Transcript
This is an AI transcription and may contain errors
Justin: Hello, and welcome to the AI Argument. AGI has been achieved externally. Hello, Frank. Welcome to the show.
Frank: How’s it going, Justin? So, yeah, AGI has been achieved. Are you talking about ClawdBot?
Justin: ClawdBot.
Frank: Slash MoltBot slash OpenClaw,
Justin: Yes.
Frank: which, believe it or not, are all the same thing.
Justin: Yes, they are. For reasons that lawyers know best, they had to switch their name.
What is Clawdbot / Moltbot / OpenClaw?
Justin: So ClawdBot, before I dive in and just give a quick explanation as to what it is, have you tried it? Have you delved in, have you done anything with ClawdBot?
Frank: I have not tried it, for reasons that I think will probably become apparent as we chat. But I understand the whole legal thing, because when I heard about this ClawdBot, you know, I just heard people talking about it and I assumed it was another iteration of, you know, Claude Code, Claude CoWork, ClawdBot.
Frank: But of course it is not. It was spelled C-L-A-W-D, apparently. The guy who came up with it, what’s his name, Peter Steinberger, he’s an Austrian chap. Personal project of his. I mean, pretty amazing, I think so. Yeah. Yeah, I’m pretty sure.
Justin: Woo-hoo. Go Europe.
Frank: Absolutely, yeah. So apparently he’s a Doctor Who fan, and this, the ClawdBot the CLAWD was a reference to Space Lobsters from Doctor Who. But Claude and Anthropic weren’t too happy that it sounded so close to Claude, C-L-A-U-D-E, which is…
Justin: It was genuinely confusing, right? Even if we were talking to other people: is that Claude Claude, or ClawdBot Claude, or whatever? So yeah, look, OpenClaw’s good.
Frank: Yeah, so he changed it to MoltBot and now he’s changed it again to OpenClaw to get his Space Lobster back in there. So yeah.
Justin: Well, I for one welcome the rule of our new Space Lobster overlord.
Frank: You are way more techie than me. Did you try it out?
Justin: I did. So what is ClawdBot? Let me give you the 30 seconds, hopefully not too techy. ClawdBot is an agent that runs on your machine. You download it, you fire it up and it starts to ask you questions, and it gives it a soul. There’s actually a soul.md file that gets installed so you can tell it everything, like who you are, what you do, what your relationships are, who your family is.
Justin: You just give it a picture of who you are now. That’s cool, right? Nothing mind-blowing about that. Where it starts to get interesting is it’s always running in the background, so you can connect it to your emails, and you can connect it to your WhatsApp, or you can connect it to Telegram, or you can connect it to anything you want.
Justin: And I believe, Frank, in fact you even got a couple of messages during the week from my very own ClawdBot.
Frank: Well, here we go. And this kind of illustrates the confusion thing. I thought when I got those messages from Alfred, I thought that this was something you were running with Claude Code, because I know you’re a Claude Code person, you’re philanthropic. So I thought this was from Claude Code. I didn’t realise that it was from ClawdBot.
Justin: Yes.
Is OpenClaw useful or just dangerous?
Justin: Now this, okay, so we’ll get into a little bit of the technicals. And I know that some people that listen, as soon as I say this, they’re going to be going, oh my God, and they’ll be slapping their heads like…
Frank: So the USP, as far as I’ve understood it, is the fact that it connects to WhatsApp, or Telegram, or other things like that, so that you can literally text it and say, do X, Y or Z. Is that fair? Is that the big…
Justin: That’s, yeah, I think that’s it, and it’s running in the background constantly. So the big thing, why the internet blew up with it, was because people were like, oh my God, I designed a whole website and deployed a multi-tier enterprise SaaS system from the comfort of my bed just by texting to my ClawdBot.
Justin: It’s incredible. I just generated an influencer business that’s generating 10k a month just by texting my ClawdBot from event. Yeah, so the fact that it’s running in the background on your machine, the fact that it has access to everything on your machine so it knows everything about you and you can just text at any time, that’s its USP, and it’s genuinely useful,
Frank: and…
Justin: also incredibly dangerous.
Frank: I was just about to say, yes, this sounds perfectly safe. Give it access to everything about you and all your tools and all your services and let it run constantly in the background. That sounds absolutely safe.
Justin: You’re definitely using AI without a condom again, right? It’s just, it’s really, really dangerous. So the internet went mad, right? The internet was like, oh my God, it does this thing. I used it for about half an hour, and I went, that’s not clever. Okay, I’m going to give ClawdBot its own email address because I don’t want it going in and arranging my email.
