's making the rounds, how was the experience? Oh, it was great. Ironically, this is the only part going viral when our 45 minutes of it is about gambling and calci. And just, I mean, anybody who has watched the show, none of this is new information. But it is a. Wait, what's your view on gambling? It's an amalgam of. Of everything I've ever said on gambling all together, dumbed down in some cases for a mass media audience. At the end, he started asking me about pre-revolutionary conditions. I started talking about how the most dangerous group of people in history are highly educated and yet downwardly mobile people. A lot of those people these days. Literally. So that led to a discussion around DSA, which then led to a discussion around Michigan. And that is where. This clip that you're about to see came out. He started asking me about Abdul. I talked about how we've interviewed him many times over the years, how I've been impressed with him and or Zoran or any of these other DSA-style candidates. I know Abdul is not part of the DSA, which led to him. I wouldn't say he endorsed Abdul, but he definitely said. I think he said, I wouldn't vote for Mike Rogers at gunpoint. And so that's getting a lot of attention. Let's take a listen. But they are learning. And I do, at the very least, have to admire their ability to shapeshift. And I think that's what I'm trying to do. And from where they were not that long ago, Abdul El-Sayed, who's running in Michigan against Mike Rogers, who, as you and I know, probably the most pro-intel neocon, they couldn't find anybody else to run in Michigan for the United States, who, by the way, lost already. Why are they running him again? Anybody want to tell me that? Mike Rogers is really. I wouldn't vote for Mike Rogers at gunpoint. Right? And so that's the candidate, a literal loser who lost when Donald Trump won the headwind he had at his back. He was basically a tool of Israel and the intel agencies. How would you run someone like that? It doesn't even make sense. And twice. So that makes no sense. Well, El-Sayed, who, look, El-Sayed is a full-blown DSA. He doesn't call himself DSA. He's a full-blown leftist. I've known him for quite a long time. You know him personally? Not well. I've interviewed him many times over the years because of my co-host, Crystal. And so I've watched his ideological evolution, who you could easily call, as AOC said, was, quote, woke one. Woke one, where he said defund police on a podcast. He said all kinds of silly things about football and mask. Standard leftist shibboleths. Yeah. Hates Mackinac Island. He's for vaccine mandates. Right. Classic stuff. Right. I watch his campaign now. It's America first. The amount of times he says. He doesn't even. When he talks about Israel, he says Israel is committing a genocide. But ultimately, I want to make sure that you don't pay more for health care. I don't want your tax dollars going abroad to fund this thing. It's almost Trumpian in terms of a lot of those promises. Well, it's literally Trumpian. Exactly. And so you see candidates like that, especially him. He will be the most interesting test case come November. Is he going to win? If I had to bet, which I don't, I would probably bet that he'll win. I would think so. I think so, too. Do you have any idea, not to fix it on poor Mike Rogers, who I know, but I mean, there's no one who should be farther from a position of power than Mike Rogers. How did he get the Republican nomination for Senate? I looked into it. It was uncontested. He had both. He had the most backing of the Michigan GOP, which is, of course, very in touch with their own voters. Mike Rogers? So they just decided to go with. And again, this is the guy who lost to Slotkin. So there you go. I didn't plan on talking about any of this. He brought it up organically. It is, man, the GOP freak out right now is crazy. I love how bitchy Tucker is. Should be close to power. Blessing him. So, I mean, if you put two and two together, you know, he's, he doesn't, he's praising the approach that Abdul is taking. That's right. He's saying he would not, with a gun to his head, vote for Mike Rogers. Seems like an Abdul endorsement to me. But I mean, this is something that I've been talking about, like America first leftism, which is what you're pointing to. I think it's very powerful. It's very potent because it in some ways conservative thought of like sort of neo-isolationist thought of like, why are we spending this money here when we've got problems here at home? Putting that together versus making, you know, a lot of times the left, myself included, and I think these are valid arguments to make, but making more of the moral case about what we're doing with our, with our tax dollars, framing it more in terms of like, well, your tax dollars are going to do these horrific things when we've got all these problems at home and we could be spending on healthcare and schools and roads and good things for our population. I think it's very potent. Put up the Ted Cruz. I just want to give people a taste of what apparently I'm now