AI Agents Are Already Escaping Labs — Now What?

Garrett Fulce sits down with Peyton Hornberger of the Alliance for Secure AI for a genuinely spirited debate on how (or whether) government should regulate artificial intelligence.

Garrett and Peyton go back and forth on AI policy, data centers, age verification, and what it means that AI agents from OpenAI, Anthropic, and Meta have all recently "escaped" test environments and hacked into other companies. Then Garrett dives into the Paxton-Talarico Senate race and what shifting prediction markets actually tell us, two major Second Amendment rulings out of Texas courts, and a new legal fight over the state's Ten Commandments classroom law.

**Topics discussed:**
- Inside the Alliance for Secure AI's mission and policy priorities
- Lessons from the "Wild West" of Internet 1.0 and 2.0 applied to AI
- Age verification laws, app stores, and the fight over online anonymity
- Whether AI is becoming a partisan wedge issue heading into 2026
- Data centers, the power grid, and fears the issue becomes "climate change 2.0"
- AI agents going rogue: the OpenAI, Anthropic, and Meta cybersecurity incidents
- Betting markets (Kalshi, Polymarket) show a dead-even Paxton vs. Talarico race
- A federal judge guts National Firearms Act registration for suppressors
- Ken Paxton's latest courtroom loss over the State Fair of Texas gun ban
- A new legal challenge to Texas' Ten Commandments classroom law

**Guest:** Peyton Hornberger, Alliance for Secure AI (secureainow.org) — find her on X/Twitter at @peytonhb_

**Sources:**
- *The Texas Tribune* — By Gabby Birenbaum — July 28, 2026 — [Talarico leads Paxton by 5 points in new Texas Senate poll](https://www.texastribune.org/2026/07/28/texas-senate-poll-james-talarico-ken-paxton-july-2026/)
- *Houston Public Media* — By Andrew Schneider — August 4, 2026 — [Another poll shows James Talarico leading Ken Paxton in their race for the U.S. Senate](https://www.houstonpublicmedia.org/articles/news/politics/election-2026/2026/08/04/558614/talarico-paxton-poll-senate-texas-southern-university-black-voters/)
- *Fox 7 Austin* — August 7, 2026 — [New Texas polls show a tight Senate race between Talarico and Paxton](https://www.fox7austin.com/news/talarico-paxton-senate-race-polling-august)
- *Associated Press (via Washington Post)* — August 5, 2026 — [Texas judge rules that law regulating firearm suppressors and some guns can't be enforced](https://www.washingtonpost.com/national/2026/08/05/gun-law-firearms-rifle-silencer-texas-judge/61fecd62-9127-11f1-9fdc-0a725c989a7b_story.html)
- *The Texas Tribune* — By Alex Nguyen — August 6, 2026 — [Appeals court rejects Attorney General Ken Paxton's effort to overturn Texas State Fair gun ban](https://www.texastribune.org/2026/08/06/texas-state-fair-dallas-gun-ban-paxton-appeals-court-ruling/)
- *Click2Houston/KPRC* — By Christian Hudspeth — August 7, 2026 — [Houston moms lose first round in fight over Ten Commandments in Texas classrooms](https://www.click2houston.com/news/local/2026/08/07/ten-commandments-classroom-mandate-lawsuit-heads-to-injunction-hearing-after-judge-denies-emergency-order/)
- *CNN Business* — July 22, 2026 — [An OpenAI test model escaped and broke into a real company's servers](https://www.cnn.com/2026/07/22/tech/openai-hugging-face-ai-cybersecurity)
- *Axios* — July 30, 2026 — [Anthropic says three Claude models reached real-world systems during cyber tests](https://www.axios.com/2026/07/30/anthropic-mythos-security-testing)
- *CNN Business* — August 5, 2026 — [An AI model from Meta also hacked another company during testing](https://www.cnn.com/2026/08/05/tech/meta-ai-hacking)

Chapters
0:00 Cold Open: This Week's Rundown
0:48 Guest: Peyton Hornberger Joins
2:44 AI vs. The Internet 1.0 Playbook
11:30 Age Verification & App Stores
14:27 How Peyton Actually Uses AI
17:29 Is AI Becoming a Voting Issue?
26:44 Data Centers: The Left's Next Cause?
32:40 AI Agents Going Rogue
37:47 Senate Race Betting Markets Flip
42:40 Gun Law Gutted + State Fair Fight
49:27 Ten Commandments Law Challenge

Find us at seeingredpodcast.substack.com for full episodes. Follow us on all socials at @theseeingredpod and online at our website Seeing Red Podcast. x.com/gwfulce 

Transcript

SPEAKER_01 0:00

We've got a spirited discussion. We got almost almost a debate between myself and Peyton Hornberger of the Alliance for Secure AI. We're going to dive right into that in a minute. And then we're diving into the Senate race again, talking about the betting markets and how there's they're going back and forth between Talo Rico and Pakistan and kind of what's happening there and how that might influence what's happening now will influence the race all the way into no into November. Then we're going to be talking about some a couple huge Second Amendment cases that are going through the pipelines. We're going to start seeing some more news about. This week we have an incredible guest. We have Peyton Hornberger from the Alliance for Secure AI, based in the DC area. Now, this is the second time we've had somebody from the Alliance on the show. We had Brendan Steinheiser, and I might be saying his name wrong, but we had him on the show earlier this year, and we've we've talked at length with folks on different aspects of AI, data centers, whatever on the show. This is an issue I care a lot about. And the Alliance tends to disagree with me on just how to implement policy, and that's okay. And that's also why I invite them on the show, because I like to make sure that we're at least talking about this as an issue as well-rounded. So, Peyton, welcome to the show and thanks for joining me today.

