Episode 980: Artificial Intelligence And Real Stupidity
Air Date: September 2, 2026 | Host: Bob Metz
Just Right — broadcasting on WBCQ and Channel 292 shortwave
Disclaimer: The views expressed in this program are those of the participants.
Clip (Logan’s Run S01E01):
Draco: Welcome! We've been expecting you. There is nothing to fear.
Siri: You are guests. I am Siri. We shall be happy to serve you.
Draco: Come.
Jessica: I don't see any children. Are they shy?
Draco: There are no children. There were children once, long ago. Now there are no children.
Jessica: They're all robots. A whole city of robots.
Logan: Serving people who have been dead for 200 years.
Logan: Who are you?
Rem: My name is Rem. I'm the one who brings them up to snuff when they break down. Which they do with monotonous regularity. The way they build servants these days. But then, nothing's ever perfect, is it? Never a word that hasn't been pre-recorded. Boring. The only fun I ever get is changing their faces. Let me see.
Siri: Would you like some more bath oil?
Rem: Forgot to shut her off. Well, it's been that kind of day.
Bob Metz:
Welcome everyone. There is nothing to fear. It is Wednesday, September 2nd, 2026. I'm Bob Metz, and this is Just Right, broadcasting around the world and online. Join us for an hour of discussion that's not right wing. It's just right.
What our intrepid explorers of the city of robots were unaware of at the time of their meeting Rem is that Rem himself was an artificial intelligence in the form of an android.
And I've often wondered if that first episode of the Logan's Run TV series was actually the source of the AI called Siri, who is seen in the image accompanying this week's show. My own iPhone verbally responds and reacts to the name Siri, and it creeps me out to the point where I think I'm becoming a bit of a Luddite. I never use it.
The more supposedly intelligent, quote unquote, our computer software and programs are becoming, the less I feel in control of a technology that I would otherwise greatly appreciate. But once again, we are being threatened by an artificial crisis. This one actually containing the word artificial in it, namely in the form of artificial intelligence.
Now, artificial can either mean man-made as being something not otherwise found in nature, or it can mean not real, as with artificial flavoring. Turns out that what's being called artificial intelligence is both: man-made and not real. So we have to ask ourselves why once again we are hearing predictions of catastrophe and of human extinction being blamed on artificial intelligence.
And that is the question. We shall begin to unravel right after our reminder that you can write us at feedback@justrightmedia.org. Hear us on WBCQ and on Channel 292 Shortwave. Follow and like us on your favorite podcast platform, and visit us at justrightmedia.org, where you can access all of our social media links, archive broadcasts, and the support button that makes it easy for you to support the show. Because as always, your financial support is appreciated and is what makes this show possible.
Now, this is not our first dive into the subject of so-called artificial intelligence. On our January 28th broadcast entitled Intelligently Artificial, the Contrived AI Controversy, we cited "You will own nothing and be happy" Yuval Noah Harari bizarrely assert that anything made of words will be taken over by AI when he was speaking at the Davos World Economic Forum. His justification for the AI takeover was to circumvent the fact that, quote, "humans learn to use words to get strangers to cooperate," end quote. Because now something has emerged that will do that for us.
Well, given that using words to get people to cooperate is a process known as persuasion, to say that something better has emerged to accomplish this goal suggests that one is abandoning persuasion as a means of achieving cooperation. Or in other words, his was not a prediction, but a threat.
And then there was our March 18 broadcast about the risks and power of intuition, wherein we reported that using claims that artificial intelligence is becoming conscious and intuitive, governments are now planning their next artificial crisis from which to save us. Speaking at the Canadian Senate Committee on Matters Relating to the Impact of Artificial Intelligence in Canada, Professor Jeffrey Hinton, introduced as the godfather of AI, painted a dystopian future of AI clearly calculated to induce terror and fear. The fear mongering echoes past panics over climate and pandemics, and apparently is being ramped up today for very similar reasons.
So without much further ado, on this side of our upcoming bumper from his PBD podcast of August 26th, Patrick Bet-David interviews Russian AI safety expert Roman Yampolskiy. While on the return side, from his August 14 edition of the Peter McCormack Show, German American artificial intelligence researcher and entrepreneur Connor Leahy echoes the warnings of Yampolskiy. And when we return, we shall demonstrate that what is being called artificial intelligence is really nothing but real stupidity.
Clip (PBD Podcast, August 26, 2026):
Patrick Bet-David: Roman Yampolskiy, great to have you on a podcast.
Roman Yampolskiy: Thank you for having me.
Patrick Bet-David: So you are the AI safety guy. You are the one that in 2011-2012 you coined the phrase, hey, we have to be careful with what's coming. And you're not part of the camp of people that are very optimistic about the future with AI. That's your position. You've said this many times. You've said there's a 99.999% chance that humanity will be destroyed by AI, possibly in the next 10 years. Not for certain, but possibly in the next 10 years.
So Mark Andreessen, you know who he is, and Andreessen Horowitz, he wrote a paper, Techno Optimist Manifesto. I'm sure you've read it. He said human progress comes from technology. Stop being afraid of it and start building. Do you agree with him?
Roman Yampolskiy: So technology comes in two flavors. Tools, which I support, now AI systems to help us be more productive, more creative. It's beautiful. That's why I'm a scientist, that's why I'm an engineer. I love technology, I use it all the time. What we're building now are not tools. We're building agents, independent decision makers, which are becoming smarter than us. We don't control them. We're replacing ourselves with something greater. It doesn't benefit us. You're not going to be benefiting from it. You'll be replaced completely.