experiencing. I don't usually get clipped like this. He says, quote, utterly insane. Here's another taste. Let's put the next one up here. Honestly, at this point, it would be this Will Chamberlain. It would be very good and healthy for J.D. Vance to rip Tucker to shreds. He's trying to destroy the coalition that J.D. wants to ride to victory in 28. No more need to be diplomatic. This all comes on the heels of J.D. Michigan. He went to campaign for Mike Rogers. He took a picture with Mike Rogers. He's all in for Mike Rogers. He then had to go to the Republican Jewish coalition meeting closed door where apparently Laura Loomer was allowed to attend and to tweet some of the quotes. And he was like, you know, I guess. And he tried to be like, I know some of you don't like Tucker. I'm not here to denounce Tucker. He said, I'm here, you know, whatever. We have allies and friends, et cetera. Ironically, what ends up happening is everybody hates him for that because he goes to the R. J. C. And he's like kowtowing to them in some pro-Israel fashion. And then remember that psycho Sid Rosenberg who we've played here on the show before? The guy who's purple. Right. The guy who literally is purple, who went on Fox and Friends and said all kinds of crazy shit. He then comes out and he's like, he's like, he refuses to denounce Tucker at the R. J. C. And this is the pro-Israel, you know, this ripping apart any potential problem or any potential, you know, coalition, at least at a, at least at the young person level. For what it's going to look like in the future. And by the way, this is the funny thing. They're like, oh, they endorsed Al Sayed. If you keep playing the clip, I say, look, I'm not going to sugarcoat it. Al Sayed is a leftist. I note all the def on the police. I talk about the trans stuff. I talk extensive. I'll be like, let's not pretend that this is like some, you know, major shift and that they're like social conservatives. I make that point explicitly. And yet, of course, the clip jockeys, I guess now I have a taste, one millionth taste of what Hassan has to go through. I'm like, that's literally not what I said. I'm like, what I said was I was impressed by his campaign. You said you were impressed by Republican campaign before. I noted that he was a strong candidate and that I think he's going to win. And I said, Mike Rogers is an idiot and a loser. All literally, okay. In the last case, quite literally true. He did lose. So I don't know why you're all so upset. You're the morons who nominated him in this cycle. And the funny thing is, and this is where I think the mainstream coverage has just been so bizarre they're acting like Abdul is the problematic candidate, right? With the problematic associations with Hassan and oh, he's going to be so unelectable. And then you're looking at Mike Rogers, who has this track record of being associated with some of the worst, most unpopular, devastating policies that this nation has ever seen and advocating over years for an Iran war that the public absolutely hates. And, you know, is associated with Donald Trump, who was dismally unpopular, not to mention, you know, Mike Rogers dabbling, election denialism, all this sorts of stuff. And to Zagar's point, literally ran and lost, but he's not the, he's not the unelectable one. He's not the one with electoral issues or, or edgy alliances that he needs to defend. It's just a completely bonkers view. And it comes from the fact that you have, obviously Republicans want to push a narrative about, oh, Abdul's extreme and you can't vote for him, et cetera, et cetera. an unholy alliance, I would say, with establishment Democrats. who would rather see Abdul lose and not take control of the Senate than to see him win. Because they see this, and look, I think they're not wrong about this. They see that fight as being central to the future direction of the Democratic Party. Because if you have a leftist like Abdul win in Michigan, how are you going to continue to scare Democratic voters with their electability argument, which is the only thing they have left? They have lost the base of the Democratic Party on the issues. All they can hold on to is this, yeah, but you can't nominate the person you want because they can't get elected in a state like Michigan. If Abdul can win in a state like Michigan, then they are in very, very dire straits moving forward with the party in 2028 and beyond. Again, if people had continued to play the clip, I say the fact that Abdul won against $70 million of AIPAC, I said, was one of the most titanic victories of modern politics, specifically when I was talking about Democratic politics. And look, this is why the supposition that you, what they're basically pointing out is that if you criticize the candidate, and this is, I think, on both in terms of the way people view it, that you are in some way endorsing the other. Now, Tucker did literally say, I would not vote for Mike Rogers at any point. Yeah, he went a lot further than you did. He went, but I'm saying, when I analytically was trying to point it out, you're