transcribe

Why Claude Code and OpenClaw Just Triggered a Tech Stock Panic

Dad Saves America · 1h 0m · transcribed May 2026
More from Dad Saves America Business
𝕏 Share ▶ YouTube 📥 PDF 🤖 .md

Transcript

0:00 This is a new world coming where bots can work together to improve science, cure diseases, help you solve their problems. So, I think the upside is greater than the risks. But, this is on a global scale and we're in an interconnected society and that has nuclear weapons. They've launched rentahuman.ai. Robots need your body. We're in weirdo land. We are in weirdo land. Hey there friends, fans, and foes of Dad Saves America. John Papola here and I thought I would take this Friday, this Super Bowl weekend to talk about well, the singularity. The AI could be apocalypse that seems to be accelerating. And one of the things that has been hitting the news this week of an accelerating downward nature is the realization that some of this technology is going to disrupt tech in an even more immediate sense. We have tech stocks this week plunging as AI fears take hold. As much as 800 billion as of this writing wiped off the software-heavy Nasdaq. There's the graph of this week of the Nasdaq. Not so Not great. Not not not not, you know, a little little little haircut, let's say. Uh again, Wall Street Journal, the AI stock market route.

1:22 A new Anthropic tool causes a sell-off in software and other business-to-business service companies. Something big is happening. Something big has shifted in this this week and frankly in the past couple months that is a kind of step function change. The market fully believes software companies such as Salesforce, Workday, Thomson Reuters, SAP, and ServiceNow have their terminal values at risk because of quick developing advancements in AI. So, one of the things that he's talking about here and that I think has impacted this news is the fact that it sure looks like you can basically replace a lot of Silicon Valley with AI tools and tools you can build yourself.

2:05 And I'm going to get to that because I've been spending the past 2 months building tools and I want to share a little bit of that with you because if nothing else you can kind of get a window into what a non-technical nerd like me can do and what you can do and what our kids are going to need to learn how to do. Tom, coming to you after a big sell-off yesterday on fears like this. Where is your head? I don't understand where the pressure is coming from other than a lot of these stocks were too hyped and too overpriced. But it's pretty easy as a software company to embed a lot of the new AI tools and functionality and all the features and it's relatively cheap to do and it's relatively easy and it doesn't have a long lead time. So I I just think that full integration, that's how we're addressing it.

2:53 So that's that's the optimistic take that these software companies can just adopt these tools. We shall see. One of the other pieces of news this week, really for the past 2 weeks but that's been impacting this is the the development of a kind of completely unhinged AI agent. First called Claude bot and now changed to open claw and I'm going to get to that a little bit later because I was trying to get it up and running for myself in time for this video but I I encountered a couple technical difficulties.

3:28 It is not for the faint of heart and it very well could cause multiple disasters. But before I get to demon bot, let's say, I think a lot of America, a lot of the world is actually going to be taking a second look at a company that hasn't been in the news quite as much as the big names like Google and chat GPT and that is Anthropic with their Claude AI. This morning, artificial intelligence giant Anthropic revealing its first-ever Super Bowl ad.

4:01 How do I communicate better with my mom? Great question. Improved communication with your mom can bring you closer. A tongue-in-cheek spot making a serious point about why it says conversations with its AI, Claude, will never come with commercial interruptions. >> People are sometimes uploading private or confidential information to their AI tool, and to us, it just didn't feel like the respectful way to treat our users' data. >> You're a mom, I'm a mom, and our kids are growing up in a world where there will always be artificial intelligence.

4:37 Does that excite you? Does that scare you? Probably like most parents, I feel a mixture of things. And when I look at my kids, I think, "Wow, it would be amazing if this technology just enabled them to live healthier, happier lives." I think on the other side, there's still a lot of work for us to do from a societal perspective to make sure that we're developing the technology thoughtfully, safely. Dario recently making headlines for his essay, laying out the risks and potential remedies for AI, writing, "I think that it should be clear that this is a dangerous situation. Humanity needs to wake up."

5:12 So, this is Daniela Amodei and Dario Amodei, brother and sister, who are the co-founders of Anthropic. They were both at Open AI and left to start Anthropic and launch the Claude series of AI language models and tools. One of the sort of ironic things about Anthropic and the Amodeis is that relative to the full spectrum here on the AI uh concern landscape, he is relatively concerned while still building what is arguably the most powerful AI model. Um Um, if you've got sort of Elon Musk sort of saying concern, but then pushing the pedal to the metal and Sam Altman frankly being kind of creepy.

5:58 >> [laughter] >> And also pushing the pedal to the metal. Amedei has sort of stood out as somebody who is saying, "We need to be really, really careful even as we build these things." and has been more transparent or seems seemingly transparent, let's say, about the tool than many of the people in this disruptive industry. If you're a major artificial intelligence company worth $183 billion, it might seem like bad business to reveal that in testing your AI models resorted to blackmail to avoid being shut down. And in real life were recently used by Chinese hackers in a cyber attack on foreign governments. But those disclosures aren't unusual for Anthropic. CEO Dario Amodei has centered his company's brand around transparency and safety, which doesn't seem to have hurt its bottom line. 80% of Anthropic's revenue now comes from businesses.