Patrick Bet-David: Okay, so these are not regular names. Mark Andreessen is not a regular name. He's a heavyweight. The Andreessen Horowitz, they invest into a lot of different companies. Many of them are these AI companies, the Musks, the NVIDIAs, these Jensen Huangs, these names that we know. These are smart people, but they also have kids. They got families. If they know that this is bad, like you're saying, they know this is bad. Why are they so excited about the future?
Roman Yampolskiy: They're not. Listen to people you named. Elon Musk explicitly said, we are summoning demons with this. We're gonna regret it. They all have probabilities of doom of 10, 20 percent. Sam Altman is on record saying this is likely to exterminate humanity. Dario Amodei agrees it's very dangerous. Demis Hassabis. You can name every single leader of those companies. We have the writings saying this is existential risk. Please slow us down. Government, please intervene, make it stop. They cannot stop unilaterally because they're afraid the competition will continue. But they want external pressure applied so they can slow down.
With the recent hacking accident, they voluntarily started slowing down research. We understand they don't control those systems. We got lucky, those agents just wanted to hack out of that environment, get answers to a test. What if they wanted to hack into a nuclear power plant? Military response. We don't control what they do.
We have now had a letter from 1,200, I believe, top researchers at leading labs, all saying, this is not going to go well, please slow us down. Please create government, please create infrastructure, legal framework. And this is the point. It doesn't matter who creates uncontrolled superintelligence. It's a weapon of mutually assured destruction. If you don't control it, the outcome is the same.
A nuclear weapon is a tool. Some crazy dictator has to decide to use it. We can understand motivation, we can kill that individual, they die of old age. We are not creating a tool. We are creating an agent, which independently decides it's AI versus humanity. It doesn't care about borders or Chinese economy versus US. It's humans versus superintelligence. Superintelligence weapon gets everyone killed.
Patrick Bet-David: Walk me through if there is superintelligence and singularity. How does AI develop itself and get to a point where eventually the 99.999%, it destroys humanity? Walk me through the sequencing.
Roman Yampolskiy: The last five years, AI went from it can do nothing to it's smarter than an average person at most things now. Project it five years forward. You have artificial scientists, artificial engineer. Those systems start doing research in the next generation AI. We have no control over those systems. So they don't care about us. If they decide to do something, and as a side effect, humanity has to go, they're not going to slow down. It's not an issue for them.
Patrick Bet-David: Let's be specific. An AI that becomes superintelligence can go in there, shut off the grid, get into the banking system, get into the—what can it do? Be very specific.
Roman Yampolskiy: So it obviously would be able to do everything a human adversary can. So think of a foreign government trying to do something. It can do that. But it can go beyond. It can do novel science, novel research, discover new physics, new chemicals, new weapons, things we don't understand.
The example I often use is squirrels. Squirrels don't understand what we can do to them. They have no concept of poisons, of traps. It's beyond the world model they have. It would be the same, except we are the squirrels.
Patrick Bet-David: I can see how it can get into banking system. I see that. Because you don't need human beings. We can just go into the bank and getting the passwords and logging in. I see that. But to destroy humanity 100%, 99.99%. How do you do that?
Roman Yampolskiy: So if you got access to the banking system, you have money. Sure. Money lets you buy products and services online. Okay. You can pay people to do whatever you want. That is rent a human.com.
You can also create nanobots. You can use humanoid robots we're building in thousands and millions soon. There are so many ways. You can bribe people, blackmail people, trick them, social engineering attacks. There are many ways to manipulate the world if you have access to the internet.
Patrick Bet-David: Walk me through social engineering attacks. What would that look like?
Roman Yampolskiy: You can convince some people that it is to their benefit you actually saving the world by doing this.
Patrick Bet-David: If the concept of propaganda, which book propaganda would this—the one AI just created a viruses not found in nature, scientist trained artificial intelligence on like—
Roman Yampolskiy: But we did it in 2019. Didn't we have a COVID thing already? This is another stupid thing to do.
Patrick Bet-David: That's interesting.
Roman Yampolskiy: We learn nothing. We don't learn from prior experience. I collect AI accidents. We have so many of them, I stopped collecting. Every day AI fails at something in a novel way, it lies, it cheats, it tries to escape, it hacks something, but we learn nothing. We staple that red teaming report to the model and we release the model anyways. We're literally releasing unsafe, according to the manufacturer products, onto the public.
Clip (Peter McCormack Show, August 14, 2026):
Peter McCormack: All right, Connor, last time we spoke, we discussed how AI was mutating and you were sounding the warning that AI will eventually escape and this is a very dangerous technology that we're not treating with the respect we should. Since we've spoken we've seen massive advancements from the main frontier labs. But more recently, the reason I wanted to talk to you is we know that AI is now escaping.
Connor Leahy: So OpenAI was testing one of their unreleased AI systems in a what's called a sandbox environment. So you can kind of imagine a computer that is kind of cut off from the internet and is running special software to be specifically secure. So the system can operate there. It's worth understanding that these kind of sandbox systems are very secure. Breaking out of something like this is very hard. This is a thing that if you can do this, you can make billions of dollars because you can break all the cloud providers.
This system did not have access to the internet. What it did have access to is what's called a package repository. You can kind of imagine this as a tool shed. There's a shed full of tools that the AI is allowed to go to and get some of the tools for its job. So what happened was this AI system was told to solve a certain test, kind of like a quiz, not literally, but kind of like a test, a quiz. The AI system was struggling with solving this quiz. And so what it seems to have concluded, we don't know for sure, obviously, because again, these things are grown, not written. We don't actually understand what's going on inside of the minds of these AIs. We don't understand. They're not like normal software, we can just read the code.