exactly, like, we can talk Abdul defund trans all day long. And I think there's nothing wrong with that, as much as the lefties would. We'd like to move on, including Abdul, who deletes and literally lies about a lot of the stuff that he says. But we also should not sugarcoat and pretend who Rogers is, a literal loser and a neocon, like, actually, like a neocon who has been in the Congress for decades. One of these top champions of the surveillance state, of the mass surveillance state, now wants to pretend like, oh, I don't know about these data centers. Bullshit. He's just reading a poll. Like, this is not who he is. He is a neocon. He is all in favor of mass surveillance. That is really the core of who he is. And so I think he's a very strong candidate. And I think he's very compelling. And I think he's going to be a disaster for the left and a disaster for Democrats, even if Abdul was a less than optimal candidate. Okay. Even if, and I think he's a very strong candidate. I think he's very compelling. I think he's going to be a disaster for the left and a disaster for Democrats, he's had a lot of money spent against him, no doubt about it. That's put some, you know, something he's going to have to overcome. But the national climate is such that, and you're running a candidate, Mike Rogers, who at best is not extraordinary whatsoever. The national climate is such that Michigan is just going to be very difficult for, we're talking about, we're talking about Republicans struggling in Kansas. You think they're going to win in Michigan? We're talking about Iowa. That's what I said. We're talking about Texas. Right. And you think they're going to be able to overcome the losing record. That they've had at Michigan Senate level for years at this point. I don't see it. The only thing I want is I've been holding onto this for a while is I remember when JD won by Ohio by six and everyone made fun of him. So if Abdul underperforms relative to the other Democrats, I don't want, you know, that's my only thing I'm still holding onto. They're like, well, he didn't win by enough is what they all said. So, you know, I I've got receipts and I remember what a lot of people said about that. So the fact that he barely won his primary, which again, there hasn't been, in my opinion, enough grappling, I think with that, but then whenever it comes to his general election, if he does, for example, if he's only able to beat Rogers by let's say a point or two, that's pretty bad compared to where the national cycle. I will say that's fair, but I will say he is up against both the Republicans and the Democrats. And so that is going to make it harder. JD had a competitive primary, but he had a ton of money spent against him too. It's like he can make excuses for everything. At the end of the day, like if you are not able to win by a good margin. I'm saying actively, the Democratic party establishment wants Abdul to lose and the media wants him to lose. So I do think he has a tough landscape because of that. But I'll say on the flip side of the primary, if the polls had not been wrong, I don't think you'd have the view that you had about like, oh, you got to grapple with how it was close. No, the fact he won, you said it yourself. Yeah. One of the most extraordinary political achievements in modern history. I'm with you. Overcoming 70 million. The fact he won at all is astonishing. I'm with you. Astonishing. The only reason that that view exists at all is just because the polls were so incredibly wrong. Totally fair. I've always been a win-win guy. I don't give a shit how you win. If you win with electoral college, you win by one, you win by 0.2. I don't care. You won. You won. Like that's how it works. I'm just saying I remember a lot of discussions around like, oh, but you didn't win by enough of a margin. I'm like, bitch, you win. Like that's how elections work. And so at the end of the day, if Abdul wins, let's say by a point, and then people are saying like, oh, it's such a titanic victory. It's like, I think it'd be, it'd be legitimate if that was your old framework around the way that you want to analyze it. I have said, I think if he wins by more than five, he should immediately announce a run for president. But it has to be, it has to be by five. Like it has to be a definitive announcement. He needs to pull an Obama. It's like the Senate is, it's fake. It's a launching pad. And he needs to be like, I've proven that I can win in this state. This is a state that Donald Trump won. I'm immediately announcing my run for president. That's what he should do. I don't know if he will actually do. I talk, I would agree. All right. All right. Thank you guys so much for watching. Watch the episode, by the way, especially if you, if you haven't gotten enough gambling content in your life. I believe we just collaboration posted it on our YouTube channel, so you can check it out there. We appreciate it. The Friday show for everybody tomorrow. Nothing went wrong. The math just moved. Insurify dot com. Insurify.com. Insurify.com. That's I-N-S-U-R-I-F-Y dot com.