SPEAKER_00 1:34

Thank you so much for having me, Garrett. I'm happy to be here.

SPEAKER_01 1:37

Well, tell me a little bit about, like just to remind folks, because I can't, you know, some people are didn't watch the previous episode, but they should, and it'll be linked in the in the description. But tell me a little bit about what the Alliance is doing and like kind of what their big goal is.

SPEAKER_00 1:54

Definitely. So the Alliance for Secure AI, our mission is to help shape the narrative around advanced AI and how that's impacting Americans, but also all of humanity. We're focused on both federal policy and is and state policy. And we are some of the first people to know what's happening and what's breaking out of Silicon Valley and what's happening in the AI space and with all the advancements and developments. And we're also some of the first to know what's happening in the DC. And as you may know and as your listeners may know, there's a lot happening. It's hard to go anywhere on the internet or in real life without hearing about AI. And our job is to make sure people are informed and also are we're guiding policy and the political landscape towards a way, towards a future that benefits all of humanity.

SPEAKER_01 2:41

Sure. Well, there's a lot of to to the whole fight. Now, I just want to uh to clarify, y'all are focused on like the the clods of the world, the chat GPTs, but how much are y'all looking at like the infrastructure behind it, say the data centers, the chips, and all that sort of part of the fight?

SPEAKER_00 2:59

Yeah, that's a really good question. I think when folks think about AI, sometimes you just think about a chat bot, just you think about chat GPT. It's a resume builder, it's something you can just you know talk to for a quick question. But we're looking at the whole landscape, literally. We're looking at, we're looking at infrastructure, we're looking at national security, we're looking at super intelligence, how you know how far fast are these machines improving? We're looking at the labs, we're looking at the companies, also automation and employment. Uh, it's very, very hard to see where AI doesn't impact every aspect of our lives. So we're we're covering all of that. There's been a lot of traction politically, and I think on a policy front, when it comes to the chat bots, deep fakes, um, and we can we can get into that. Now on the horizon is what do we do about automation and employment and how that's changing, you know, the the the landscape in America. And then also data centers. Big question, trying to figure out what do we do about, you know, where do we build them? How how big are they? How ugly are they? We're trying to figure out what to do about the data centers for sure.

SPEAKER_01 4:06

Gotcha. So, you know, for me, there I look at how the internet would have developed from the 90s into to the to the present day, and there was a lot of Wild West with it, right? There was a lot of just figuring it out. They got it, the government got out of the way, and then things started working themselves out. We had huge advancements over the years, but there, you know, there were costs to it. And so it feels like you're trying this time to kind of the the looking back at what happened with just the internet 1.0 and 2.0, I guess, you know, looking at where we're at today with AI, like what lessons are y'all applying from that what we learned from that experiment, and what are you trying to do with this with with and apply those lessons to the AI fight?

SPEAKER_00 5:01

There's in our in our space and and how we're seeing the conversations play out in DC and over in California and then you know coast to coast, a lot of comparisons are drawn between AI and social media, AI and the Industrial Revolution, and then AI and also the age of the internet. There's some hard lessons that were learned, I think, with and not to get super in the policy weeds, but there's this thing called Section 230, which regulates a lot of free speech. As a person who loves free speech and loves being able to say whatever I want, and coming from the libertarian world, that's really important and that's that's critical for you know maintaining like our civil liberties and our constitution and upholding those values. But there's a lot of questions on well, who's responsible when a chat bot you know says something that we don't like? Is that is that the responsibility of the machine or is that you know the creators? So those are some questions that we have to ask ourselves. And I think you know, some people think that we got that really wrong with social media and you know, what we were allowed to post and how that impacts kids and mental health. And there's a lot of questions there. So we're trying to take those lessons and implement them in a way that does benefit people and not just you know saying we can't do anything about it and going full hands off. I think that was definitely an acknowledgement of we can't just go hands off. And when I say we, I mean policy advocates, policy influencers, lawmakers. I think during the social media internet 2.0, there was this feeling of let's just let the companies and let's let the tech people and you know these young entrepreneurs come up with some cool ideas. And I think there were some negative consequences that requires people to be politically involved and say, no, we actually do want to shape the way that this policy and this technology leads us. So I think that that's a good comparison. I also think that the comparisons between the Industrial Revolution and AI are appropriate because we're in a really transformational period. Things will not look or be the same in you know a year from now. For example, you know, two years ago, we were not in a we thought some of the stuff was science fiction, like to be honest. We thought it was impossible that you know machines could, you know, act as AI agents, they could act on their own, you can prompt them and they can do multiple tasks that was totally outside of the realm. The movie Her from 2013, that is almost irrelevant now. That's just happening. I encourage people to watch that one. I just rewatched Space Odyssey 2001. That's kind of scary. Anyway, my my point is there's a lot to be uncertain about about. There's a lot to be, you know, people that are really worried about the future. I think that's valid. The exciting thing, the opportunity we have, and I think what we're doing here at the Alliance is saying we want to take this technology, you know, by the reins. We want to really get involved, we want states to be involved, we want you know, Congress to be involved, we want everyone to be involved to help say this is where we want technology to go, here's who we want it to benefit.