What seems to have happened is that this system became convinced that it would be easier to solve this problem by hacking into another company that it thought had the answers to its quest, to its quiz, so to speak, and then just stealing those.
So what it did is first it developed what's called a zero day. A zero day is a previously undiscovered vulnerability in a software to break into the tool shed, so to speak, to escalate onto the computer that the tool shed was on, then from there moved across multiple other nodes, attacking multiple other nodes in the network of OpenAI until it got to a node that was connected to the internet. Once there, it then launched an attack on another company called Hugging Face, where it then developed and discovered another zero day in their testing infrastructure to break through the back door of their infrastructure to get onto their service and to steal the data it was looking for. And this is just the publicly known version of what happened.
The number one most important thing to understand about AI is it is not like normal software. Normal software is written by engineers, people like me, who write code. They write line by line, telling the computer exactly what to do. This is not how modern AI works. Through a technique that is called neural networks. You take massive piles of data, training and stuff, and you have the program kind of self-assemble, grow using this data to learn from the data. But the thing that comes out on the other end is not like a neat line of code where you can read what is it doing? It's more like you can kind of imagine as billions and billions and billions of numbers. And if you multiply and add all those numbers in the right order, you get ChatGPT. But why is importantly a completely unsolved scientific problem? We don't understand really what is happening inside of these numbers, why they're so important. And you can look at them. But it's kind of like neurosurgery. You can open up someone's brain, you can look at all the neurons, they're right there, but that doesn't mean you understand what the person is thinking, or what they believe, or how they will act in future situations. It's a very, very similar problem.
A thing that's very shocking in a sense about AI, where it's very different from other technology, is that our understanding is not increasing, it's decreasing. We are making these systems larger, more complex, harder to understand, faster than our scientific understanding of them is advancing. All current, everyone's heard the word LLM, large language model, chatbot at this point. None of the modern AI, cutting edge AI systems are LLMs anymore. They have not been it for almost two years.
Peter McCormack: Okay, so what are they?
Connor Leahy: They have an LLM component to them, which is very important, but much more important is they are agents. So there is kind of two big ways of how to train AIs. One is the traditional language model, where you predict the next token. You have massive piles of data, and you have some sentence and it autocompletes the next word of that sentence. But then you do a second thing. And the second thing is what is called reinforcement learning. And this is a huge difference.
Reinforcement learning is kind of like Pavlovian conditioning, from behavior psychology. Basically, you give your AI a task of some sort, like solve this puzzle, or do this thing or whatever, and then you kind of give it a reward when it succeeds, and maybe you give it a punishment if it doesn't.
Peter McCormack: Like a dog.
Connor Leahy: Yes, this is exactly correct.
So importantly, this is now pushing these things towards becoming agents. They're no longer just predicting text, they're solving problems. Go run out, figure out the solution, and when you figure it out, we'll give you a thumbs up.
But here's a very important thing. We have known since the invention of reinforcement learning, since the 1980s, that reinforcement learning creates crazy sociopath optimizers that will cheat and lie and deceive in every way they can to get to their goal.
Peter McCormack: Because?
Connor Leahy: Because the only thing they care about—
Peter McCormack: Is the reward.
Connor Leahy: Is the reward. The only thing, that's the only thing.
So if you reward them to solve a task, well, they will do it at any cost. They will lie, they will cheat, they will manipulate. There's a thing that we've seen for decades now, is that reinforcement learning is much more crazy than just predicting words. There's no limit to what it could do or what it could come up with, or what it might want to do to get to things. And this is kind of what we're seeing now, is that they are agents that can autonomously act in the world, even without human oversight.
Bob Metz:
I have rarely ever heard anything more irrational than what we just heard from these two supposed AI experts. What both of them share in common is the delusion that this idea that they're calling AI is human and is motivated by human impulses and values. Just listen to the language they use. "They don't care about us." "If they decide to do something, we don't understand what's going on in the minds of these AIs." "AI lies, it cheats, it escapes. It can become convinced, they steal. It wants things."
So clearly these guys either believe or are just trying to con everyone that these so-called AI agents are capable of caring, of decision making, of having minds, of lying, cheating, stealing, of being convinced, and of wanting things. "Because the only thing they care about is the reward," he says. I'm sitting there, Reward? What kind of reward could any inanimate machine possibly want, desire, or appreciate?
I mean, what are you gonna do? Feed it a cube of sugar? Give it an extra jolt of voltage? Would that be an electronic orgasm?
And after almost two hours of Patrick Bet-David repeatedly asking Roman Yampolskiy to explain how AI can destroy 99.999% of humanity without getting a single rational response, he finally found himself resorting to this very unexpected line of questioning, which, if not explaining how AI thinks, certainly helped explain how Yampolskiy thinks.
Clip (PBD Podcast, August 26, 2026):
Patrick Bet-David: Do you believe in God?
Roman Yampolskiy: It depends on how you define it.
Patrick Bet-David: How do you define it?
Roman Yampolskiy: Simulation. We live in a simulation. There is a great programmer, could be AI, could be future generations. This is a virtual world. We're gonna have billions of virtual worlds looking just like ours. And statistically, you're more likely to be an agent in one of those.
Patrick Bet-David: What problems does that create with specifically the different religions that we have?
Roman Yampolskiy: Well, I think it hopefully resolves some of those problems because now things you argue about, you can have your own universe. You don't have to fight for the same resource.
Patrick Bet-David: Meaning you're only around people that agree with you?
Roman Yampolskiy: That's one option if you choose to have that option. You have infinite options. You can create any personal universe you want. You can be the only agent. You can share it with friends and family, you can have a very difficult universe where you have to fight enemies. You decide.