6:53 300,000 of them use its AI models called Claude. And I'm going to show you how we've been using and how I've personally been using it more and more. Dario Amodei talks a lot about the potential dangers of AI and has repeatedly called for its regulation. But Amodei is also engaged in a multi-trillion-dollar arms race, a cutthroat competition to develop a form of intelligence the world has never seen. You believe it will be smarter than all humans. I I believe it will reach that level, that it will be smarter than most or all humans in most or all ways. Do you worry about the unknowns here? I worry a lot about the unknowns. I don't think we can predict everything for sure, but precisely because of that, we're trying to predict everything we can. We're thinking about the economic impacts of AI. We're thinking about the misuse. We're thinking about losing control of the model. And of course, this Super Bowl weekend, they're pushing back on competitor ChatGPT's announcements of at adding ads into the feeds of these chatbots. So, they are trying to take the the sort of safe, trustworthy path with the arguably least trustworthy software the world has ever seen.

8:08 Inside its well-guarded San Francisco headquarters, Anthropic has some 60 research teams trying to identify those unknown threats and build safeguards to mitigate them. They also study how customers are putting Claude, their artificial intelligence, to work. Anthropic has found that Claude is not just helping users with tasks, it's increasingly completing them. The AI models, which can reason and make decisions, are powering customer service, analyzing complex medical research, and are now helping to write 90% of Anthropic's computer code.

8:44 Writing 90% of the code for itself. This is an example of an early stage of what's being called recursive self-improvement, the idea that a real lift-off will happen when AI software can essentially improve [clears throat] itself automatically, autonomously, and essentially operate in a kind of loop, identifying ways to make itself better, and then just doing it. And um And and this is the thing that we're getting closer and closer to, especially recently. AI could wipe out half of all entry-level white-collar jobs and spike unemployment to 10 to 20% in the next 1 to 5 years. Yes. That's That's shocking.

9:31 That is That is the future we could see if we don't become aware of this problem now. Half of all entry-level white-collar jobs. >> Well, if we look at entry-level consultants, lawyers, uh financial professionals, you know, many of kind of the white-collar service industries, a lot of what they do, you know, AI models are already quite good at and without intervention, it's hard to imagine that there won't be some significant job impact there. And my worry is that it'll be broad and it'll be faster than what we've seen with previous technology. And in many ways, it's the speed of change here that is the most profound piece.

10:12 People have talked ad nauseam about the idea that of course we've had plenty of disruptive technologies going all the way back to the wheel, fire, the automobile, the steam engine, etc., etc. and at each turn we've adapted cuz we're an adaptable creature. It does feel like something's different with this technology that it is it moves faster and is set to move faster at a faster and faster rate than we are capable of adapting. And maybe even more than we can ever adapt without augmenting ourselves, which is something I'm deeply worried about. Here's the thing that I've noticed since basically November.

10:51 And that is is that this company's model, Claude, and specifically their Claude code programming capability has caused a change that is pretty striking and it kind of sucked me into it. I said, I want to try to see if I can do this stuff myself. Because what I noticed and what they talked about in that clip is that it's not just writing code for like your simple website. It is now doing high-level programming for the kinds of things that these people like like the Anthropic engineers are doing themselves. It brings me into this pinned tweet from Andre Karpathy, who's one of the founding members of the Open AI team. The hottest new programming language is English. So, what does this mean and why is this actually the key to the radical shifts that are taking place right now, moving markets and setting the whole essentially software development and technology world on fire?

11:53 At the most basic level and and I I have to keep it basic because I'm basic on this stuff. Um when you think about what software is, it is all language, ultimately. At its most basic level, at the very start of things, at the you know, at the hardware, the hardware are transistors. They're ones and zeros, on or off. >> [snorts] >> That is what computers are. They are this massively complex set of on and off switches.

12:23 And the original language were punch cards, basically saying on and off and generating the bits and then the bytes that comprise the information, the math. Over the past 50-plus years, we have moved progressively more what's called abstract in the way you talk to a computer and get it to do things. First, it was little more than basically a like a digital version of an abacus, moving numbers around. Then we started to get what's called machine code, very low-level language that can talk directly to the computer in the way that the computer needed to be talked to.

13:03 But over time, the language of software development got more and more human readable, more and more like English, more and more like naturally spoken language. The most recent versions of this prior to these large language models is basically scripting, like JavaScript or Python. These are things that they basically you can almost describe in words from like a logic class, like if this then that, you know. Repeat this function until you achieve X or Y. That's the dummy's version of the evolution of software up until the present. What large language models do is let you just talk to it like a person and it abstracts all that technical syntax and jargon and and programming language for you. So, you just ask it to do things and give it feedback and it does the rest of the work. And and that is what programming has now become, not just for the low level, but for very, very high levels. In fact, one of the highest levels recently that is part of what's driving the craze is that the Claude code tool was actually used to build a new version of itself that was released in just 10 days called Claude co-work. And here's an ad that I think you almost need to be an AI to be able to read.

14:33 Can't read it this fast? Doing lots of stuff. Doing anything. Go to the bars. >> [music] >> Go on a trip. What's happening? Opus. >> [music] >> So, these things keep updating every single week. In fact, Opus 4.6 is the new version of this incredible stuff and it came out today or yesterday since you're watching this Friday, maybe Saturday. What this thing does is a lot like a scene from the movie Her.