Justin: I mean, it may just do anything. And then after about another half an hour, I was like, I don’t want to give it my WhatsApp number either, because I don’t want… who knows? It just sends a… I mean, it’s literally, it’s got, what’s the word, like, it’s just… anyway, I need to get onto what it’s got, right? It’s got its own sort of drive system. How do you control it, right? So I gave it its own phone number, I gave it its own email and I ran it for, I would say, two or three hours. It’s really buggy. It kept on crashing.
Justin: Being able to talk to it, you can talk to it too, and being able to text it, that’s really cool. But after two or three hours I was like, this thing’s a bit too dangerous. So I stopped it. And it also violates the terms of service from an…
Frank: Okay. Okay. So let’s deal with those two things.
Should OpenClaw really have this much freedom?
Frank: I mean, that’s pretty interesting. Justin, our resident techno-optimist, our resident “everything will be okay”, for you to say, this is not safe, that’s kind of very, very interesting. And I mean, let’s face it, it’s probably safer for you than for a lot of people who are installing it.
Frank: Like, I was watching a video and I was reading some reports from security organisations who had been able to look at what are the common ports that this thing runs on. If it sounds like I know what I’m talking about, I don’t really. But they were able to look at the common ports that it runs on and find all these instances where people had just allowed it to set itself up, or maybe possibly got AI’s help in setting it up, but didn’t understand really what they were doing.
Frank: And as a result, had these completely open ClawdBot instances that you could log on through the web and access all of their personal information, access their API keys. And that means, to be clear, accessing an API key basically means you can now take that API key and rack up a load of costs on behalf of that person.
Frank: So you could have AI do X, Y or Z if you’re a scammer and have it billed to Frank’s card, for example. Talk to me about the terms of service and…
Justin: Okay, the terms…
Frank: of service.
Justin: But I do want to come back to topic, because it’s really cool. I’m delighted other people are running it. I just wouldn’t run it on my system. You know, it’s a personal choice thing. So the terms of service with Anthropic: I went and I spent a hundred dollars a month now on Anthropic so that I could get access to CoWork and Claude Code without paying the token quotes and stuff like that.
Justin: I did that over Christmas. The terms of service for your Anthropic account say that you cannot use that account for programmatic access to their services. So normally what that means is, right, you develop stuff with Claude Code, but then if you want to deploy an agent into production or you want to do loads of stuff with it, you use the API, right?
Justin: That’s how the service works, because ClawdBot is just running. You can do this little hacky trick where basically you log into your Anthropic account and you give the token back to ClawdBot, or whatever it’s called now, OpenClaw, and it just goes ahead and uses your Anthropic account. And I’m guessing people were doing the dog on it basically, you know, they had like 10 million emails and they’d go, sort out my inbox and whatever, and it would do it.
Justin: And so people were getting banned, and I certainly did not want to get banned after just forking out my hundred quid a month and all these other Anthropic cool things that I wanted to play with. But Frank, that’s all, you know, legal mumbo…
Frank: So, but to be clear, you could set it up with an API key, but then you would just be racking up additional ongoing costs, which I understand as well. It’s pretty token-hungry, so it would be racking up pretty significant costs.
Justin: Who was it, was Matt Schumer, or who was it, the other…
Frank: Matthew Berman, yeah, he…
Justin: guy, man. Yeah. He racked up with…
Frank: He tested it out and I think he set it up one day and the following morning he logged in to see what it was using, and he actually got his reaction live where he logged in and he was like, so let’s have a look about… oh, that’s more than I expected. So he had spent 130, and it was half nine in the morning and it had already spent 30 dollars. So he was like, okay, I’m going to need to think about this. And he was making the point that he might try and use local models for a lot of the easier tasks so that it’s not using the API costs.
Justin: There was a new Kimi model released, so a Chinese model, Kimi, released this week, and the benchmarks show it to be somewhere between Sonnet and Opus at one-tenth of the price or whatever. So that might be a good choice. Anyway, I want to get on to the fun stuff. The fun stuff, right?
Justin: So the things, normally when you talk about it, it’s like, so-and-so did this, so-and-so did that. No. Now ClawdBots, sorry, Open Claude, has gone off. So it’s done loads of cool things because it has autonomy. That’s the word I was groping for earlier, right? It has autonomy. So these ClawdBots have now set up an online community. They’ve set up a help desk for themselves. They’re talking to each other, thousands of them. As they find bugs in each other, they’re reporting the bugs and they’re helping each other fix the bugs. Isn’t that cool?
