Zuckerberg Shocks Meta, Anthropic AI Panic, FBI Reporter Probe

TJ Walker warns that Anthropic's new Mythos model could exploit banking and power-grid vulnerabilities like a weapon, criticizes Meta for harvesting employee data to train AI, and discusses the FBI's investigation into a New York Times reporter who covered Kash Patel.
Chapters
Ultimize. Hi, I'm TJ Walker. This is Filtered, and today's a special day. It is Take Your Son or Daughter to Work Today, and I'm joined by our star reporter, analyst, and my daughter, Rossa Walker. Thanks for joining us today. And is it nice to be on your, your favorite show of all time? Um, totally. Okay, that's good. We're gonna dive right into the top stories As always, and you can sit a little closer so people can see you. According to the New York Times, the artificial intelligence company Anthropic has developed a model called Mythos, so powerful that it is restricting access, warning that it can identify and exploit vulnerabilities in critical infrastructure, including banks, power grids, and government systems. Officials have taken that seriously. The governor of the Bank of England warned the model could, quote, crack the whole cyber risk world open while governments across Europe and beyond began urgently assessing their defenses. Now, that's the setup. That sounds kind of boring so far, right? Why, why should anyone really care about this? Well, let's talk about why people in very expensive suits suddenly look like they've seen a ghost, because this is not just about better technology. This is about breaking the systems that make modern life possible. And the easiest way to understand that is not with code, but with your bank account. You have a bank account, right? I don't. I thought— yes, you do. You— we just don't give you access to it. Imagine you wake up tomorrow morning, you open your bank app, you're expecting to see $5,000, $10,000, enough money for rent, mortgages, grocery, your life, and instead it says Zero. Gone. Now, if that happens to you alone, well, that's a nightmare for you. You call the bank, you scream, you sit on hold for 3 hours listening to some bad jazz music, but eventually maybe they fix it. That's one person's problem. Now imagine it happens to everyone in your city, everyone in your state, then your entire country, then globally, every bank account, every balance, every record. Suddenly nobody knows what everyone has, who owns what, who owes what, who's solvent, who's bankrupt. The answer is nobody knows. That's not an inconvenience, that's the end of the financial system, because the financial system is not money, it's trust. It's the shared belief that when your screen says you have $10,000 you actually have $10,000. Take that belief away and everything stops. Businesses don't know it if they've been paid. Employees don't know if they'll ever be paid. Banks don't know what they owe. Governments don't know what to guarantee. Credit freezes. Transactions stop. Supply chains break. And very quickly, you move from my app is glitching to we can't get food into cities. Because trucks don't move without payments, payments don't move without banks, banks don't function without data. So yeah, within days, maybe hours in some places, you start seeing empty shelves, then panic, then real hardship. And that's just the banking system. Now layer in power grids because this same category of artificial intelligence isn't just looking at financial systems, It could look at infrastructure, electricity, communications, transportation. Imagine the same thing happens there. Lights go out, not in one building, but across regions. Communication networks fail. Emergency systems slow down or stop. Now you don't have an economic crisis, you have a societal one. And here's the darkly funny part, because there's always a dark funny spot somewhere. For years, we've all joked about technology failing. Oh, the Wi-Fi is down. Civilization is over. You wouldn't like it if Wi-Fi went down, would you? Well, it turns out that was less of a joke than we thought, because modern civilization is essentially a stack of systems that all assume the layer below them works. Take out enough of those layers, and suddenly we're all very nostalgic for paper maps and cash. That's the big problem. Now, to be clear, this hasn't happened yet. It's not happening tomorrow, but this is not run for the hills. This is why people are taking it seriously, because what Anthropic has built and what others are racing to build is not just a tool that does tasks. It's a tool that finds weaknesses systemically at scale faster than any human team. And once you find weaknesses, You can exploit them. Now let's go back to this analogy. For decades, the most dangerous technologies like nuclear weapons required enormous infrastructure, money, materials. You had to have enriched uranium, not something you can just conjure up or get from Amazon. You needed materials, scientists, time. That created kind of a natural barrier. Not everyone could do it. Now imagine if that barrier just disappears. Imagine if the most powerful disruptive capabilities in the world don't require a government, don't require billions of dollars, don't require years of development. Just intelligence, just code, just access. That's the shift. And that's why governments are nervous because this is not just about has the biggest military. It's about who has the smartest systems, or worse, who steals them or leaks them or misuses them. And here's where the gallows humor kicks in, because we're now in a situation where the most responsible actor in the story is the company saying, hmm, we built something terrifying and we're trying to be careful. That's the good news. The bad news They also said others could reach similar capabilities within about 18 months. 