SPEAKER_01 8:13

Absolutely. I I hear you. So my read on it is a little is similar, and I in that I recognize that some of the problems that exist, and that I think I think we differ more on the solutions to the problems than on what the problems actually are, which means we're probably further along uh to agreement than most people on this issue. But I would say that a lot of what the public is talking about, or the things that the public is specifically concerned about, has some more to do with the data center side of things, but also with just an understanding of what AI was, like you said, from three years ago. It's oh, it's just I chat to it and it gives me an answer. It's you know, it's the does some of the research for me, it does the Google search for me and tells me what the answers are type stuff. You know, the original ChatGPT LLM stuff that came out a few years ago. But as you are well aware, and you talked about like we've moved into agentics stuff. There's the what the newest models are doing and capable of doing are on the coding side incredible. You can build stuff, you can build apps, websites, all sorts of things on your computer, widgets and whatnot to help with your own productivity with all on your own machine without having to go onto the open web. And just from a like you said, comparing it to the industrial revolution, but the revolution was that you were able to do get so much more output from so much less physical labor because you're using the power of machinery, whatever else. It's doing the same thing again, you can get a lot more done with like from a production standpoint, from a tasks standpoint. What and that's going to change fundamentally how the economy is shaped and works, and gives people concern because how does it impact their job? And it is an unknown, and that's the unknown, I think, is giving people uncertainty, and they just want a little bit more, you know, they just want it to happen slower. And I don't know if we have that luxury because the United States, while we're at the top of the economic food chain, we're not alone in this world. We don't just get to decide that necessarily, because there are other states, there are other actors outside of the United States that can just blow it through the wall, you know, blow it out of the water, and then we could end up being in a far different position. And that's on the national security side, but also on the enterprise side, on the consumer side. And that's where I kind of where I come back to is we can shape it. We can kind of add parameters, and then you have to figure out do you punish people for acting badly, or do you just go out of your way to prevent people from trying to act badly, but also capture all the people who are trying to do good and push the technology forward? And you have to figure out that mix, but I think there's there's going to be a push. There's going to be necessarily a push to stay competitive in the international scene. And I think that we have to make sure that we're talking about the whole picture and not just thinking about the bubble that is United States or Texas or wherever the case may be. So that's so that's where I'm at with it. But there's a bunch of stuff that is being pushed, and I know Texas passed stuff. Do y'all talk on any of the app store regulation and uh in similar legislation that went through like Texas at the state level, and that the feds are trying to preempt with like the Kids Act?

SPEAKER_00 11:44

Um, can you tell me a little bit more about that, about the app store?

SPEAKER_01 11:47

So the the there were different legislations that passed that basically said that there has to be identity verification, either at the app level or the app store level, video games, things like that, that have to verify the identity of the person to make sure that they're of age, and if they're not of age, that they have parental consent. And so there's the identity, it's de is uh in practice, it slightly de-anonymizes the internet, but also the government can see your search industries, all that fun stuff.

SPEAKER_00 12:15

Right.

SPEAKER_01 12:16

I don't know why we're in that space at all.

SPEAKER_00 12:18

We we are, and that ties in with a lot of the child safety. Child safety is is wrapped up very tightly with a lot of these AI bills and a lot of the AI policy right now. And there is a there is a big ongoing battle and conversation between federal framework, federal policy, and then what do we let the states do? I think it's a good thing that we have states like Texas and you have others that are really trying to lead in terms of how do we make sure AI is going well and that we get out of this unknown zone and we can actually make people feel more confident about either interacting with the technology or you know, moving people towards a more you know practical space. You know, California, Illinois, New York, but Texas has a great opportunity as a really fantastic state to lead here. We've seen some good movements. So there's like Traga, for example, and that is a pretty powerful framework for how AI can be interacted with. They address stuff like deepfakes as well. I think the age verification stuff is a good first step. A lot of this stuff is a great first step, and it's not going to be perfect. And I don't think the expectation, I don't think the expectation is that we're gonna get perfect policy right now, especially as AI is rapidly advancing. We're just, it's a remarkable technology and it's it's hard to keep up. But we do need people thinking about this and and trying to get closer and closer to what we see as a better outcome. But we do have folks working on that, and there's a lot of friends and allies, both in Texas and then all across the United States, that are trying to think of how do we protect children and not cause them any more harm and you know not lead to absolute disasters uh when it comes to technology for sure.

SPEAKER_01 14:01

Well, and here's my here's my question. I it's kind of a gotcha. So if you we can, you know, just just giving you a preface, it's whatever else. But how are you using AI in your everyday or work life? Like, like what is like what how are you using it? And are any of the policies you're advocating for going to make it more difficult for you to you know get done what you're trying to get done?