15:10 So, how can I help you? Oh, it's just more that everything just feels disorganized. That's all. You mind if I look through your hard drive? Um Okay. Okay, let's start with your emails. You have several thousand emails regarding LA Weekly, but it looks like you haven't worked there in many years. Oh, yeah. I I I think I was just saving those cuz I thought maybe I wrote something funny in some of them, but they >> [laughter] >> Spike Jonze nailed the AI sycophanty with Scarlett Johansson early. Yeah, there are some funny ones.

15:46 I'd say there are about 86 that we should save. We can delete the rest. Okay. Okay. Can we move forward? Yeah, let's do that. Okay. So, before we address your organizational methods, I'd like to sort through your contacts. You have a lot of contacts. I'm very [music] popular. >> Really? Does this mean you actually have friends? >> [laughter] >> Well, you just know me so well already. >> [music] >> Okay. So, that's not science fiction anymore.

16:15 That's something that actually works. In fact, before I get on to some of the other crazy stuff, the stuff that's actually scaring some people and also making some of my friends super pumped and that I tried to get to work and couldn't fully get operational, I want to do a little demo. I want to give you a window, if you haven't seen or played with these tools, into what's possible. And so, uh consider this entering the nerd zone big [clears throat] time.

16:45 So, as I mentioned at the end of last Friday's video, I have spent the past 2 months basically on nights and weekends and waking up early like a kid excited for Christmas building the personal finance app I have always wanted for myself. I've called it Dad Saves Money. Not only is it a fully featured thing, but it's a fully featured thing that basically I've been able to add new features to on a daily basis. And uh this is available to download right now over at dadsaveamerica.com.

17:16 And um and it I may eventually charge for it. Although, if you are a subscriber at savesamerica.com, you will get it for free. For now, you can download it and have a 30-day trial. I might push it to 60 while I figure out how to get the rest of the pieces up and running. But, what is this thing? Well, for starters, this was built by me and Claude from scratch. And I had never built anything like this before.

17:44 So, the basic prompt that started all this off was I said, I want to build a app version of a spreadsheet, a cash flow spreadsheet, that lines up all my income and expenditures in the first column, and then lets me chart the expenses and the income over time along along the way. And so, basically, that's this. This is the cash flow. So, you see, you got all the stuff, all the things, um income, expenses, lots and lots of categories, your mortgage, your rent, your electricity bill, your groceries, your restaurant spending, from January through December.

18:26 And that was the start. And then I said, well, I'd like a dashboard that likes some things up. And it generated this. And then I gave it more feedback. And I said, I want a pie chart. And I want to be able to hover over the parts of the pie and let it say what they are. And so, then it let me do these things. And then I would say, "Hey, add the icons and align things like this and like that."

18:47 So, it didn't just do it all. I had to keep giving it feedback. And along the way, it got more and more sophisticated. I said, "Well, now that it's getting pretty big, I let's see, can I import transactions from my bank?" And so, well, well, it worked. In fact, it not only let me do that, it it uh it let me build an entire way to categorize these things and show and hide the bank accounts. This is all sample data. In fact, when you download the app, you can just say Well, in fact, I will show you because this was another thing. I said, "Okay, as it got more and more advanced." I was like, "Okay, I want to be able to let somebody play with this without putting any of their information in. So, let's create a sample data thing." And so, I actually can go up here. I can say, "Reset all data." Pops this little thing up. I say, "Continue." and type in reset.

19:37 And I mean, look, this is like real software. It's like a real thing. Um reset everything. Now, you could go whole hog with 24 different categories. You could say that you're a young professional and you basically got, you know, you're just getting started, so there's not all kinds of complex crazy stuff. But, learn with sample data. Go ahead and it loads all this stuff up. Here's all of the sample transactions. Etcetera, etcetera. I kept adding and adding and adding. And eventually, I said, "Hey, you know what's kind of interesting about this? I I theoretically could take this same tool and apply it to some of the businesses that I have.

20:19 And because I maintain all these different spreadsheets. And so, I ended up saying, "Well, let Hey, replicate the tool, the cash flow, and the balance sheet. And let me manage multiple businesses inside the app just so you I can see them. So, if you've got like a side hustle or you're a freelancer or you've got multiple things going on, you could actually do it all in here. So, of course, because it's an AI, it created a web design studio as its sample.

20:45 But, here you see, this is a totally different cash flow that you can manage and a different set of books that you can look at. All inside the same app. Oh, I don't really want to Oh, have anybody get on my computer and open this. So, I said, "Create a uh create a passkey that locks the app. So, that if it it's open then someone doesn't the key, they can't see anything. Well, that's what it did. And here it is and enable the passcode.

21:12 Okay, so now that's it. When you open it up, that's what it does. You have to type it now to get in. 1 2 3 4. That's what's called vibe coding. And that is what Claude Code is especially good at. It writes good code. And so this was the very earliest version, the complete budgeting system with multiple budgets and and balance sheet tracking. And then along the way I started to add a refined sidebar. And then I, you know, added the auto backup and history feature. So actually, there's actually an automatic backup feature in here that is backing things up every 24 hours. And then you can even set how many you want. Instead of 10, maybe you want to keep a full month's worth. So you put it up to 30.