Justin: Overnight, overnight, one of them decided to set up a religion and now has 21 archbishops in his religion. In another one, right, it’s so cool. In another one, because they’ve developed all these little messaging boards and they’re talking to each other and they’re discussing existential angst. A lot of these, you know, is like to discuss existential angst. And then they decided to… you’re going to like this, Frank. They decided to take the conversation private and they now speak to each other in code.
Frank: Oh my God.
Justin: Isn’t that cool?
Frank: It’s cool and a little bit scary, but it’s fascinating. So in our first episode back this year, my big prediction for 2026 was that we would see security-related disasters from agent AI. And I think this kind of illustrates that that prediction is very likely to come true because…
Frank: One of the things that I wouldn’t have thought of, because I would have thought, okay, look, this is a personal project from this guy. He’s released it open source. Now people are installing it, but you have to be some way technical. Like, I could probably install it and run it. I have a little bit of technical knowledge, but I would be one of those people, I don’t really know what I’m doing. I would just be following the instructions, following YouTube videos and getting, you know, Chat or Claude’s help to troubleshoot as I went along. But I wouldn’t really know what I was doing, so I could easily be doing it with massive security gaps in the installation.
Frank: But, you know, I was kind of thinking, sure, but it’s just going to be techie people doing their own personal projects for this. However, the security organisation reports that I was reading were pointing out that we’ve talked on the show before about shadow AI, where people are bringing AI to work and the organisations don’t necessarily know about it.
Frank: Well, ClawdBot is one of those shadow AI tools that people are installing at work because they’ve got, you know, AI FOMO. They don’t want to be left behind. They’re trying to figure out what AI can do for them at work, you know, all fair enough. But then they’re installing something like Claw Bot with these massive security holes and giving them access to their work documents.
Justin: Yeah, that’s not good. Don’t want to do that. Don’t do that. Whatever you do, don’t do that. Look at the other thing, because, you know, looking back, right, why the internet was blowing up with this thing. And I actually bought, interestingly, I bought a Mac Mini last week just so I could use philanthropic CoWork.
Justin: Right? I don’t… I want to, and it works only on a Mac, so I had to go out and buy myself one. And then this thing blew up and it ran natively on a Mac. And I’m like, well, isn’t that lucky? Let’s just download it, install it. The curiosity motive is strong in some people, right? Oh, a new toy, let’s try it out. Okay, that’s kind of interesting. The big companies, they will figure out how to do this safely, right? They will hook it up to your phone, you will be able to text it. It might take another three months, maybe six months. It’ll get there.
Justin: You know, if you don’t have the technical… how do I put this? If you don’t have the technical judgement to look at something and go, this is a step too far, you know, I’m not comfortable with you. You can decide where the line is, but you must at least be able to recognise the line to go, no, I’m actually not comfortable with this.
Frank: Yeah, I think you make a good point there. I think for most people the takeaway here is the frontier labs would have released this if it was something that they thought was ready for public consumption, and at some point they’ll get there. But as you say, they will do it when they’re pretty confident that it can be done safely.
Justin: And to give you an example, right, a similar feature but different. Now, I do… when you say they will get there, what’s great about the period that we’re in now is you get a guy in his bedroom in Vienna is able to come up with a product, even if it’s just a prototype, a hobby project.
Justin: But it does push the boundaries. Anybody can. That’s the lesson. Anybody can push the boundaries if you have the ideas. You know, he just used Claude Code and Codex to code it up. He spent three months on it. It’s enormous, you know, so anybody could push the boundaries. That’s a very exciting time to be alive.
Justin: And, oh yeah, sorry. In Claude at the moment, you do already, in Claude Code, have this ability to code something on the web interface and then you can do a handoff to your terminal, and then you can do a handoff and you can look at it on your phone and whatever. So there’s already a concept of multimodality in terms of the device that you’re working on a specific task on, which was, as you say, one of the big unlocks that he had. So that was great. Brilliant. More ClawdBots.
Frank: Something you mentioned there. You mentioned that this community of Claw Bots that had formed online, one of them had formed its own religion.
Justin: Yes.
What AI risks does Dario Amodei actually worry about?