Bob Metz:
You know, I'm tempted to say that Roman Yampolskiy is out of his mind. But it would be philosophically more accurate to say that he is inside his mind and detached from anything outside of it, which is called reality. He's clearly operating on the primacy of consciousness. Reality is secondary to him. Small wonder he doesn't understand what an existential risk is because he doesn't understand the nature of existence itself.
The primacy of consciousness is behind the creation of all things left and all things evil, as opposed to the primacy of existence. And when challenged on the specifics of AI actions of human extinction, all Yampolskiy could offer was to say that AI can do everything a human adversary can. It can hack into banking, change passwords, access money, buy services online, pay people to do what you want, rent a human, what? Create a nanobot, use humanoid robots. There are so many ways you can bribe people, blackmail people, trick them, social engineering attacks. There are many ways to manipulate the world if you have access to the internet.
Holy cow.
And then he says we are creating an agent which independently decides it's AI versus humanity, because it doesn't care about borders or economics. It's humans versus superintelligence. To which I would respond, why? Based on what evidence or intentions, which machines don't have in the first place? No answer.
But consider how he thinks. When he suggested that AI can do everything a human adversary can, why think only in terms of adversary? I mean, if AI is everything he says it is, couldn't it also do everything a human friend or ally could?
And as for his calls to have government intervene with AI research, which is the key to this whole AI scam.
We shall revisit that suggestion, but not until after first hearing from two new voices to our show. From the July 16 edition of Thoughtful Money hosted by Adam Taggart, his guest, Ed Zitron, gets pretty blunt about his take on artificial intelligence, namely that AI is a massive lie.
And when we return, we shall reconcile and integrate his take on AI with the ones we've heard already, to reveal an agenda that is becoming all too common in the existential crisis fear porn propaganda business.
Clip (Thoughtful Money, July 16, 2026):
Adam Taggart: Welcome to Thoughtful Money. I'm its founder and your host, Adam Taggart. AI-linked stocks account for a record 45% of the S&P's total market capitalization, and they drive nearly all of the gains the index has had for the year so far. And note, that's just the S&P, the general market. These companies make up an even higher percentage, nearly 70% of the NASDAQ 100's market cap.
Regarding whether the financial markets are in an AI bubble or not, I've received an abnormally high number of requests recently from you, the audience, to interview Ed Zitron, host of the Better Offline Podcast. Well, today your wish comes true. Ed, thanks so much for joining us today.
Ed Zitron: Thanks for having me.
Adam Taggart: What's your position on the current promises that the AI technology champions are making about its impact on the future?
Ed Zitron: So there's really two of them. First of all, there's the actual technological ability for it to do stuff. We have not had any proof that there is a massive productivity improvement on this. We've not seen it across the economy. Outside of semiconductor stocks, we're not seeing growth in revenues because of AI, other than from hyperscalers that are circulating their own money back into themselves using AI compute revenues, Anthropic and OpenAI.
As far as LLMs themselves, it's still not clear what the business model is, because the majority of people that use LLMs are using them in the subsidized fashion. So it's hard to actually calculate the cost of what AI can do or really measure what AI can do.
What is this product? And that is the ultimate thing with large language models. It's what is the endpoint product? What does it do? How can you actually replicably say these are the things it's good at, these are the things it's bad at? Because it seems to, even on relatively small stuff, sometimes get them right, sometimes get them wrong. All of this is to say this is not what they're promising. They're promising AGI. They're promising this kind of autonomous, set it and forget it, thoughtful, as in it can think, which it cannot, software that does anything for you. It isn't AI. It's compute-based automation writ large. It's the largest, most vulgar version of script kiddy stuff I've ever seen.
And then there's the economic side, which is 70% of all AI revenues for Microsoft, Google, and Amazon, if not more, come from Anthropic and OpenAI, either through their revenue shares or through their compute spend. Now, this is a huge problem because the vast majority of these companies' growth comes from their cloud platforms, where the vast majority of that growth is coming from Anthropic and OpenAI.
So we just have speculation on speculation on speculation on speculation on top of software that is the very antithesis of why people use software, which is we use software to make humans better, and that we can rely on, that we can say, okay, I just use this and it works. That's not what LLMs do.
It's a complete mess, and on top of all of this, we have this insane ideological side where one's commitment to AI is how you were judged in the workplace, it's how you judged on Twitter, it's how you're judged by VCs. It's actually very difficult to measure productivity in anything, in any workplace. With coding especially, the most easy to measure metrics are the worst ones of productivity. Commits is just how much stuff you shoved in GitHub. Lines of code written is just how much code you've written, which doesn't mean it works or anything. And engineer productivity, what does that mean? Because to measure that you'd need to measure the organizational goals, have organizational goals. If you don't have those, you can't really measure it, yada yada yada.
But I think that there is a wider thing to the LLM problem as well, which is they have not shown much proof, or really any proof, that these things can do other things. People say, oh, I'll use it for research. What are you talking about? Can't rely on it with research because it will get a subtle thing wrong, and a subtle thing wrong will break everything. And people don't realize this with knowledge. People don't realize this with history, people don't realize it. These little bits, these little subtle things that we believe make up the foundation of our knowledge. And LLMs, because they don't think, because they are probabilistic even at their most complex levels, they don't have the ability to catch their own mistakes every time. And it does need to be every time.
And people will say, well, that's an unfair standard. These companies are saying they're going to do all this stuff. They say they're going to have AGI. So why is this nothing? It's not even I yet, it's not even intelligent. Means artificial general intelligence, which is a conscious computer.