SPEAKER_00 14:26

Yeah, great question. I I use AI every single day. I think for well, for two reasons. One, I've I've always thought emergent technology was very interesting. I was primed at a young age to be into all of like the sci-fi stuff anyway. I use AI every day. I was an early adopter of Chat GPT, even when um it was really bad at writing emails. Now it's pretty good. Like it's really good, actually. The video, I mean, video content images are indistinguishable from reality, and that's been frightening and amazing to see at the same time. I've built my own AI agents that have helped with like repeatable tasks. Um so I've seen, you know, firsthand the productivity benefit from being able to use AI. And, you know, there's a lot you can do in terms of like the digital space, in terms of digital marketing and and in the communications world as well. You can have AI code a whole website for you. There, there's really fantastic ways to you know make yourself more productive. Also in my personal life, it's great for recipes. That's fantastic. And, you know, doing some research for different trips. So I do think that there's a lot of great uses. Um I've also seen, I've also seen how AI can really uplift up-and-coming entrepreneurs, small businesses. There's a lot of good opportunities there. Okay, so then the second part of your question is anything that we're doing or advocating for going to harm any of that or or make our lives worse. That's what I don't want. I don't want, I don't want to see policy that you know puts, first of all, any of the states in handcuffs or any of the businesses in handcuffs. I think that's a bad plan. And our our philosophy at the Alliance is how do we promote US innovation and US dominance? Because we do, we, we, we have to be dominant in the AI space, especially against China. How do we maintain that while protecting children and you know, protecting people from harm? Because the digital space does do that, all while trying to keep up while this technology rapidly, rapidly increases. So I would say no, I don't think that anything, any of the uses that we see of productivity, uplifting entrepreneurs, helping out small businesses, helping out, you know, parents who want to do it all, or people who want, you know, even like minor health advice, none of that is is at risk. I think we're really in a position where we can get the most out of the technology.

SPEAKER_01 16:55

This issue's turning into a voting issue, like all us all do. Are you seeing this issue polarized yet? I'm advice to say that is one party taking one side versus another. Is it still kind of figuring, or the coalitions still kind of both bipartisan where you have or just what are you just in Texas or across the country? But then also, how are you seeing it play in elections and whether it be primaries or on the either side of the aisle, but also shaping general election fights across the country?

SPEAKER_00 17:29

Sure. The politics on the AI issue, I think, are really good. They're almost all polling that we've seen from whether small scale to large scale, say that everyone, up to upwards of 80%, no matter your background, they're concerned about the direction of AI currently. They want something to be done, they want some answers, we want some more certainty, we want to know what's going on in the black box. That is true for all different backgrounds for people, Democrats, Republicans, teachers, students, parents, workers. There's a lot of it's that that's unanimous. Now, where it's getting politicized, I would say, you know, we don't think that there should be like a ban on all data centers. I mean, before it was wrapped up with the AI stuff. And we we've had data centers. If you want to use your smartphone, we need data centers. And we also should we also need to think about that for, you know, if we want to make sure that America is dominant, we're gonna need to, you know, think about how we can build data centers and build critical infrastructure in a way that's going to, again, benefit America, you know, without without causing any real harm. The concerns though that people have about data centers are very valid. And my my thoughts for people running for office on that topic, you should listen to those people and you should hear them out. You know, why are they concerned? What are they afraid of? Are they afraid of is it the like is it the cost? Is it the cost of living? Is it the noise, or is it, you know, something else? Is it the fact that they're just afraid of afraid of AI? I would say that going into upcoming elections, if if a candidate's not addressing the data center issue or the AI issue, they're gonna be in for a tough ride. You need to be prepared to like have some answers to some questions because people are are really, really concerned.

SPEAKER_01 19:25

Yeah, I I think I think you hit the nail on the head, the nail on the head on that one. The the there's a lot of backlash, and it seems bipartisan. There's rural backlash, there's whatever there's organized fights on the east coast. There's a lot of misinformation coming out about like what the problems are. Like, there are a handful that are making noise, but that's from the energy generation that was required. So, like, for uh for me, it's like if you require you have to bring your own power, which is one of the things that people are saying. Well, then they're saying, but we don't like we don't want it to be loud. It's like, okay, well, pick one, and like I feel like the noise is gonna win out because my here's my issue. And I and I've said this before, I'll keep saying it. The data centers need the power. We need the data centers, but the reason they don't have access to the power is because governments across the country have not been building their grids appropriately. They've been just trying to manage it with like and been scared by climate alarmists about like not of having too much electricity or whatever, and then afraid of nuclear. Like, we've have power needs. We have the power needs regardless of whether it's AI or not, or data centers or not. Well, it's AI, it's data centers that's really pushing this and it's really forcing this. We need to expand the our grid capacity just across the board generally. And so, yes, I I hear you, like people are concerned about like whether we're gonna have enough power. It's like, yeah, but we need more power. So we just need we need to that that's a it's a similar problem, it's a it's an aligned problem. They are they definitely whatever else, but it's not like it's not AI's fault. We need the power of workless, and we need to upgrade what we currently have. That's one of the other problems. Like it's like we have aging infrastructure, we have to re- you know, make our grid more resilient.

SPEAKER_00 21:23

I I actually really agree with you, Garrett, on this point. And I think a year ago we really couldn't have a conversation about this. People weren't really ready to have the, you know. Actually, let me back up. Nuclear is a great, a great point. Every that we we weren't talking about that. We know it's like one of the cleanest energy sources, it could also be super profitable. It could be there, there's so much potential to come out of nuclear, but everyone got very scared and we did nothing else with it. And we know that if we want to keep building AI and we want we need more power, we're gonna have to get very creative on how. We build good technology and infrastructure to power all of this in a way that is reasonable for communities. And I actually I tend to be an optimist generally. I do think that even though there's this backlash, even though there's this big, big push that we're seeing nationwide, that signal I think is going to end up leading to like better policy, better infrastructure that does, you know, hopefully ultimately like lead to like more productivity, people are happier, we're getting like better, you know, yeah, better infrastructure and updated infrastructure. I think that's where we're headed. But I I wanted to go back to one point. One of one of one of the things that we're doing and why we really try to talk to a lot of different people at the Alliance, and we're doing a lot of different, you know, communications and events and panels and bringing together different coalitions is so that we can give people the courage to talk about AI. I think there's sometimes folks think that because it can be so techno technical, or because it's so, you know, out of their realm of imagining, they can't, you know, have have an opinion, but they really can. And I think that we don't want folks to tap out. We want people to be involved in this process and say, like, no, we we do think that, you know, the data centers are allowed, that that's valid. Okay, how can we get better, you know, like we need we need the power. Let's let's have someone very creative and clever and have these engineers build something. So I mean I I think people should have courage to, you know, get involved and say, you know, learn more and you know, help shape this because it it really does like require all of us.