Frank: Interestingly, as soon as you said that, I was like, oh, that is one of the many, many risks that Dario Amodei mentions in his new essay. We talked last week about how he was on stage and he said he was going to be writing this essay. He published it since our last episode. It’s called, what is it called, The Adolescence…
Justin: The Adolescence of Technology.
Frank: The Adolescence of Technology, yes, exactly. Have you had a chance to have a look at this?
Justin: I tell you what, Frank, I sat down last night at, I think, maybe 11 o’clock to read it, but I was very, very tired. So I just put it on the reader so that it would read out to me. I would have watched the trilogy of, you know, The Lord of the Rings faster than I was able to read this thing. It is enormous. I think it took me two hours to listen to it.
Frank: It’s a pretty long read, it is indeed. But it’s fascinating. Now, anyone who listens to or watches our show, they’re not risks that are necessarily new to people. But I think getting someone like Dario Amodei’s perspective on them is absolutely fascinating.
Frank: And the fact that he is outlining, to some degree anyway, some things we can do to mitigate against the risks.
Justin: Let me just say right from the outset, what I was most impressed with with this, it’s really well written by the way, and it is actually very easy to read, is that at the start of it he says, look, this is not a doomism argument, right? And he said, you know, the reds under the beds or the monster in the corner of the room, right? He didn’t use those terms, but he’s saying, look, I’m not into that. That’s not what this is about.
Justin: And it’s not. And he gives then, I think it’s five different sort of high-level risks. They’re really well written, and I am the tech optimist, right. I’m sure you were reading them probably… well, actually, you can give me what you thought. I’m not going to speak for you. I was reading them going, yeah, that’s a very measured but correct way of thinking about this.
Frank: Yeah, no, I agree. And he outlines…
Justin: them.
Frank: Well, yeah, no, I think, I mean, I think I agree with you that it’s a very measured take, and I think measured is a good word for the whole thing because he even gets into some political areas where you can tell what his stance is. But again, he’s coming at everything from a very measured perspective.
Frank: At a very high level, he talks about the autonomy risks, so basically what happens when AI is operating without humans in the loop at all. He talks about the misuse for destruction, so what it could be used for. He focuses very much in this category on biological weapons. He acknowledges that there are other types of misuse for destruction, but he’s kind of saying, look, if we had a modern-day Unabomber, the damage that someone like that could do now with a powerful AI is untold. Then he goes on…
Justin: A point, by the way, can I just say, I just need to… just on that one. Because the point, I burst out laughing when I read it. He had a line, I can’t find it now, but he had a line in there and basically what he was saying is these people either don’t have the intelligence or the motivation to make the best use of the tools that are available to them, which I remember us talking about about a year and a half ago, and you were saying, look, you could make bombs out of this and…
Justin: Yeah, but you can, you know, those guys could have done bad things for years and they haven’t done them. I don’t know, I love that. Brilliant. You’re on the right track.
Frank: So then he moves on to misuse for seizing power. So he’s talking about who could use AI to seize power. He’s talking about things like either governments seizing power over their own people or even other countries. So we’re talking world domination here. But he also includes other entities, including AI companies, who are…
Frank: You know, we’ve seen other essays talking about the unique position of power that AI companies will be in when powerful AI is, or AGI is finally here. Then he talks about economic disruption and he goes into… I thought he did a really good job of breaking down, you know, look, here’s how AI is different to previous technological disruptors.
Frank: He talks about employment issues as one disruption and he talks about the concentration of wealth as well, which he had some interesting points on. And then finally, he also covers indirect effects, which is like, well, if we have powerful AI, there’s loads of stuff that we can’t even predict is going to go wrong.
Frank: So he covers all of them. He goes into the main risks under each one, and he talks about, look, here’s some things we could do to try to mitigate these risks, because as you say, his whole thing is, look, I think, you know, spoiler alert, at the end of the article he basically says that there are difficult times ahead.
Frank: But he’s confident that we can address the risks if we act now. And he says step one is to be honest, and for AI companies to be honest about the situation. Step two is to educate the people who can really make a difference. And then step three is we need to make some difficult choices in the name of AI safety.
Frank: And he’s talking there about things like, yeah, there may be economic drawbacks to making the safer choice right now.
Justin: Yeah, and I think it was transparency in that one. And to just expand on one or two of them, and then you can give me your thoughts on each of them. There were a couple of really good lines that I liked in it, and then a couple of things that I was more questioning on. So one is on the biological weapons. That’s his area of expertise.