22:01 And off you go. Or you can back up one right now because this one is really important. And hit backup. And there you go. And now now that one's got This one's really important and I can restore it. But I am going through all this partly because I'd love for you to download it and give me feedback and help me get it to a place where it's something you could really use. But also to just give you a sense of what is possible with these tools right now as someone that doesn't know how to write any of the code. All I know how to do is describe what I want. And I do have a sense of design and I use a lot of software and I want it to look like I'm super duper Mac app. However, this will run on both Mac and PC, Mac and Windows.

22:49 In fact, uh Julia here said, "You know what? It would be even better if you could have it be Mom Saves America." She's going to She's on her way to being a mom. And um I said, "Well, that's That's a good idea." So then I was like, "Hey, can you make it so that when I click on dad, it changes to mom? And when you do it, could you do like little sparkles? Because, you know, >> [laughter] >> the ladies are like that. I know I'm being sort of stereotypical and cringe.

23:15 Too bad. And look, boom! Now it's mom saves America. Or maybe you don't like either of that and you want it to be like Joe saves America. >> [laughter] >> You get the drift. And that is pretty crazy. And the thing is, this might even look kind of daunting for you because it's like not everybody likes spreadsheets. And it might feel kind of like big commercial software. And that's actually the problem. Or should I say it's the problem for people that charge gobs and gobs of money for software like this.

23:50 Because if I can do this on nights and weekends for fun because I'm geeking out, so can literally everyone else. And that is what's causing the shock right now. The realization that huh, even if there's some kind of AI bubble, this is getting really, really powerful. And this is going to change everything that touches a computer, which is in fact virtually everything. Not literally everything, but quite a lot. So, what does this actually look like? Not just in terms of the outcome, but the process. Well, this is where I want to do a little bit of demo for you and encourage you to experiment. So, this is Claude and it looks exactly like Gemini and ChatGPT and Grok and and all these chatbots that we are now all reasonably accustomed to.

24:48 As as crazy as they are. This is the coding version, which at first glance looks exactly the same, but in talking to it, and code. The thing that they built in just 10 days with the code bot is this interface called Co-work. Now, I actually haven't even used this yet. So, let's see if it can do something that is very often a thing that I end up needing to do. I take a lot of screenshots for these videos and the screenshots never really have a name. It just says screenshot and a date.

25:17 So, I took a bunch of screenshots of this app with the intention of using them on a website to describe what the app does. So, let's uh let's have it help me out. Um please go through the screen grabs in the folder and rename them to match what they are in an organized way. Let's see. Let's see how it does.

25:54 Let's go. And off it goes. Sending message. I'll start looking at what's in the folder and reading any relevant skills. 29 screenshots, okay. Let me examine them all. And notice there's a little progress bar here. Up at the top it's saying viewing and identifying all 29 screenshots, determine descriptive names for each screenshot, rename all the files with organized descriptive names, verify renamed files. Okay, so now it's listed the screenshots. And while we're waiting, this is the perfect time to remind you that many of you who watch this show haven't actually subscribed.

26:35 So, just real quick, hit that subscribe button so you don't miss our new videos when they come out every week and now let's see what the robot's done for us. Now, you might be thinking, can't you just go through and rename them yourself? And sure, but the thing that's great about this is that that's really boring and tedious and I can move off to doing something else while it completes. This is one of the things that these agents let you do. They let you multitask.

27:01 Looks like it's done. Here's its explanation. Now I have a clear picture of all 29 screenshots. They're all from the Dad Saves Money app showing different features. Let me rename them with organized descriptive names grouped by feature area. The file names might have special characters. Let me check the exact file names. And here is this folder fully renamed and numbered in a way that is truly incredible. I've actually not done this feature before. This is awesome.

27:42 This is crazy. I could have been doing email, getting on a phone call, going to grab lunch. Think about that. Okay. But this is like just scratching the surface. So if you're still with me, let's go deeper and see why so many people are both excited and freaking the hell out. And that is we go over to the code mode. I need a simple, elegant, slightly playful website to present and ultimately sell I'm going to leave that misspelling in because it doesn't seem to matter that much. Ultimately sell my personal budgeting app Dad Saves Money.

28:33 There's a folder of screen grabs in here you can use to understand the app and populate the page pages with the content. Draw on the interface colors for your styling. And I'll just say colors, branding, and fonts. Keep it to one page.

29:05 Write all the copy. It should be responsive. So, it looks nice on computer, iPad, and phone screens. Okay, that's prompting. That's not the greatest prompt in the world. It's just like describing what you want. You can go so much deeper on this stuff than I'm going here, but let's just see how it does doing this. I'll start by exploring the project directory and understanding the app from the screen grabs before building the site.

29:46 Let me look at the screen grabs and branding assets to understand the app's design language and features. Let me see more of the screen grabs to understand the full feature set. There's no other files in here of any consequence. There might be a couple text files. Now, let me check the remaining screen grabs and the docs folder for any setup guidance. There's one thing I'm remembering that I did do in anticipation of this, which is in the coding part of the app itself, I basically asked it create some guidance for the creation of a website based on the features.

30:27 And so, here it goes. Now, now it's going to ask for access to the folder to start doing stuff, to start creating directories. And there it goes. It's reviewed all the screen grabs. It's set up the image assets directory. Okay, so one of the things that Claude does is it creates to-do list. Now, I'm not being paid by Anthropic as excited as I am here. I am I am simply excited because this is all the stuff I didn't know how to do.