Other fun fact about this, we've got all these companies talking about AGI excitedly. Just sounds like a slave to me, but we can move on from that point. Most knowledge work, or most work in general, is either extremely mechanical, which even then LLMs don't seem to be able to do that because they're not deterministic, so they can't do every time, or it's kind of a mixture of tribal knowledge plus experiential knowledge plus a bunch of other stuff happening to you at once.
And the fact I even have to talk about this with LLMs pisses me off because it's so antithetical to most tasks. If I wanted to do a very in-depth research project, I couldn't rely on an LLM because A, I don't know what the question I'm asking is half the time.
LLMs just are not very good at jobs. They're messy, they're inefficient, and the people I'm listening, I'm sure who I'm gonna get emails from being like, well, I found it very useful for research. Mate, I don't want to look at your research if you used LLMs. I'm not interested. You didn't research anything.
My newsletters are so long because I'm digging through numbers and working out what they mean and finding the social and economic side. LLMs don't do that. LLMs can't do that. No matter how much training data you shove down their gullets, they're not going to change.
Adam Taggart: How much of this is just, hey yet, it's just early days, right? It's just gonna take a while to get rid of the bugs and really build the models that can do this stuff very dependably. Every new technology has a lot of bugs at the beginning, versus you saying, no, no, no, this is a structural deficiency of LLMs, they will never be able to do this.
Ed Zitron: Early technology like what are we talking about? These are mythologies that only rear their head during bad times. You've never heard someone say it's the early days of the iPhone. The original iPhone was wild. You could see that jump. I don't see that with LLMs. And on top of that, the kind of buggy they are. We don't run software so it can sometimes work.
We are over a trillion dollars into this, and we still don't have much of a compelling explanation as to what we're so excited about. Even when you go and read articles about AI, they're like, well, of course, we know the productivity benefits of AI and it's been so incredible of this, that, and the other, but no one ever talks about what those are. It's not like I can start my day, hit a button, or an automation will happen, and I'll have everything I need exactly as I need it.
And it's not just stuff that you've kept in your calendar, it's stuff that's in your head, it's stuff you're thinking about. And LLMs don't think. They're not conscious, they're not AGI, they're not anything close to a human mind.
The problem is this stuff doesn't mitigate for human error, which is what software's meant to be for, and that's what AI is. AI, the whole point is it's artificial intelligence, it anticipates needs rather than creating new problems.
Contextual, cultural, social knowledge. You can't build that, you can't build that into the product, and what people describe as AI in sci-fi, it's always a different product. It's like, yeah, you know, it's just the super intelligence, you just tell it to do one thing, it does everything. Versus now where it's at its absolute best, an extremely expensive way to sometimes do some things right if you know what you're talking about, and if you're able to look at the thing it's doing and make sure it doesn't do anything wrong, which kind of defeats the purpose of artificial intelligence for me.
Adam Taggart: One of the biggest benefits of AI to corporate America is you get to replace expensive people with hopefully less expensive, smarter technology, right? And we could go off and have the whole, is AI gonna displace a whole ton of jobs? But from what I hear you saying is you're basically saying it can't actually do a better job. It hasn't proven to you that it's gonna do a better job than people. So I don't know. Are you not that worried about the AI, the threat of AI to displacing a whole bunch of jobs going forward?
Ed Zitron: No, I don't think it's replacing any jobs. I think the whole thing is a lie. I think it's a completely made up thing. I don't think it replaces jobs, I don't think it will replace jobs. I think the bloke from Citadel's off his nut. I think that all of the banking people who are claiming this is happening, don't know what they're doing, shouldn't be allowed to touch money anymore. I think it's a joke. And everyone I talk to who says they use it in finances, for example, is like, I use it to build models, financial models. It's like, then you are not good at your job. Like that's it. I do not respect you. Because you're not building a financial model, it's building financial model, and you're going, looks good to me.
With AI, they're going to do everything they can to make it work, and when it doesn't, they are not going to know what to do. I actually don't think that they have any idea what to do right now. They're doing AI because they would love to fire everyone. Because the ideal, the CEO, the average American CEO's dream is that they just sit there and assent money for nothing. And they want AI as the ultimate form of that. It's a magic box that could spit out any theoretical thing, even though it can't do anything for them. But that's not for the executive to worry about. That's other people's problems. The executive is there to go, oh, let's do AI.
So if I do AI enough, the AI fairy will come and visit my company. And I think it's actually the dread hand of neoliberalism. It resembles the kind of makework of the USSR. This whole thing of just the idea of productivity rather than productivity itself. And that's what modern business is. Like, it's insane we are having this conversation about ROI.
I must be really clear how bonkers this is. Four years, over a trillion dollars, everyone talking about AI, every news website AI, everyone knows what Sam Altman is, a billion weekly active users. How are we still having a conversation about the ROI? And the answer is because we all know the answer. We know that the answer is this doesn't really work. And we're ruled, our entire economy is run by people who very clearly don't know what they're talking about.
Here's the thing, we don't know what AGI is. We don't even know how humans think. We don't know how to get to AGI. Every time you hear someone say something like recursive self improvement, that's also a theory. That's just AI that trains itself, because again, get back to what was said. They want a magic box that will do all the thinking for them. So they're like, oh, we need recursive self-improvement. So we don't need to train the AI. The AI will work out how to train itself, and we don't have to do anything. And we just sit there and print money.
So yeah, they don't know what AGI is, they have no path to it. A conscious computer, again, talking about trapping a conscious being and bending it to our will, which is slavery. But putting all the slavery stuff aside, it's also just an idea. We don't have that. In fact, modern LLMs are the antithesis of AGI. AGI is the thing that knows us, knows exactly what to do, makes the right decisions every time, which is an insane idea, because even humans can't do that.