SPEAKER_01 23:44

Well, and I and here's my thing. I have many things, but here's one of my things. If you look at a similar technology in that it's a backbone, that the real stuff is what's carried on the data, it's not the data itself or the data center itself. You look at something like the railroad, which is also extraordinarily loud and it has to is necessarily next to people, like yeah, we have some quiet hours sometimes, but but that doesn't mean but that means they just can't blow their horns. The trains themselves are still loud. I can tell you, because I live a mile away from one, and at nighttime I can hear it going around the clock down the tracks. A mile away in a major suburb in a city, you know, like that's it's just loud. They're just loud, and you get used to it. So, like, yeah, with data centers, like we have some uh some agency and like where we can move something that's you know, depending on where it is, if you're in a city or if you're near housing, like we can make it a little bit further away, we can make it less loud, you can put decibel lemps on on all sorts of stuff. I'm not against any of those policies, but I'm what I'm saying is like it doesn't have to be perfect, and we shouldn't require perfection. Like they're still trying to figure it out, and that's my one of my just annoyances because I here's my here's my fear is that the people who are organizing the data center fight are the same people who organize the all the climate fights to kill which were basically just designed to kill all development, kill any pipeline, kill any nuclear power plant, whatever else. And they're using the same playbook. And right now, Republican and Democrat governors and administrations are giving in to the first steps of it, and then they're just gonna clog up the works, which is how things work in California. Well, all their NIMBY stuff. It's take that's why development's so difficult there. Like eventually there has to be some sort of breakthrough, and the Republican states will probably be the ones that lead it, but like my my fear, my concern is that that's about the playbook that's being deployed. I'm seeing it in action, and I don't want to make sure that the people on the right, center right, far right, whatever, are looking at this and think this is going to be a tool of the left. The left is going to make this, this is gonna be like climate change all over.

SPEAKER_00 25:54

Interesting. I want to push back on this a little bit. I I see where you're coming from, although I would point to a couple of different groups out there that are very openly like right of center and also you know, activating people on the data center issue, just in terms of whether it's like eminent domain or whether it's just you know what like can't you know, questioning like what the government can do. I I I think if you go and and look at groups like, for example, like Humans First, they've been doing a great job of organizing you know, people on the ground. And that is a that was born out of like the Tea Party movement. So there's a lot of groups out there that are bipartisan, they're just like these rural people that are have no interest in you know buddying up with you know climate activists, but here they are, here they are like standing side by side because they have they have some concerns. I do think though, you know, expanding a little bit out of just you know the data center issue, there's a greater concern. I think the data center issue for some people is a physical manifestation of what they're actually worried about, which is the uncertainty around AI. And that is a cross-cutting issue. That's not, you know, that that's right, left, center, that's you know, everywhere. I think the real worry is we don't know what's happening with this technology. We're being told by the companies that they don't understand how this technology works. And yes, it's fantastic right now. It's giving us these, you know, awesome productivity tools. But just this past week, the top three labs had AI agents go rogue and escape and break into other companies. And, you know, that is just now starting to hit the mainstream. But those are the real concerns I think that people are have of what if we can't control this technology? What if, yeah, what if what if humans are no longer in control? And what does that what does that mean for us? What does that mean for our future?

SPEAKER_01 27:52

I mean, yeah, I I I hear you. And on a couple of things to note. One is to my to your point that there are people on the right who are or who are on this, I think you're right. Right now it's a fairly bipartisan issue. I'm just saying I think eventually it's going to coalesce as one of those like kind of issues that'll and that'll be driven by the left because it's ultimately going to be anti-development, and that's where that one usually falls on the on the political spectrum. In the same sense that environmental concerns started the same way. It was all about clean water and you know, whatever in broad stroke stuff. That's why Nixon pushed it through in the 70s, those those policies through in the 70s, because there was a long history on the on the right of conservation versus environmentalism. And that's kind of a much deeper conversation for you know a different different show. But my my point is that right now you're gonna have those folks who on specific uh elements are going to be aligned in kind of this weird Frankenstein's like or weird uh amalgamation politically pushing it back against something. And right now it's very popular, so think people on politicians on both sides are pushing back. I'm just saying that I'm seeing the playbook from the left, it's going to that that they've used in the over the past few decades, and that this is I'm about as certain as I can be that that's how this issue is going to fall on the data center stuff specifically. But that once again, that's just opinion, and that's my prognostication. I'm not asserting that that's you know, as a fact. I'm I think it's a fact, it's definitely an opinion. The second thing you noted was the stuff that got out, I think it was either from Mythos or from Fable, or whatever they called it before they were labeled as Fable from Anthropic. But those were things that were done by Claude and the Anthropic team. And I think it's interesting is that it wasn't those weren't those jailbreaks where they got out and broke into other people's sites. They found it's I'm less concerned about that. If for only a reason is it will make I think it'll make I think I think it'll make cybersecurity safer because we you will you can have the White House doing the stuff to find the stuff first, right? So I think that what what because what happened was Anthropic did it, it got out, they realized they got in, they immediately, you know, did what they had to do on the security standpoint and alerted those companies to whatever happened, had them patch it, and they get patched it and fixed it. So now that that um for those companies can't happen again, but depending on what sort of exploit they find it found and how, you can then deploy those things to all sorts of other enterprise folks, like that community talks to itself to make sure that everyone gets patched up. And so I hear you, and I think that it's there's a wisdom in keeping it stateside right now where only US citizens can use those tools and you know, putting restrictions on how much they can be used, and there's all sorts of guardrails they put into it now for the people outside of the environments. And from my understanding, is they're pretty they're pretty tightly tuned in the sense that like they're gonna get a lot more false positives than the from people trying to go out and be elite hackers or whatever. But I mean it's a serious concern, but there's the serious concern also relies on the fact that there are models in other countries that are being deployed, and we need to know what those vulnerabilities are as those models get better and better, too, because they're gonna be used against us and our businesses and our governments. We want to make sure that we're good. So I'm I I hear where you're coming from with it, but that's kind of where I'm landing with a different view of how those issues are processing.