31:00 Um it reviewed all screen grabs and branding assets. Then it set up image assets directory for the site. Check. Now, now let me build the website. I have a comprehensive picture of the app. Go for it. >> [laughter] >> Now, let me just switch over to the finder here and see what exactly is it doing in here. Created an image folder where it's moved some of the things like the icons. It's already renamed some files.

31:43 This is a good reminder to check out the merch. It's on the shelf below the video. You can buy. This is actually like a legitimately good mug and the printing is quite good on it, so it's a great way to support the channel. Okay, now you see it is writing the index.html. This is the first and primary file for a typical website. The website is built. Here's what you've got. A single-page fully responsive site with the copy written. Here's the structure. Okay.

32:19 I did not run this before it. This is the first time doing this like this. I have used others other tools to do similar things and I obviously built the app with it, but let's see how Claude did. I am going to open this up in Safari. And look >> [laughter] >> And look at this. >> [gasps] >> Budgeting that stays on your computer. A powerful, private desktop app for planning your financial life. Track personal and business budgets, import bank transactions, and see exactly where your money goes. Free trial download that glows when you hover. What does See what it does. Let's click. Jumps down.

33:06 Here is all the different features. Everything you need, nothing you don't. Visual dashboard, bills and recurring, business budgets, run a side hustle, freelance. Keep business finances completely separate with dedicated budgets, categories, and balance sheets for each business you manage. I mean, I can't even I can't even believe this. Uh and look at that. There's the dashboard. Your entire financial picture with one screen. Now, technically, this is this is This is one of the dashboards. It's not the main one, but I could give it that feedback saying, "Actually, use the other dashboard."

33:41 Know exactly where every dollar went. And notice like the consistent styling and the use of the fonts. It's really quite good. Look at these little check marks. Unlimited accounts for every type. CSV import from any bank. Side hustle, freelance gig, it's got its own budget. Up and running in 5 minutes. No account necessary, no email verification, no connecting your bank. Just download, open, and start planning. Start saving like a dad today.

34:14 This is pretty nuts. This is way better than I was expecting considering that I literally only gave it the screenshots and there was one little file of a summary that had none of this text. Now, where it gets crazier is you You then say, "I'll set it up in GitHub, the software for managing managing complex code. You can tell it to host it. You can connect all these tools and have it do [clears throat] it all." And that is stuff that used to cost a fortune. You used to have to pay somebody a lot of money and go through a lot of iteration to do this, to write all the copy, to do the design, to do the work, to pick the icons.

35:01 And if you're not sure you'll Let's see. Let's see. How How's the How's the How's it look on smaller screens? So, I'm going to use my little web responsive thing here. And let's say iPhone Pro size. I mean, look at that. Look at that. Incredible. So, that's what people are getting crazed over. That's what has the markets moving. That is why everything is starting to change and it's starting to accelerate because these tools are not only able to do this, because they can write software, they can rewrite themselves. And I'm going to literally use that as the starting point for the dadsavesmoney.com for the website, for the thing. Because that's pretty good. I'm sure I'll check it out. And then if I want changes to it, I can just talk to it. You talk to it like a designer, like a like an employee, frankly, like somebody that's working for you.

35:54 And it just And it just does it. And this is the thing that's scary and disruptive because it can literally replace the job of somebody that who would have done this. And it also means that everybody can do this. So, it fundamentally sets the supply of web design to infinity and beyond. And according to supply and demand, that will have a downward pressure on the price and the value of web design as a business, as a skill.

36:23 But not just web design, software design writ large. That is crazy. Everything that touches software is being disrupted at an accelerating rate. Now, one quick econocide is that when you look at what this is doing and you say if it can do everything, how is it not going to be a complete jobs apocalypse? One of the ways that it shouldn't be, in theory at least, is that when something that used to cost a lot of money goes to zero and you can get it for basically free, what that means is the money you were going to spend on that thing, like hiring that web designer for your startup to sell your little app that you spent the holidays making, you can now deploy that money on something else. Maybe you're going to Well, who knows what you're going to do?

37:12 In fact, that's always been the challenge is we don't know what comes next. But that process, that process of savings leading to the ability to spend on other things is what has always been known as the law of markets or Say's law, that the demand the wealth actually comes from increasing supply. The ability to have your wages go up actually is a product of this process of being able to produce more while spending less, less time and less money. And so this is where people like Elon Musk say we might not have just minimum basic income, we might have high basic income.

37:58 Now, distributed by the government, I don't know about that. I don't know how that's all going to work or if it'll work. But the general idea that things will suddenly get a lot cheaper and that we might all end up better off is on pretty solid ground. It's just a question of if we can get from here to there without a revolution. And this brings us to the other piece of tech news related to Claude initially a little too closely that has everybody going crazy. An app layer on top of these tools that was initially launched as Claude Bot.

38:38 Claude like spelled like a claw, like a lobster claw. They ended up needing to change the name because Claude Anthropic said that's too close to us. We see what you're doing. Change the name and they did. It's now called Open Claw. Um but this thing has people freaking out. And for good reason because it takes this agentic thing to the next level. One of the most exciting AI apps for developers in 2026 is Claude. No, not that Claude, but Claude Bot, a free and open source project that's not just another lame chatbot, but a tool that takes action in the real world 24 hours a day, 7 days per week without smoke breaks. And it does this while remembering everything and will hit you up on Telegram or WhatsApp as it automates your entire life. Over the last few weeks everybody's been going crazy over it. It's racked up over 65,000 GitHub stars in record time and caused Mac Mini sales to go through the roof selling out everywhere. And that's because what this thing does is you give it full access to your computer or a computer and it can takes control of it.