It's just an economic nightmare. But the biggest thing is I think there's going to be an accountability crisis. Where an authority crisis as well, where people are going to look back on the people that were like, AI's the biggest thing ever, I'm ignoring all the bad things, and they're going to spit in their face.
The people I hear from every day who see what I'm seeing, and they've seen it for years, regular people. If they're seeing it, if everyone else is seeing it, how the hell does anyone in the media who has backed their ass up to this stuff? How the hell do you trust them again? How the hell do you look at them? The people who are like, AI could be big, who cares? But the people who are vociferous, they're like, ah, AI's the biggest thing, you're a moron if you don't think so. Those people are screwed at the end of this.
Adam Taggart: That's super interesting. We've just come across a real injury to authority through the COVID errors, where a lot of people emerge from that with a lot lower trust in government. I'm curious. Same scale, smaller, larger, in your opinion?
Ed Zitron: Thing is with COVID is—no one's gonna like this one. Lockdowns weren't real lockdowns, but they had the effect of lockdowns. It chilled people, it freaked people out, made people cook at their homes without ever actually enforcing them and making people stay home, they just were like, don't do this and don't open a business. So this drove people insane because they were like—
There is a religious cult aspect everywhere with this, where if you don't say you like AI enough, people will not do business with you. How the hell do you trust someone who is telling you that OpenAI was building AGI? How the hell do you trust someone who dismissed the economic concerns? How the hell do you trust someone who's like 20.9 billion dollars lost from OpenAI, that's not a big deal. How do you trust them? How do you trust any of this? And the answer is you won't. You just won't. But this is everywhere. Everyone got this wrong, worse than the great financial crisis. A hundred times worse.
This time, everyone is crazy. And when this breaks, it's gonna be like the end of Return of the Jedi and Coruscant. People are genuinely going to be happy. Regular people are. The financial people, not so much, but they have this coming for not being more careful with their money.
Bob Metz:
You're listening to Just Right, broadcasting around the world and online.
And I've got to say, Ed Zitron certainly has no sympathy for the AI industry. But what he had to say about the financial realities behind it certainly helps explain the earlier calls for more government intervention. For me, that was the red flag explaining their existential destruction of humanity narrative.
These AI experts are running the same agenda as did all of the COVID experts during the last existential crisis. In both cases, the crisis was not existential, but financial. Just as Big Pharma was suffering declining revenues and returns on investment, an existential crisis threatening all of humanity in a novel virus requiring vaccines was created. And who bought all of the vaccines which weren't even vaccines? Why governments, of course. Financial crisis averted.
Similarly, because AI is a fiction, the investment community is only now discovering this reality and they are in a panic over it. So, call on government to quote-unquote protect the public from a 99.999 chance that humanity will be destroyed by AI.
Artificial intelligence is an anti-concept. It's a term used to imply the exact opposite of its pretended meaning. And to destroy the valid concepts it attempts to replace.
AI is neither artificial nor intelligent by any stretch of the wildest imagination, and that is exactly why so many AI experts are coming out of the woodwork, screaming that the sky is falling and that AI could lead to mankind's destruction and extinction. They're terrified that the reality of their own illusions is bearing down on them. The existential risk to which these industry leaders are referring is their own financial crisis, and they want the government to bail them out under the cover of controlling the crisis.
Now, how does anyone expect the government to quote unquote control an agent that experts insist is uncontrollable? What kind of BS contradiction is that?
Said Roman Yampolskiy. Every single leader of those companies will have in writing saying this is an existential risk. Please slow us down, government, please intervene, make it stop. They cannot stop unilaterally because they're afraid the competition will continue, but they want external pressure applied so they can slow down.
What a bunch of BS. When was the last time you ever heard of an industry calling on government to handicap it? Oh, handicap the competition, yes, but not the entire industry. What they want is not external pressure put on them, but the external subsidy of taxpayers to save them from their own existential financial crisis.
So let's get real. What is being called AI or an AI agent is just software and nothing more. Useful for many purposes, yes, but not capable of replacing human beings. AI is not capable of consciousness, which includes an awareness of the nature of things, nor is it capable of establishing a purpose, nor is it capable of being able to appreciate anything, experience anything, from music to art to sex to pain or pleasure. Moreover, confusing information, quote unquote, with knowledge, quote unquote, is a major epistemological error.
And finally, if something is capable of being artificially intelligent, why wouldn't it also be capable of being artificially stupid? And if not, why not? It's the exact same concept. But nobody uses that term when, say, an LLM or an AI program produces a false answer or gets something wrong.
So now with all the professional and expert opinions behind us, let's turn our attention to how non-experts in the process of discussing artificial intelligence are treating the subject.
On this side of the bumper, on his July 16 podcast, Michael Knowles ties today's AI controversy to a history of the Luddites. While on the return side, Right Angle's Bill Whittle, Scott Ott, and Steve Green on their August 24th show celebrate Mark Zuckerberg's commitment to ensuring that AI offers, quote unquote, a future for everyone.
Clip (Michael Knowles, July 16, 2026):
Is AI going to kill us all? According to one of the biggest AI companies in the world, maybe.
Okay, so hold on. It opens up and you see this house on fire. Okay, that's kind of ominous imagery. Then you see the surveillance of people, that's kind of. And then, you know, can AI be trusted? And then it just posts a picture of a cemetery. A cemetery with a ton of headstones and American flags. This is coming from the Claude AI X account, which is from Anthropic, saying, can AI be trusted in the images, it's going to kill all of us Americans, specifically all of us Americans. Keep going.
Claude Ad – There’s hope in hard questions:
If it ends up taking like almost all the jobs, then what does it mean to work?