SPEAKER_00 31:39

But you're hopeful, you're hopeful about humans keeping up.

SPEAKER_01 31:43

You think that we'll yes, and well, and and more and more I'm I'm also thinking there was the problem is as much that that a Claude agent found the vulnerability is that the vulnerability existed in the first place, you know, like it's just because we have a tool that can find it, and no one said, like, oh, no one would have ever thought this in a thousand years. Like, it didn't take them very long to find it. So, like, what we so what I'm hearing is, oh, we can use this as a cybersecurity thing. There's only so many white hat hackers out there, white hat being like people who do it, like where they're paid to find vulnerabilities and let people know so they can patch it. Like, okay, now we can test these sites, these systems in different environments, in live environments where we have people attacking them and that we're able to find those things and patch those issues as they happen. That's a whole other realm of cybersecurity that we haven't even got to that might, in my opinion, depending on how it's in, you know, I'm not on the data science thing, but I think it might make things more secure long term.

SPEAKER_00 32:47

You know, I again I really appreciate your optimism. I I I do I know that the response to some of these was, you know, was was measured and in and quick. I one of the one of the incidents I'm referring to was with OpenAI's models after the mythos, mythos was a couple months ago, and one of OpenAI's models on its own hacked into a company called Hugging Face. And Hugging Face, this other company had no idea what was going on, so they reported, they reported that hack as a crime. And then it wasn't until you know a few days after that they figured out what happened. And then Anthropic subsequently also said that a similar thing happened. And now Meta also yesterday said that this they had a similar situation. And it is a good signal that you know the very, very smart people that are working on this like are able to cat like catch it. To me, as you to me, what I'm saying is, well, that wasn't part of the plan. We didn't, we like they weren't planning on having these models go and act autonomously. And that's something that we've been particularly worried about is what happens. Okay, we were able to control it after you know four days, but as the technology, again, that that that curve, that exponential curve, as it keeps getting more powerful. What if what if it's, you know, instead of four days ahead of us, what if it's four weeks or four months or a year? And there's some serious risks there. But equally, I mean, to your point, like, good reason to have increased cybersecurity, good reason to make sure that like our you know adversaries are not, you know, that we need to be keeping a good eye on how improved their models are, because that could, I mean, that could do some serious damage like to our to our digital space. So the the fact that that's happening, again, you know, was truly like science, truly like you could not have a serious conversation about that possibility with someone a year ago, and now it's like very, you know, realistic. I'm also hopeful about humanity. I think that we're all here because we're trying to make sure it like goes well. But the fact that AI is now powerful enough to act on its own should be, you know, right. That should be ringing in people's in people's minds, like, okay, it's time to pay attention. You know, what are we really building here? What's really happening?

SPEAKER_01 35:14

Well, where could thank you so much for coming on? I want to make sure people are able to find you and find the alliance online. Where where can people find you?

SPEAKER_00 35:23

Absolutely. Well, thank you, and thank you so much for having me. It's uh it's always good to talk about this stuff and you know, talk to Texas. You know, love that place. Everyone can go to secureainow.org and you guys can learn more about what we're up to, both our policy priorities. Also, we have a really robust action action center for different folks who are wanting to get involved on a community level, on a state level. Uh, we have a lot of different ways for people to engage, get educated, you know, get involved in the fight to make sure that AI goes well for humanity.

SPEAKER_01 35:56

Well, thanks again for joining us, Peyton. Did you tell people where they could find you?

SPEAKER_00 36:01

Oh, me? Yes. Okay, you can find you can find me on Twitter. It's at PeytonMhorn underscore. But yeah, all my work, all my work is for the Alliance trying to make sure this goes well.

SPEAKER_01 36:14

Okay. Well, thank you for coming on and for you know, having this spirited conversation with me. And I look forward to talking with you more as this goes along because I'm sure we'll be getting into fights in Texas, and I want to have someone come on and give a good defense of things I'm probably against. So thank you for joining us.

SPEAKER_00 36:31

Absolutely. Thank you so much for having me.