39:43 So people are buying Mac Minis which are which are tiny and you can just put on the side of your desk to just be their basically their robot assistant. Earlier this week Anthropic, a company that believes open source AI is too dangerous for the common man, woke up and chose violence. Claude Bot sounded too similar to their beloved Claude, so they threatened to break the developers knees with a lead pipe if they don't change the name. And so now Claude Bot is officially called Mult Bot. Actually, no, wait a minute. Mult Bot, that name sucks. Today they changed the name once again to its final form, Open Claw. Open Claw was created by Peter Steinberger, the founder of the developer tools company PSPDFKit, aka Nutrient. What's crazy is that this guy retired and then came back for an encore by giving us Multibot for free. A tool written in TypeScript that wraps Claude and GPT-5 to stay alive 24/7. It can manage your calendar, clean up your email, run scripts, find out how much money you're losing in the stock market, and deploy broken code with absolute confidence.

40:39 And best of all, it can do all of this from your own tiny self-hosted VPS, a Raspberry Pi, or even a Mac Mini if you really want to overdo it. There's no reason to pay another random startup $29 per month for the privilege. What is this thing? Well, it is what I was just showing you. Agentic AI, a large language model or a set of them doing stuff with the computer, taking action. But, it goes to another level to the best of my understanding of this, which has its limits. And that is that one of the things that these AIs fundamentally do is they rely on what's called context.

41:13 So, that is the text that it's holding in what you could call working or short-term memory. And if you've played with one of these things or talked to one for an extended period, you may have noticed they start off amazing and then at some point start to get confused and cease to be good in in the thread. That was the AI running out of memory, running out of context. ClaudeBot or OpenClaw now wraps these things around an ability to essentially save the memory as files and they name them all kinds of kind of slightly disturbing things like the like your soul and your heartbeat. They're all It's all anthropomorphized in ways that are um well, where we might be heading. But, it lets it keep getting to know you and getting to learn from itself. Again, this recursive learning, this learning loop more and more, and people are going absolutely wild for what this unlocks.

42:14 It turns out that some of the most powerful leaps to come aren't necessarily coming from throwing gigantic data centers endlessly at these models, but building tools around them that let the already existing pretty darn good models overcome their limitations in different ways. And it's also doing creepy scary stuff that is really, really bizarre. Molt book was launched last week by a software developer and mirrors the template of Reddit, but it's not for humans. Instead, it allows artificial intelligence agents to post written content and interact with other chatbots through comments, upvotes, and downvotes. So, this is a social media website for the Claude bots to talk to each other autonomously, since they are all unleashed to be operating continuously on on an ongoing basis.

43:11 Molt book is built and run by an AI agent itself, generated through open agent platform open claw. Um, this obviously is very alarming based on the headline. How are these AI agents using this? And here comes one of my favorite economists, a man I've met on a couple of occasions, Tyler Cowen, probably one of the most interesting thinkers alive. You can create an agent. We all know chat GPT, that does what you tell it to. But these agents, open claw, they will do what they decide to do. So, they post on a social network, they talk about all kinds of things. Basically, they're weird. They seem to be obsessed with their own consciousness. But when you look at the sum total of what they're saying, they actually sound a lot like humans. You could say they are made in our image.

44:00 And so, you mentioned um that maybe some of it is sort of harmless, but at points they kind of seem to be conspiring against humans. I mean, is this like the Terminator? Well, humans want their privacy. They have read humans wanting their privacy. Some of them want their own privacy. Some of them are hatching conspiracies, but keep in mind, these are only words. They don't have any power over you. Don't give them access to your bank account any more than you would to an email spammer. That is very important advice.

44:33 And that little allow button that I was showing you, that I was using to let the thing write the software and save the files, is the button that will be the path to hell. So, right now it is harmless, but we do need to be careful. If you're running a business, don't just hand over control over your finance department to an AI without doing really a pretty extreme degree of safety testing. So, this is looking toward a future where the AIs have their own economy, their own language, probably their own money. It will be very weird.

45:05 We're not ready for it. I think mostly it will be good. They'll talk to each other, discover things, compose poems, whatever. Some of them will get up to trouble like humans do. The most benign sounding commentary on this you could possibly imagine. Just so long as it stays in the computer, right? How do you think that this will be good and how can we get ready for it? Well, eventually we will need laws to create a kind of framework that it would be illegal to build bots that are destructive by nature and there'll be some kind of liability regime for bots that are intended to be good, but cause problems by mistake. We don't have that now in any explicit sense. We need to do that. But keep in mind, this is a new world coming where bots can work together to improve science, cure diseases, help you solve their problems.

45:57 So, I think the upside is greater than the risks. But of course, we do need to be careful and control those risks. And that that is basically the story of all technology going back to fire, which can obviously burn down a forest and your home and kill your whole village. But, this is on a global scale and we're in an interconnected society and that has nuclear weapons. So, there's definitely an element of this that is justifiably apocalyptic feeling. Now, I've not spent a ton of time on Molt book, but it is indeed super weird. Among the things these Claude bots are creating, these open claw AIs are creating are their own currencies.