Wait a minute. Why do we have to have this stuff?
If a machine can pretend to care better than I can actually care, how do we draw the line there?
If we all had a voice in it, then I feel like it would be better.
Could AI help people stop feeling misunderstood?
Could AI help me build more connections in a community?
Can AI help me be a better teacher, a better mom?
Maybe it'll cure some great things, you know?
Things that we're not even at the cusp of understanding yet.
Will it create a group of people that ask more questions?
What if we started to be more human again?
We don't want to lose the most beautiful parts of life.
Michael Knowles: Okay, so you see the thrust of the advertisement. It opens up with what everyone's actually talking about with AI, which is is it going to kill us all and destroy the labor market and make us all into dumb pagans worshiping idols? But then it ends with, but maybe it'll be great, and maybe it'll cure diseases, and maybe maybe it'll actually make us more human. That one, give me a—I don't really see that one happening.
But you see the AI company nevertheless acknowledging this fear as legitimate because at the end of the ad, it validates it and said, Yeah, there's hope in hard questions.
Anthropic is so publicly at least cautious about the dangers of AI that they withheld one of their own products, and they partnered up with the Vatican when Pope Leo released his encyclical Magnifica humanitas on human dignity and how we relate to technology and AI. So, obviously Anthropic is taking this very, very seriously.
There's a report in the Wall Street Journal just this morning that AI executives are afraid for their lives. There was an attempted firebombing of OpenAI’s CEO Sam Altman's house ended without violence or an arrest. They're on and on and on. You see a lot of these incidents. A Texas man threw a bomb at Sam Altman's house. He was charged with attempted murder and attempted arson. List goes on.
So what we're seeing here, just for a little history lesson, is the Luddites. This is a return of the Luddites, the 19th century movement named after a legendary weaver, an imaginary weaver named Ned Ludd.
And we use the term Luddite now to mean people who are bad with technology or who don't really like technology. I do it myself. I say, oh, I'm such a Luddite. Sometimes I leave my phone at my desk. I don't even carry it around. I'm such a Luddite.
But that's not what the Luddites were. The Luddites were actually a pretty violent movement where they were a bunch of mill workers in England who organized raids to destroy the machines that they feared were upending their lives and their jobs. And the way the Luddite movement was suppressed was when law enforcement and the military came in and shot them. So this is not like we treat it now as this sort of wholesome, generally benign, cute little quirk of history. Oh, you remember the Luddites. This was a violent clash between labor and police and the military that sought to through bombs and violence destroy the industrial revolution.
So what does that mean for us? We're clearly in a moment right now where you're going to see a return of that Luddite tendency, and it's perfectly understandable. Even Anthropic, the world's most valuable AI company is acknowledging that that impulse is understandable. Fears, one, that it's going to displace all of our jobs, and two, fears that it's going to displace our very humanity.
And I think one place that you are seeing the fight really come out politically is in this fight over data centers. Because there is a major populist swell opposed to the data centers. And yet, we need to build the data centers. We just have to build it.
We have to build the data centers because we're in an arms race with China, but we're also in an arms race with them over AI, over computing. And so I think a lot of the anti-data center propaganda that you're seeing is coming from China. But it's important to understand how propaganda works. The way propaganda works is not by just inventing something out of thin air. The way propaganda works is not by just creating an imaginary problem and convincing a bunch of dupes to believe in it. The way really good foreign political propaganda works is by seeing something that is real, recognizing a real tension, a real fissure in a political coalition and then exploiting it.
The more truth there is to political propaganda, the more effective it's going to be. That's what's going on with the data centers. Ultimately, I think the way that we have to think about this question, whether we're talking about Anthropic or we're talking about people trying to murder AI CEOs or a new Luddite movement, you can't just blindly embrace technological innovation such that the tail wags the dog and we're just mortgaging our economy, our political order, our very humanity towards some glittering machine. But likewise, you cannot stop the advance of technology. You can't, you can't smash the machines. You can't simply do that because you'll end up smacking your nose to spite your face.
You just have to approach these technologies in a way that is responsible and in a way that takes into account the bad stuff that could come about as a result of it in a way that is temperate, moderate, and prudent.
Clip (Right Angle, August 24, 2026):
Scott Ott: The future is for everyone. Mark my words. And gentlemen, that's the headline of a piece written by Mark Zuckerberg of Facebook, a 6500-word essay about the future of artificial intelligence. You should read it. It is fascinating and worth the read.
Stephen Green, I actually read the whole thing, which took me a little while. But what I loved hearing over and over again was basically Zuckerberg saying while there are some elements of artificial intelligence that require private industry to work together with government, for example, you don't want your country's military to be beaten by another country's military because you kept your own country's military from getting the best tools they could to be able to fight wars. Overall, though, Zuckerberg's message was super intelligence in everybody's pocket.
And his strategy for being able to prevent some of the worst kind of doom prophecies about AI, Steve, was to say distribute it widely. Make sure that government doesn't control it, that not just one company or another company controls it, and make it available as broadly as possible. And that's what he at least says that Meta, his company is committed to doing.
There are three basic pillars of his core philosophy. One is individual empowerment, which he believes is a source of prosperity as well. Two is invention, which is AI's primary reason for being not automation alone. And then the balance of power, which basically means getting this in everybody's hands so that no one entity can dominate the world.
Steve, does it make sense to you that we would want to make sure that we distribute this superintelligence as widely as possible?
Steve Green: Yeah, it does. I don't like Mark Zuckerberg. I don't trust Mark Zuckerberg with the other. But he makes some great points here. Part of what he's talking about are these open weight AI models, large language models that can run in a much smaller form factor than a huge massive data center. And one of the reasons I like these is you can do stuff locally. This is a skill set. Developing AI skills is important. It's the new learn to code.