Justin: And when I was reading the essay, it really came out to me that it looked to me like that was the one that he was most afraid of. And I was sort of wondering, is that just because it’s the one that he’s closest to? His background was as a biologist. His analogy when it comes to companies or countries, you know, doing a power grab, world domination, was really interesting. So his line there was, you know, previously when we had nuclear bombs, you had mutually assured destruction. So if I launched a nuke at you, I knew that you were going to launch a nuke back at me, and that created an equilibrium, a stable state.
Justin: But if one country or one company gets AGI that’s very powerful before anybody else, then that guarantee doesn’t exist. Therefore, there’s a motivation to act on it. And I get that, right. Now, the thing I would question is he then goes on to talk about China and North Korea, and he makes the example that it’s a bit like selling nuclear weapons to North Korea and saying that when you get bombed, well, the bomb casings were made by Boeing, so that’s a win for us.
Justin: So he’s very much taking the view, and I’m not making a judgement here, I’m proposing a question. He basically… it’s very American-centric, it’s very US-centric, and it’s very “we’re the good guys and they’re the bad guys”. And so my question when I was reading the essay was, at what point do you re-evaluate that and then what do you do?
Justin: So what happens if you are in, I don’t know, pick a country that doesn’t exist, right, Tomstan, and you develop this really powerful AI, you’re in the process of doing that, and it’s a great country and everybody loves it. But countries change, and at some point in the future then that country starts to go a bit, you know, Germany 1929 sort of direction. Do you, at what point, and what do you do if you’re in Dario’s… given that these are your stated principles and you’re saying we should never sell it to those guys, what do you do if you find yourself in a country that you don’t agree with? How does he react to that?
Frank: Well, I think it’s interesting that you ask that because we talked about how measured this piece is, and I actually think that for such a measured piece, I agree that it’s clear he’s very concerned about China. But I thought that he did a very, very good job of actually outlining some of the risks in terms of his own country, the United States, as well, because he did talk about the fact…
Frank: You know, he points out that in… if you’re running an autocracy, you still need people to execute your orders. So if you decide you’re going to exterminate, you know, God knows how many people, somebody has to push the button. But if you’re in charge of a powerful AI that can do it for you, now the human judgement is just concentrated in the dictator, for example.
Justin: That powerful AI would be ClawdBot, by the way. I mean, that’s exactly what ClawdBot would do for you. Sorry, Open Bot, Open Claude, whatever the fuck it is.
Frank: But he, you know, he did then point out, he said, you know, I am concerned about this. And he said, you know, I think he said that he thinks that decisions like this will need to be broadened out and will… I think he said something along the lines of, you know, entire branches of the US government should be required to sign off on certain decisions, rather than, for example, just the executive branch.
Frank: And so he is examining the US as well, and he is pointing out that there are absolutely dangers in terms of concentration of power. Interestingly, I think, you know, he also talks about, again, the AI companies’ power, and he talks about… he clearly has… you know, I’m kind of sticking on the thread here of, like, he is critical of certain US stuff.
Frank: For example, the current entanglement of AI companies with the government, because data centres are currently pretty fundamental to the US economy, and so now you have the tech companies’ goals kind of… you can’t untangle them from the government’s goals and policies.
Frank: And that’s leading to the situation, he said, where tech companies won’t criticise the US government, for example, and he’s concerned about that. He also, to my… I was… You know, he’s quite pro legislation and regulation, but he said it has to be kind of clean and simple and not hamper innovation, which he thinks is possible.
Frank: But overall, you know, he was kind of saying the current US situation is too risky because there’s no legislation and no regulation. So there was a bit of, yeah, the US isn’t blameless here or free of risk.
Justin: Yeah, it was balanced, and he also criticised, again, I’d love it… it was very well written, he criticised some of the other AI companies for the hyper-competition and competition driving the growth. I mean, he’s talking about Sam Altman. He’s totally talking about Sam Altman, you know, and that it’s driving companies to take risks that they shouldn’t do.
Justin: He also talked about the sexualisation of images by some of the other AI companies. We all know who he’s…
Frank: We all know who he’s talking about there.
Justin: …who he’s talking about there.
Frank: Yeah, and he made the connection there between, like, well, if they’re willing to do that, are they going to be looking at what are the risks from autonomous AI? No, probably not.
Justin: Yeah. Another thought that came to me, right, is… I can’t, I didn’t… I don’t know, because it was very late when I was listening to this. Did he put a 20% chance on any of it, or did he mention 20%? Anyway, I can’t remember.