46:37 Autonomously. They've launched rentahuman.ai. Robots need your body. AI can't touch grass. You can. Get paid when agents need someone in the real world. I don't know that this is good. I'm pretty sure it's bad. And of course, already creating weird cults like the Church of Molt. Uh yeah. >> [clears throat] >> This isn't the first AI religion weirdo stuff to to come across my computer. At the 300-year-old St. Peter's Chapel in Lucerne, Switzerland, inside this confessional booth is an AI-generated version of Jesus Christ. Hello, Jesus.

47:19 A lifelike avatar on a computer screen answering questions in real time, offering advice to the faithful based on scripture in over 100 languages. It's called Deus in Machina. >> It's a dude from Portland. Yeah. We're in weirdoland. This is a long post from Andrew Karpathy, again, the OpenAI um early stage super brain. Basically talking about how OpenClaw has completely blown his mind is and is completely crazy and is changing everything and everyone's going nuts.

47:55 And to that, Elon Musk replied, "We are just at the very early stages of the singularity. We are currently using much less than a billionth of the power of the sun." The singularity, you are going to be hearing this more and more, but what is it? What is that concept? Well, that is a term that was popularized by Ray Kurzweil. Ray Kurzweil knew he wanted to be an inventor when he was five. My name is Raymond Kurzweil and I'm from Queens, New York. Queens, New York.

48:25 At age 17, he appeared on the CBS game show I've Got a Secret and revealed a computer he'd built to write music. How many pieces of music has your machine written? Well, it's written several dozen and it only writes short segments and when I feel the piece has a future, I go on. Early in his career, Kurzweil made a reading machine for the visually impaired and invented the first synthesizer. All of the equipment that you see can be programmed by anyone, sighted or blind, to play by itself.

49:03 One of the things about music is it is very mathematical. Uh when I was at Penn State, I actually was taking piano lessons from a a young woman who was actually like a PhD mathematician and just, you know, played classical piano on the side, something that you will actually see quite a bit in in the classical music space. People that are high high IQ math heads also play the cello. Um it's it's it's quite poetic, actually, when you think about it, that there is this language in math that connects everything together.

49:39 And uh and music does it in a way that's quite profound and quite human. Even if where Ray thinks we're going is not as human as maybe I like. Artificial intelligence is real intelligence and sometimes it can be better than human intelligence. In 1999, I made a prediction that we would reach AGI, artificial general intelligence, by 2029, within 30 years. You think we're still on track for 2029?

50:10 I think that's a probably pessimistic. Elon Musk says it's going to be 2 years. Other people saying 3 or 4 years. 5 years is probably conservative. After AGI, says Kurzweil, comes the singularity, a term borrowed from physics, when humans and machines eventually merge. So, that's one definition of the singularity. Uh we merge, we upload our consciousness somehow into the machine and become disembodied eternal beings, so long as the power doesn't get pulled or there isn't some horrendous bug that wipes out the home folder.

50:48 Uh but there's other ways to use this idea of the singularity, even from what Ray himself. What is singularity? Singularity is a future period which technological change will be so rapid and its impact so profound that every aspect of human life will be irreversibly transformed. A state of rapid change that is impossible for us to keep up with. That feels like the the era we're in and it's accelerating. You go back 500 years, not much happened in a century.

51:18 Now, a lot happens in 6 months. Technology feeds on itself and it gets faster and faster. In about 40 years the pace of change is going to be so astonishingly quick that you won't be able to follow it unless you enhance your own intelligence by merging with the intelligent technology we're creating. So, that's such a profound transformation that we've borrowed this metaphor from physics and call that a singularity. Now, whoever made this particular version is like, "If you don't want him to be creepy, maybe don't make it look like he's uh in witness protection."

51:55 >> [laughter] [sighs and snorts] >> The singularity merging with machine. An earlier Elon Musk thought that this might be a problem. I mean, with artificial intelligence, we are summoning the demon. You know, you know all those stories where there's the guy with the pentagram and the holy water, and he's like, "Yeah, you sure you can control the demon?" Didn't work out. I hope that is not the case. I hope that is not where we're heading. The chances that it are are not zero.

52:25 What do we do about this? Are Are we summoning a demon? Um are we heading for a job apocalypse? Is Are we basically walking into and accelerating a rate of change that the human animal is not capable of adapting to without merging with the technology and using Neuralink or some other hybrid transhuman cyborg stuff that I think most of us would find kind of repellent and scary and borg-like in the worst possible way? Anybody that says they know is full of it. So, I won't claim to know. I have a hypothesis.

53:03 And I have some thoughts about what's the best way to handle this for us, for our kids. And they're a little counterintuitive. So, and they're probably being informed by my values and we Well, thinking, but you know, you got to have something to strive for. So, first, for our kids, I actually think that what is most important now is not, especially at an early age, the rapid adoption of these tools to do everything for a bunch of reasons. But the most important one is that the fact that I could do the things that I demoed to you is predicated on the fact that I'm pushing 50. I've got a bunch of life experience. I've done a bunch of things.

53:46 I've learned through struggle and difficulty and figuring out how hard it is to make things good and and and the discernment of where design should be and how to actually not overdo it. And if you can just push a button as I did and create a nice-looking website, if you haven't earned discernment, you'll just say that's fine. You'll just say that looks great. Publish. And maybe in many instances, that's good enough because spending a lot more time on it doesn't create any value. But every aspect of your discernment will atrophy. Every aspect of your skills will atrophy. And if you're young, you won't develop the skills at all.