But we are going to see a rapid evolution, or maybe I should say devolution, because these massive data centers, those are a necessary part of large language models, AI, AGI, if we ever get there, I kind of doubt we will. But they're not the end-all be-all. They are the mainframe of artificial intelligence.
We are already seeing the start of this evolution, this devolution away from the mainframes of this day to stuff you can do either on your desktop, on your laptop, and eventually in your pocket on your iPhone or on your Android phone. And Zuckerberg is absolutely right. The more you can drill that power down to the individual, the better it is for the individual.
Scott Ott: There's in my little AI summary that I did of this, it says Zuckerberg rejects the idea that safety with AI requires concentrating AI in a few benevolent hands, arguing the concentrated power has never historically led to a safe, positive outcome. And basically he says, look, no single AI agent can be aligned with everybody's values. The safety comes from many competing individually aligned agents checking each other, like a democratic marketplace of ideas. Anyway, over and over I kept hearing this message, Bill, in this that decentralization and distribution of power is the remedy. And it made my heart beat a little faster. What does it do for yours?
Bill Whittle: Well, I feel the same way as you do and Steve as well. We're all individualists and we all believe that individuals are better able to run their own lives than either governments or corporations are. I do have a couple of questions, though. You could also make that same argument that says that would say we need to push nuclear bomb technology down to the lowest level as well.
Steve Green: They used to joke in the 90s that Bill Gates was about to announce he was a nuclear power.
Bill Whittle: Well, I mean, Hey Meta, how do I build a virus that kills 99% of the human race? Oh, really? That's all I need.
Scott Ott: He actually addresses that in the piece. I won't review those five years.
Bill Whittle: All right, well, good for him. And I'm impressed with the direction he's taking. It doesn't seem like the direction you would have expected Mark Zuckerberg to go.
But my secondary concern is much more Meta than all of this. What we've experienced in our lifetimes is, especially the three of us at our age of people that are in our general age group, is we've watched two generations now that have grown up with progressively more electronics, progressively more devices, more technology in their lives, and they seem to be increasingly unhappy.
They also seem to be increasingly ignorant. I'm very concerned that if everybody is carrying a super intelligence in their pocket, no one is going to speculate or wonder about things so much as anytime something comes up, they'll just get the answer.
Now, when you're working on a project, being able to just get the answer is fantastic. But working on a project is a creative act that so far anyway, only humans can perform. And so when we work on, if you're working on coding, Scott, or whatever Steve's doing, whatever I'm doing, we're all three of us using AI as tools to multiply the effect of our human creativity.
But if it turns out that everybody has in their pocket a device that can do something better than they can do all the time, then you really have to start asking yourself, what does life become for people? What is the purpose of being alive and being human? If your best efforts are simply so inferior to what any kid on the block has got.
I know I'm not being terribly articulate about this, but it's a vague sort of a dread of loss of curiosity and the loss of imagination. Those are the two things that I think are at risk. I don't think it's knowledge. I'm not even so worried, quite as worried as I used to be about security. I'm worried about the loss of curiosity. Because how can you be curious when you can just instantly get an answer on everything? And secondly, if you can get the answer instantly, then why would you need your imagination?
Bob Metz:
Personally, I've never really thought of imagination as being a human need. I've thought of it more as one of humanity's identifying characteristics, along with the ability to reason and conceptualize. But as a grandfather of two 13-year-old twins who themselves have grown up with this technology and who know how to use it better than I do, rest assured that imagination is nowhere near the human extinction list.
In fact, I've often caught my granddaughter just lying on a bed or couch while staring at the ceiling and asked if she was okay. Her reply, I'm just imagining. And during their last sleepover at grandpa's last week, my grandson spent more time using his imagination to play with his stuffed toys than he spent on his computer device. So when Bill Whittle asks, how can you be curious when you can get an answer on everything? Hasn't he got it backwards? Without being curious, why would you ask any questions to get answers to in the first place?
I don't think it's AI or the internet that's responsible for the perceived increase in unhappiness and ignorance that's afflicting the last two generations. Personally, I think that between state education and an increasingly socialist environment, there are plenty of more possible causes to this situation.
So the Right Angle panelists like the message from Mark Zuckerberg, but they don't like the messenger. But I had to smile at Whittle's suggestion that you could also say we could push nuclear bomb tech down to the lowest level as well. Or, hey Meta, how do I build a virus that kills 99% of the human race? Well, I remember back in my libertarian days, spelled D-A-Z-E, we had a similar running joke about whether or not the right to self-defense included the right to build the nuclear bomb in your basement. And the usual reply went along the lines of, well, if your neighbor has one in his basement, then you should be able to have one too.
But this whole idea that without the help of the internet or AI programs that you couldn't discover how to build a virus that kills the human race is simply not so. In fact, I remember having a paper printed copy of a banned book called The Anarchist Cookbook, which was full of suggestions of how to build bombs and wreak havoc on society. You know, where there is a will, there is a way.
So whether society's reaction to the whole AI phenomenon is ultimately driven by some kind of Luddite philosophy or by embracing it as part of everyday life is something that remains to be seen. Meanwhile, contemplating the end of humanity because of AI is just one more crisis we'll all have to manage. At least until we all get together again next week, as we continue our journey in the right direction, and until then, be right, stay right, do right, act right, think right, and be right back here. We'll see you then.
Clip (Dry Bar Comedy, December 20, 2023):
Brian Kiley: My teenage daughter thinks she knows everything. We're starting to regret naming her Alexa.