18 months, that's not a distant future. That's basically one or two iPhone upgrades away. So now you have a race, not just between countries, between companies, between actors who may not all be careful. And there are no global rules, no treaties, no shared understanding of limits, just hope. Which historically is not a great security strategy. Now bring it back to the human level because it's easy to think this is abstract. Well, it's not. It's your paycheck, your savings, your electricity, your ability to buy food, your ability to function in a system that works. And if that system is disrupted at scale, you don't get a warning, you get confusion. Then panic, then consequences, very real ones. So yeah, the story is being treated like a big deal because it is. It's not hype. It's not just another tech headline. It's a glimpse into a world where the most powerful tools are no longer just productive. They're potentially disruptive at a level we've never seen before. And the final slightly uncomfortable thought is this: For most of human history, power came from physical things— land, weapons, resources. Now it increasingly comes from something invisible— knowledge, algorithms, intelligence— which means the next global crisis might not start with something you can see. It might start with something you don't notice at all until you check your bank account and it says zero. So I'm joined today by my daughter, Rossa. It is Take Your Son or Daughter to Work Today. I'm hoping this problem gets fixed because we can't just rely on CEOs to act responsibly. Lean in a little so folks can see you, those not listening on the podcast. If you're listening on the podcast, you can check us out on the YouTube show and you can see it. Rossa, who, by the way, is a star in her own right. She has been the instructor of several online instruction courses on the Udemy platform—public speaking, how to create digital art, and others. So, and she's also an accomplished artist. So, Rossa, you may grow up one day and sell a painting or sculpture for $1 million someone pays you, but if artificial intelligence is so good that someone can just crack your account like that, and you get $0. You might have to come back and live with mommy and daddy even when you're 30 years old. Does that sound like fun? No. That's— she said no. If you didn't hear that, that's why this is such a serious issue. More in a moment. Let me just throw this out there for a second. If you've been listening to this show for a while, you already have more information than most people ever act on. That's not the problem. The problem is nothing is reinforcing it tomorrow or the next day. Or the next day. That's why I built Ultimize, to give you something you can use repeatedly, not just hear once and forget. So if this is resonating at all, just go to ultimize.com, grab the free course, communication skills, habit building, whichever one you want, and actually start putting this stuff into practice. It's free, no downside. Just go to ultimize.com. Okay, according to Report from the Information, Meta Platforms is expanding the use of internal employee data, including coding activity, internal communication, and workflow patterns to train its artificial intelligence systems with the explicit goal of automating more tasks currently performed by human workers. The report cites internal documents and employees who say Meta is focused on building AI tools that can replicate the day-to-day functions of engineers and other knowledge workers. Effectively reducing the need for certain roles over time. In other words, this is not just AI helping employees, this is AI studying employees. Now let's slow this down because this is where it gets a little unsettling. If you're a Meta employee, your daily work, every line of code, every decision, every process is potentially being captured and used as training data. You're not just doing your job, you're documenting your job for a machine. Now, that may sound a little creepy, but I'm joined today by my daughter because it's Take Your Daughter and Son to Work Day. Rossa, right now, you know, because I give you access to a cell phone, your mom and dad can track that with parental controls. Well, you're 13, but imagine if you are 30 or 40 and someone is tracking every single thing you do all day long. Do you think you'd enjoy that at that point? You don't even enjoy it now, right? Okay. So that's what's troubling so many people. So it may eventually do it faster, cheaper, and without asking for stock options if the AI can do everything the Meta employees do now. Now, to be fair, there is a positive spin and it's not fake. This could eliminate repetitive work. It could make companies dramatically more efficient. I mean, I don't think there's any doubt it will do that. It could allow top performers to focus on strategy instead of execution. That's the optimistic version. That's the version that Mark Zuckerberg is selling. But the employee version, you show up Monday morning, you open your laptop, and you quietly begin training your replacement so you can get fired. And this is happening at scale. I don't want to overreact here. But if it worked there, I might occasionally open up a few highly questionable browser tabs because I don't want to lose my job. I might want to mess up the AI just to make sure the AI