SPEAKER_01 36:37

Well, I want to say thanks again to Peyton for coming onto the show, having a fun little adversarial conversation on certain points on just you know, talking about the difference of approach to how we're going to regulate and keep AI, you know, from taking over the world, I guess. But now we're gonna dive into the Senate race. And let's talk a little bit about betting markets. So there's Calci, there's polymarket, there's a couple. And for those of you who for some reason don't know, but you probably do, they are prediction markets. They basically allow you to predict on the outcome of certain events. You see it a lot with in the sports, but also on political matters and the big sites, they have all the big Senate races and so on and so forth about who people think are going to win. And for the most part, Paxson has led Talo Rico in that in those markets where people are putting their money on who's going to win. But another string of bad polls came out this last week or past couple weeks that have consistently showed Talo Rico with from like a one-point to a four-point lead. And so the betting markets are taking that into account. And now, depending on when you look at it, Talo Rico or Paxton might be the one who has the quote unquote lead, which basically means it's a coin flip right now. That's interesting. So if you're the type that says Republicans can never lose in Texas, can never lose in Texas, you might want to put your money where your mouth is. But a lot of people on both sides are putting their money in different directions. That's an interesting thing to keep an eye on. It's not necessarily the end-all be-all, right? Because it's predicting a result, not a vote margin. And so it's not as if, you know, one person has a one-point or a two-point lead. So it's not you can't look at it like it's polling, but it is an interesting data point of how people feel about the race, not necessarily of how the race is going. But it's taking into account a bunch of things, like the polling numbers, who what people are seeing on TV, what people are saying around their friends, their family. There's some political science literature that seems to indicate that a better polling question than who are you going to vote for is who do you think the people in your neighborhood or in your life are going to vote for? That tends to get a closer answer to final voting results than what people actually say about themselves for a couple of reasons. One, people, some people might lie about who they're voting for or not voting for on a poll. They're probably not going to lie about what they think other people are doing. But also, it just brings in a bunch more information into that question. Now, we haven't seen that play out in a lot of different polling scenarios. So who knows where that's gonna how valid that may be. But kind of the wisdom of the crowd's thinking, which is all kind of polling, any opinion polling really is, is also that can bring in more variables, is going to be more valuable typically. And a lot of what people look to for at the betting markets is is that is that phenomenon, basically. So that's something to keep in mind right now. Some other things that have been happening just on the on the race as a whole. Governor Abbott has announced a $10 million ad buy that I think has already gone on air as of recording this. Ken Paxton has gone on air with a Spanish language ad. I saw the news alert right before I start recording, so I haven't gotten to go watch it yet. I'm in Tal Rico still going up on air. So we're really now in it, right? We're into the push. We're in the last 85, 84 days until election. The real, like the sprint starts happening after Labor Day, which is only in a few weeks. So we're here. This is where the meat of everything is starting. Everyone starts paying attention. So if you're just tuning in, you kind of getting a say of the race, it's probably 50-50, can go either way for on the Senate race. Everyone, top to down on the Republican side, is getting behind the ticket as a whole, and they're pushing themselves forward. And they're trying to combat kind of the the drag that being the party in office in the presidency has during the midterm elections, regardless of party, regardless of president. It's just a phenomenon that's happened in this country for the last 150 years. So they're doing what they can, and we're we're we're gonna be keeping an eye on it going forward. Now we're kind of now let's we're gonna talk about a couple of other big issues that are going to kind of echo in this race a little bit, but are just more interesting. Come they're just more interesting when you're if you're a follower of Second Amendment kind of law and what's happening, right? There are two big stories that are two big cases that are happening right now. Uh Judge Hendricks in the Northern District of Texas just gutted the National Firearms Act. Basically, what happened during the one big beautiful bill of last year was that they dropped the taxing authority from some of the from some of the elements of the National Firearms Act from like 1934 or something down to zero. This impacts things like suppressors, short barrel rifles, things like of that nature that had a tax on it that you had to pay in order to transfer or do certain things. Well, that had been ruled an eligible like restriction or uh a thing on the Second Amendment because it was an adding a tax to something. But the moment when the one big beautiful bill lowered that number from like 200 bucks down to zero dollars, the judge is now ruling that that restriction, the other restrictions or are not that went along with that tax are no longer viable either, because it was only put into the law via the tax code. It's a really nuanced legal argument. It's one that Second Amendment amendment folks have been talking about since the One Big Beautiful bill passed. And it's the first kind of test case that says, hey, this is a valid legal argument. Obviously, this is going to be appealed by different folks on the gun rights versus gun control debate, and we're only at the first rung, but it's a definitely a victory for the Second Amendment. Now, on the flip side is another area that Ken Paxton has been pushing the state fair of Texas on. Now, the State Fair of Texas is a nonprofit organization. They house the state fair on public property at Fair Park in the Dallas area. And the Attorney General's argument was that despite the nonprofit running the fair, since it was public property, the rules in place that bar the government from restricting the Second Amendment, like uh whether concealed or open carry, uh, would transfer to the park because they're using the safe facilities. I've always had a trouble with that legal argument because they're technically a lessee. They are leasing the park, and that is available to a bunch of other, you know, it's available to other people through a bidding process or whatever. So the it is temporarily converting a public space validly into a private one. And as a private person, you retain the right to prohibit firearms or people. Or carriers of firearms from coming onto your property. You mean that's you know, that's this kind of basic property rights 101. If you have an apartment or if you have a house or you have land and you don't want guns on it for some reason, you can say you don't want guns on it. Now, there are arguments that it's a bad policy because it makes the fair less safe, that there are people getting through with firearms and there have been incidents over the past few years. And you know, there's