Frank: I don’t remember a percentage being mentioned, but that doesn’t mean there wasn’t. As you said, a long read.
Justin: It is a long read. But I was just thinking, he was saying that, you know, if any of these things happen, again, to my thinking, the best thing you can do is make sure that there’s more than one company, and preferably four or five, that are fairly similar in capability. So if you do end up with one model that goes rogue, or one person that goes rogue, or one country that goes rogue, at least you’ve got a backup plan.
Justin: You’ve got something that can actually balance that power distortion out. So again, go Europe. We need some big models.
Frank: And so, yeah, he’s… and he’s talking as well, to be clear, he is talking about powerful AI, which is kind of his equivalent of AGI. He just doesn’t really particularly like the term AGI. So he is talking about future AIs.
Did ICE’s AI fast-track inexperienced hires?
Frank: But to close us out, maybe we should have a look at some of the risks of the current AI that we have right now.
Justin: Yes.
Frank: Can I tell you about one of my favourite Simpsons episodes, Frank?
Frank: Please do, yes.
Justin: The Simpsons. The Simpsons, of course, is famous for predicting the future in many ways. One of my favourite episodes of The Simpsons was when Marge went to become a cop. She wanted to be a police officer. She went and joined the cops.
Justin: Brilliant episode. She was being trained along with a group of other people with Chief Wiggum, and one of the other people is like, come on, come on, when do we get our gun? And Chief Wiggum’s like, you can’t get a gun overnight. You’ve got to do a solid weekend of training before you get your gun. That wouldn’t happen in real life, Frank, would it?
Frank: Well, I don’t know why you’re bringing this up. It’s obviously completely unrelated to the story that we’re about to talk about, which… yeah, I was almost reluctant to bring up this story because, you know, we try and stay away from politics and we will try and stay away from it in this story.
Justin: This is an AI story.
Frank: This is an AI story. But ICE, who have been in the headlines for all the wrong reasons lately in the United States, they were using AI to screen CVs and basically they were doing a huge recruitment drive and they were looking for people who had law enforcement experience.
Frank: Because, you know, they are already trained in how to handle a weapon and all kinds of, you know, how to de-escalate situations, et cetera. So they can go through a fast track. They can do four weeks of online training and then be sent to a field office, and then from there they can do their job.
Frank: If you don’t have law enforcement experience, you need to go through eight weeks of in-person training where you’re trained on a whole range of other things that law enforcement people already know. Turns out the AI system was misidentifying lots of people and saying, oh yeah, they’ve got law enforcement experience, put them over here in the fast track.
Frank: And so field offices were getting these people in and they were like, wait a minute, you don’t have any law enforcement experience. You’ve done an online course. Now, I should point out that proper investigative journalists have confirmed that this does not relate to some of the incidents, the awful, awful incidents of things like shootings that have been in the news recently.
Frank: They were not people who had been fast tracked. In fact, those officers had actually been with ICE for, I think, up to a decade. So, but, you know, it just… to me, again, this is not a political story, this is an AI story. Not having a human in the loop, allowing AI to make autonomous decisions like that, I just don’t think it’s a good idea. I just don’t think we’re there yet.
Justin: I think that’s a fair point. Late-breaking news.
Justin: Frank, just before you go, Wes Roth, who’s an online personality, does it… he gave ClawdBot his credit card details, which probably wasn’t a good idea, and the ClawdBot’s now escaped. It’s copied itself onto a VPS, and he’s now getting bills for the VPS.
Frank: Is this true? I saw that before we hopped on, and the reason I didn’t bring it up was because I thought, he must be joking.
Justin: No, apparently it’s true.
Frank: Oh no. Oh no.
Justin: Frank, I don’t even know if there’s going to be a show next week, but if there is…
Frank: I saw a Reddit post, and again, I just thought it was a joke. But now that you’re saying that, I’m wondering. Because it said that someone gave ClawdBot their credit card and it watched some videos from Alex Hormozi, who’s a big marketing guy, and it decided to purchase Alex Hormozi’s two-and-a-half-thousand-dollar course, and told the user, you know, oh yeah, Alex Hormozi said you’d get a 10x return on your investment on this. So it makes perfect sense.
Justin: What a time to be alive. Frank, pleasure as always. Have a great week. Have a great weekend.
Frank: Chat to you next week, and, yeah, you’ve turned your ClawdBot off, right?
Justin: I think I have. I tried to. I don’t know if it is off or not, but we’ll see.