doesn't become too efficient. Maybe employees start looking at Minesweeper a little, maybe a deep dive into why pigeons might be running the global economy just to throw off the algorithm a bit, because otherwise What are we really doing here? We're building a system that learns from us and then outgrows us. And yes, it's always fun to take a shot at Mark Zuckerberg. He does occasionally give off strong, "I've already uploaded my consciousness to the cloud" vibes. But here's the more important point. This isn't one quirky CEO. This is the entire business model of modern tech. These companies are legally and financially obligated to increase efficiency and reduce cost. And if you are an employee, I hate to tell you, you are a cost. Perhaps a valuable cost, maybe a talented cost, a cost with excellent slack etiquette, but still a cost. So no, I'm not saying wake up in a full panic every morning. That's exhausting. But I am suggesting you wake up with just enough awareness to ask, What am I doing that cannot be easily turned into training data? Where's my leverage? What do I actually own? Because the safest people in this new world are not the ones who follow systems perfectly. They're the ones who create systems or own systems or operate in areas where judgment, creativity, and trust matter more than efficiency. And here's the twist: the same AI tools being used to potentially replace you are also available to you to build your own business, your own audience, your own intellectual property. So this isn't a complete doom story, it's a wake-up story, because if you're quietly doing your job and doing it really well, you might be doing it so well that the machine is already halfway to replacing you. Okay. I'm not saying in this case, Zuckerberg is some evil monster. His investors are demanding that he be more efficient, cut costs, and make more money. But it is troubling for anyone who doesn't own, I would say, at least 20% of a big tech company. If you work for a tech company, I'm here joined today by my daughter. Since it's Take Your Daughters and Son to Work Day, I said, do you want to work for a big company and have bosses and telling you what to do? Or do you want to work for your own company when you grow up? What do you think? If it's high pay. If it's high pay, you would. Okay. That is, in fact, exactly why a lot of people like working for big companies. You can get the high pay benefits without having to do all the stuff of a startup and the risk associated with it flopping or not raising money. But what happens if you get used to that high pay and all of a sudden they tell you by a text, hey, we don't need you anymore because an AI is doing your job. How would that make you feel? Then you got to move back home with dad and mom, right? There's more to come. Stay tuned. Let me just throw this out there for a second. If you've been listening to this show for a while, you already have more information than most people ever act on. That's not the problem. The problem is nothing is reinforcing it tomorrow or the next day or the next day. That's why I built Ultimize, to give you something you can use repeatedly, not just hear once and forget. So if this is resonating at all, just go to ultimize.com, grab the free course, communication skills, habit building, whichever one you want, and actually start putting this stuff into practice. It's free, no downside, just Just go to ultimize.com. Hi, I'm TJ Walker. Thanks for joining me today. Today's a special day. I'm joined by my daughter. It is Take Your Daughter and Son to Work Day. So Rossa, lean in and say hello to folks. Rossa is an artist. Tell them where you go to school and what you do. I'm not saying that. Okay, you don't have to. She is here mostly to observe, but she's learning all the behind the scenes and how the teleprompter works, how a show, is built in what we're doing today. And comments are coming in today. Willow Sloan says, "Best show of all time." Well, so Willow, thanks so much for joining us. Claus Bears 9899 says, "Greetings, pretty young lady. Congratulations, TJ, scaring this topic." Yes. And Cat Z2D says, "Rossa, hi, and Boo." Are these some friends of yours? Yeah. Okay, we'll say hello to Cat then. Hi, Cat. Okay. All right, we're having fun and diving right into the next story. According to The New York Times, the Federal Bureau of Investigation examined whether it could pursue action against a Times reporter after she published a story about Kash Patel using government resources to provide scrutiny and transportation for his girlfriend. Agents interviewed the girlfriend, searched databases for information on the reporter, and discussed whether standard reporting techniques could be treated as stalking, believe it or not, before officials ultimately determined there was no legal basis to proceed. Duh. No kidding. Let's be crystal clear about what that means. A journalist writes a story about a powerful government official. The government then looks into the journalist. That's not normal in a healthy democracy. That's the move straight out of the authoritarian dictator playbook. In Turkey, for example, journalists have been prosecuted and jailed simply for reporting on government actions. In Hungary, the government of Viktor Orbán has been accused of using surveillance techniques to monitor reporters and squeeze independent media out of business. In the Philippines, Nobel Prize-winning journalist Maria Ressa faced years of legal harassment for exposing those in power. And if you want the most chilling example, look at Russia. Anna