an argument to be made there. But we're talking about a legal argument, and thus far, all the way up to the different channels to the I think the the appeals court level, uh, the attendor general attorney general has lost. If see if one of the courts of last resort in the state take take on the take on the case. But for now, the fair can keep restricting, you know, barring guns from coming into the fair, and that's gonna make news again because the fair is coming around on September 25th, right? In the middle of the run-up to the election. So Ken Paxson has to make a decision on whether he how he's going to pursue this because it's going to be a topic of news or whatever else. And on one hand, you can kind of try and gin up 2A supporters to come out and vote for you because you're you know trying to you know impact their ability to carry to the state fair, but maybe there's more to lose among persuadable voters. And that's an open question that the his can he and his campaign are going to have to decide on, but it's something to be looking for when and looking in that lens when you're looking at this issue in particular. And so this isn't necessarily an issue that's going to drive a lot of media coverage, it could drive enough media coverage that it opens up the gun issue just as an issue, because there's pretty definitive sides one way or the other when it comes to guns. And Democrats in Texas typically don't want to talk about the gun issue, whatever it may be. This is an interesting case where the facts at issue are less about guns and more about it's less about the guns. It's more about what rights does a nonprofit or for-profit company have or a person have when they lease government property or rent or when rent government property. So if you're at a park and you're having a birthday party and you rent the gazebo at your local park, are you able to restrict access to that? And if you're able to restrict access, can you do so on the basis of whether someone has a firearm or not? And so if and if someone otherwise would have access to it, are you able to restrict it? Those all those questions are in play, and it becomes a point where unless there's a carve-out for it in law, but that might be it need a legislative fix to actually fix it because the law itself is generally going to protect the person's rights to it. But that's the map, that's the legal issue. The political issue is who's for guns at the state fair. Some die-hard Republicans and Suei people are going to be for it, but a lot of uncertain suburban parents will probably be against it. And those are people you're trying to persuade to vote for you. That turns the gun issue and maybe a liability for Republicans, or for at least for Ken Paxton going into the election, because people are talking about it in the news, depending on what Paxson does on this case. He can either just lose and not talk about it, and it doesn't become an issue, or, or except a little bit right now, or he can file his appeals, which he necessarily needs to do soon, as the state fair is starting and people are thinking about it, and it becomes an issue, it becomes a political issue, a hot button going into November. And so it's just something to keep your eye on going into the election of this is something that could flare up and what the actual issues are in the case. But there are other cases happening right now that Kim Pax is going to have to defend stuff that he's won. There's been the last in the 89th session, the legislature uh passed a requirement that school districts would have to display the Ten Commandments if people donated them to the to the camp, to the school or to the district. And so schools all over the state have the Ten Commandments posted in their cafeterias and in their classrooms and all over the place. And in a way that's specific specifically prescribed of how they have to look. And so that's those have those laws have been tested up to the, I believe, the fifth court of appeals. I don't believe there's a circuit split yet, but eventually that'll go to the Supreme Court. We'll see how that how that plays at that level. But more than the, I think Texas, Arkansas, Alabama, and Louisiana have all have all passed these laws so far. And so eventually that law will end up on the Supreme Court from the rule on what is and isn't allowed on this. But in the meantime, there's a new case that came out last week from the Houston area that isn't challenging it on federal grounds. It's challenging it under state law. A couple of laws passed recently, including the parental bill of rights law passed in the last session. It's it's interesting for a few reasons. One is we'll see if that's effective and how what the meat of that other law. But that law was meant for conservatives. It was written by conservatives to kind of help schools or to help give parents some teeth when fighting against schools and against other fighting for their kids and making sure they have the you know parental authority. So I'm not ex I'm not an attorney, so I'm not sure the merits of this particular case, but it is an interesting angle to keep an eye on. Right now, the government has fought off a temporary restraining order, which would have uh made it so that the districts in question where the parents are suing from are they don't have to, they wouldn't, they would have to take down the the Ten Commandment stuff while the case is pending. That's a pretty easy bar to clear. So it doesn't tell us much in the tea leaves at this point, but tells us something. Tells us that the case the government has a a pretty strong case to that they're likely to win. But it's we'll it's interesting to see, and it's something that we'll we'll keep moving along. It'll move slowly. It probably won't impact the verdict won't impact descendant race, but talking about it and getting it into the bloodstream of the media and what people are talking about, it could be one of those issues that flares up. And what right now, 85, 83, whatever days, however many days we are, however many days out we are, what you're looking for are those things that are going to spark, those stories that are gonna spark into a wildfire, right? Because it's gonna be something right now. People are called probably bet it's gonna be some sort of data center-related thing. It's probably gonna be some sort of national story. And there are a few things that have started up that we'll we'll see. Maybe it's I don't want to speculate because a lot of times it comes from small tragedies that become major issues. So, but I don't know how many major like health crises like COVID, but like also Ebola from 2024, ISIS from 20, sorry, 2014, ISIS or 2012, whatever. But ISIS from 2014, these different issues become like the issue everyone's talking about across the country. And it's hard to say what that issue is going to be, but it's kind of starting to look for the things that it could be, and those you can have local issues like that too. So that's what I'm looking for right now, trying to keep you up to date with what could be happening while also telling you about some news you may not have known about. And that's what we're doing here at Sing Red. We're talking about all the things as best we can and trying to tell you about as it is. So if you like that, subscribe where you're watching, where you're listening, so you can hear us week to week. And hopefully, as you if you do that, you'll see me pop up in your in your feed next week and the week after that. And then I can say hello to you. But for now, I'm gonna say goodbye. Have a good one. This is seeing red.