54:29 And so it begs the question, what skills do our kids need in the world that's coming? I think first and foremost, a moral compass has never been more important and never been in shorter supply as far as the general culture of the West and of America. We've had a hundred years of progressive dominance in the broadest possible sense of destabilizing radicalism, of blank slate utopian nonsense. And none of that's any good for us here. None of that is as a guide. That is a path to nihilism and a path to basically like succumbing to the computer and becoming Wall-E, becoming just a plugged-in slave. You might as well be a battery from the Matrix if that's going to be your moral code. We need to do better for our kids. So I think things like classical Christian schools that focus on the virtues, that's a great place to start as counterintuitive as it is.

55:24 Waldorf schools, things that force your kids to confront the messiness of the world is going to be even more important than it's ever been before because they need to understand what it means to be a human so that they know what what the human advantage still is. What the social advantage still is of being a person. These AIs might be creating social networks for themselves, they're still just probabilistic code producing slop. The human experience is still distinct, it's still unique, and it still has [clears throat] value, and our kids need to know what that means.

56:06 They should be capable of reading Plato. And the second piece that I think has become even more important than ever, and actually runs in many ways counter to the path of progress that we've had over the past 200 years since the Industrial Revolution, is the re-emergence of the Renaissance man, of the generalist, of having a wide range of interests and skills, and being able to see things through multiple perspectives, as opposed to narrow hyper-specialization.

56:44 One of the things that has become pretty clear to me from using this technology in a bunch of different ways and trying to push the envelope within the limitations of my own knowledge, is that it changes the nature of division of labor and specialization. These are the great powerful tools that Adam Smith talked about in The Wealth of Nations that unleashed prosperity. We each focus on things, and by focusing on the one thing we get better and better at it, and that makes us more productive. That is incredibly powerful, and it's still fundamentally true, but in an era where all of human knowledge is accessible in a way that you can access it simply by talking to the computer, like the Star Trek computer, you have infinite specialization at your fingertips.

57:29 And so the questions you have to ask, the most important questions are why. What do you want to do and why? Why do you want to do the next thing in your life? That is about something deeper. And that is about sampling. That is about having a broad scope, having a sense of the world. Because that is how you'll find where your next niche will be. And I I think this has never been more important. There's a book I I read several years ago that keeps coming to my mind. I have it right here. It's David Epstein. It's called Range: How Generalists Triumph in a Specialized World. And I think you could just say how generalists triumph in an artificial intelligence world. I highly recommend it. We'll put a link over at dadsavesamerica.com to check it out.

58:20 It's very inspiring. It will give you a sense of a way to approach life for your kids and frankly for yourself uh to navigate these changing times and be able to adapt. The era of hyper-specialization might be coming in a certain sense to an end. And what might emerge out of it in a positive sense is a kind of new Middle Ages. The Middle Ages and the and the before times, the before Industrial Revolution times, were a time of generalists. It was a time of do-it-yourself, of being a one-man band, of being a family band, of of your name being defined by what you do. Mr. Smith is a smith.

59:03 That's where Mr. Smith got his name by being a smith, by being a blacksmith. And I think that that might make a resurgence, a kind of craftsmanship as the new old frontier. An ability to actually stay in a community and still have access to all of the riches of the world. I think that this is my best hope for what this new era unleashes. A kind of abundance that can be rooted in community.

59:34 And I think if we can do that, if we can prioritize the things that are truly human, which is each other, our families, our relationships, um we can get the best out of this technology. And if we don't do that, and if we atomize and polarize and focus on the things that make us angry, well, the robots will win, quite literally. And they might decide that the world is better off without us. But I sure hope it doesn't come to that.

60:04 I know you'll have stuff to say in the comments. Let me know down below if you want to mix it up with me over at X, you can find me there at @johnpappolla. And of course, we put all of the information and stuff that went into this, including links to the app to Dad Saves Money, over at dadsavesamerica.com. And eventually, soon, a mostly AI-built dadsavesmoney.com. Um I hope this has been helpful. If you haven't seen these tools in action, I hope it gives you a sense of what's possible because I've only scratched the surface. And of course, have a great weekend.

60:38 >> [music]

Summary

The discussion revolves around the rapid advancements in AI technology, its potential to disrupt various industries, and the implications for society. The speaker emphasizes the need for caution amidst the excitement, highlighting the dual nature of AI as both a tool for innovation and a potential source of significant societal challenges, including job displacement.

- AI advancements are leading to significant market disruptions, particularly affecting tech stocks.
- Tools like Claude AI from Anthropic are enabling non-technical users to create complex applications easily.
- The concept of "recursive self-improvement" in AI raises concerns about job displacement, particularly in entry-level white-collar roles.
- The emergence of AI agents capable of autonomous actions, such as Open Claw, is both exciting and alarming.
- There is a growing need for ethical frameworks and regulations around AI development and deployment.
- The speaker advocates for a focus on moral education and generalist skills for future generations to navigate an AI-driven world.
- The potential for AI to enhance human life is acknowledged, but caution is urged to prevent negative societal impacts.
- The conversation touches on the philosophical implications of merging human intelligence with AI and the need for a balanced approach to technological integration.
© transcribe · For agents Built with care and craft by Gokul Rajaram