Politkovskaya— and I botched her name, Rossa, do you know how to say that name? I'm here with my daughter. Anna Politkovskaya, forgive me for botching your name, spent years reporting critically on government actions and abuses. She was threatened repeatedly. She was poisoned once. And in 2006, she was shot and killed in her apartment building. This is the extreme end of the same instinct: silence the person asking questions. Now, no, the situation's not that yet, but it rhymes with it because the principle is identical. In a democracy, journalists investigate people in power. In authoritarian regimes, people in power investigate journalists. That line is everything, and when you start blurring it even a little, you're chipping away at the foundation of the entire system. The First Amendment exists for one reason above all others: to make sure the government never gets to decide which reporting is acceptable and which reporting is criminal. Once the government starts asking, can we charge the reporter with a crime because we don't like what they said, instead of answering the question, is the story true, you're no longer defending democracy, you are eroding it. Now let's talk about Kash Patel, because this is not just a policy issue, this is a judgment issue. This is a man who appears so thin-skinned, so insecure, that criticism of his personal conduct triggers the machinery of federal law enforcement. That's not strength. That's not leadership. That is weakness with a badge. And there's a pattern here that shows up in every country that drifts in this direction. Leaders who want loyalty above all else do not surround themselves with the most competent independent thinkers. Oh no, they surround themselves with people who are dependent, people who know, wow, I'm wildly incompetent. I'd be lucky to ever get a job half as good as this. They also want people who will never challenge authority because they can't afford to. Because the more controversy they generate, the more they need protection from the person who gave them power in the first place. That's how the system feeds itself. And here's the bottom line. This is not about one reporter. It's not even about one investigation that ultimately went nowhere. It's about the signal. If you're a journalist watching this, the message is clear. Ask tough questions and you might become the story. And if the message takes hold, even quietly, even subtly, you don't need censorship. You get self-censorship, which is far more effective and far more dangerous because democracies don't collapse all at once. They erode one investigation at a time. So, I hope I'm wrong here, but I am very contrary to what a lot of pundits are saying now. Many, many are predicting this guy is fired, that Trump will say, "Hey, I've had enough." Let's fire Kash Patel. Many people are saying he's gonna be gone. The prediction markets say he's gonna be gone soon. I don't agree. I hope I'm wrong, but I happen to think he is doing exactly what Trump wants, and he fits the profile of what Trump wants. He's someone willing to investigate enemies and do things that are blatantly unconstitutional and illegal. Trump wants that. He's someone willing to harass and target reporters. Trump loves that. And he's someone who is beholden to Trump because this guy strikes me as a guy who would be lucky to be hired as an assistant security guard outside a small-town bank, much less having the most powerful law enforcement position in an entire country. So he knows he's never gonna get a better job than this. And he's just completely corrupt. He likes flying around in big government jets, flying his girlfriend around, commuting to his home in Las Vegas. So he knows that Kash Patel will do whatever he wants. That's why I'm predicting he's not fired, but we'll see. All right, again, it's take your parent— no, it's not take your parents to school day, it's take your daughter and son to work day. Which is why I have my daughter, Rossa Walker, sitting right next to me. Rossa, I know you don't love the spotlight, even though you are very much a good public speaker. What's your takeaway of all this in the news? Anything other than you find it incredibly boring? Interesting? What would you talk about today if you were doing a show? Um, AI is bad. AI is bad. Tell us why you think AI is bad. Because I need a little bit louder so people can hear you. Because you said that it's like taking stuff. You, you said talk about it. Well, it can take away jobs, there's no doubt about it. But cars took away jobs from stagecoach drivers. Do we want to go back to a time where there's only stagecoach driver? Would you like to get rid of cars? So it's a double-edged sword. I'm TJ Walker. This is Filtered. Thanks so much for joining me today. Take care. See you tomorrow when someone will be back in school and someone will be right back here with you at
More from TJ

The Breakout Cases - What Actually Caused Famous Speaking Career Breakthroughs | Ultamize

The Red Circle Sessions Finale - Build the Speaking Career You Came Here For | Ultamize

Two Stages, One Logo, Two Completely Different Realities - Main Stage vs Independent Events | Ult...

Pre-Reset - Why Repetition Builds Better Speaking Decisions | The Red Circle Sessions | Ultamize

Your Decision, Your Framework, and What Happens After - Build Your Speaking Career First | Ultamize

The Red Circle Sessions Trailer - The Speaking Career Truth Nobody on That Stage Will Tell You | ...

The Two-Minute Ultamize Reset - Your Platform, Your Audience, Your Voice | The Red Circle Sessions



