IM 816: Flappy Jeff - Harper Reed, Cluely, Bluesky Bluecheck

Summary

Episode Summary: IM 816 – Flappy Jeff with Harper Reed

Podcast: Intelligent Machines
Host: Leo Laporte
Guests: Jeff Jarvis, Paris Martineau, Harper Reed
Release Date: April 24, 2025
Recorded: April 23, 2025

---

Introduction to Harper Reed and AI in Coding

The episode kicks off with Leo Laporte introducing Harper Reed, a seasoned technologist and entrepreneur known for his role as CTO during the 2012 Obama campaign and his ventures in the e-commerce sector, including his company's acquisition by PayPal. Harper shares insights into his latest venture focused on AI-driven coding tools.

Notable Quote:
Harper Reed [04:21]: "We are building agents that are personified and a little bit unhinged. They'll do some work for you and help you out and live on your computer and maybe talk shit."

Understanding Vibe Coding and AI Collaboration

Harper delves into the concept of "vibe coding," where coding becomes an expressive and less rigid process, facilitated by AI. He contrasts his comfort with AI-generated code against using AI for content creation, highlighting the structured nature of code that allows for effective testing and verification.

Notable Quote:
Harper Reed [07:01]: "I stumbled across the nuance that I would never have AI write a blog post, but I have no issue with AI writing my code."

AI's Impact on Software Development Practices

The conversation explores how AI tools like ChatGPT and Claude are transforming software development. Harper discusses the shift from extensive spec documents to more iterative, prompt-based interactions with AI, likening the process to traditional waterfall models but accelerated.

Notable Quote:
Harper Reed [15:04]: "Instead of taking AI and shoehorning it into a service that already exists, to think, what should this product look like with these new automations?"

Challenges and Misconceptions in AI Coding

Harper addresses common misconceptions about AI-generated code quality and emphasizes the importance of context and detailed prompts to achieve desired outcomes. He also touches on the limitations of AI in handling large, existing codebases compared to greenfield projects.

Notable Quote:
Harper Reed [28:02]: "We just don't know yet. Most of the people who are doing code generation stuff right now are doing greenfield work."

AI, Automation, and the Future of Work

The discussion shifts to the broader implications of AI-driven automation on employment. Harper expresses concerns about job displacement, drawing parallels to past technological shifts like desktop publishing and emphasizing the need for societal adaptations such as Universal Basic Income (UBI).

Notable Quote:
Harper Reed [42:02]: "I think there's a reality here that as much as we are benefiting from this, there is a downside, which is I'm personally not hiring as many people. Some really scary answers."

AI in Education and Skill Development

Addressing questions from Jeff Jarvis, Harper provides recommendations for educators and students navigating the evolving tech landscape. He advocates for experimentation and hands-on learning with AI tools to build practical skills and adapt to new workflows.

Notable Quote:
Harper Reed [23:29]: "If you want to build something weird. Build a mobile app. Build an iOS app. Just try it. You can achieve almost anything with enough patience."

Social Media, Mental Health, and AI-Generated Content

Later in the episode, the hosts discuss the impact of AI on social media, particularly concerning content authenticity and mental health. Drawing from recent Pew Research findings, they highlight the disparity between parents' and teens' perceptions of social media's effects.

Notable Quote:
Leo Laporte [112:10]: "To say I have a problem with Bill Maher in general."

Flappy Jeff – An AI-Generated Game Experiment

The episode features an interactive segment where the hosts experiment with "Flappy Jeff," an AI-generated clone of the popular game Flappy Bird, personalized with Jeff Jarvis's likeness. They discuss the challenges of creating and debugging AI-generated games, emphasizing the iterative nature of working with AI.

Notable Quote:
Paris Martineau [84:19]: "Flappy Jeff rofl app. No, it's in... we're watching a video."

Reflections on Technological Evolution and Ethics

Concluding the episode, the hosts reflect on the rapid advancements in AI and their ethical implications. They express both excitement and apprehension about AI's role in shaping the future, emphasizing the importance of responsible adoption and continuous learning.

Notable Quote:
Leo Laporte [41:45]: "But what happens is it's up to all of us to maintain our mental health by tuning out the stuff that we feel is causing those problems."

---

Conclusion

Episode IM 816 offers a comprehensive exploration of AI's transformative role in coding, software development, and broader societal impacts. Through engaging discussions and expert insights from Harper Reed, listeners gain a nuanced understanding of both the promising and perilous aspects of the rise of Intelligent Machines.

---

For more details and to explore the topics discussed, visit the Intelligent Machines podcast on TWiT or subscribe through your preferred podcast platform.

Transcript

[Leo Laporte]

It's time for Intelligent Machines. Our guest this week, Harper Reed. Longtime coder, he was CTO for the Obama campaign in 2012. He has started and sold companies to the biggest and the best. But now he's all into AI coding. We're getting an insight into how AI can help everyone write code next on Intelligent Machines. Podcasts you love from people you trust. This is TWiT. This is Intelligent Machines, episode 816, recorded Wednesday, April 23, 2025. Flappy Jeff, it's time for Intelligent Machines, the show we cover the latest in, well, intelligent machines. Whether it's AI in your computer or an IoT device that seems to be thinking or a robot that's making your bed right now. Joining me as always for Intelligent Machines, the wonderful Paris Smart. Not. Hello Paris.

[Paris Martineau]

Hello, Leo.

[Leo Laporte]

With her broken lava lamp, her latest find on the streets of New York.

[Paris Martineau]

Listen, you know, sometimes you walk by a lava lamp sitting on a stoop and you're like, I must have this.

[Leo Laporte]

Did they put a sign that said free or were they.

[Paris Martineau]

Oh, yeah, no, it's a free. That's a table that a neighbor has had out for years where they, and I assume other people put free things. Usually it's largely junk, but today there was a lava lamp missing a top and now it is mine.

[Leo Laporte]

And thanks to our wonderful audience, we have a top which is a cowboy.

[Paris Martineau]

That's true. We'll see. Come on.

[Leo Laporte]

It actually is really good looking. I think that's.

[Jeff Jarvis]

Yeah, that's it.

[Leo Laporte]

That's definitely it.

[Paris Martineau]

I mean, someone in the chat has told me that I should perhaps reconsider this whenever I get heating things, as it might get hot enough to melt.

[Leo Laporte]

But no, you know, there's an insight though. Take the hat off because look, it's like a little bottle cap on the top, like a soda bottle.

[Paris Martineau]

Yeah, this is how you like.

[Leo Laporte]

Who knew? Because you gotta, you gotta fill it up with.

[Paris Martineau]

Fill it up with stuff.

[Leo Laporte]

There's gooze and liquid. Don't drink the liquid, though.

[Paris Martineau]

I mean, I could if I'm looking for. If I'm. If I'm jonesing for a drink and there's nothing else, I could just crack it and take a big sip.

[Leo Laporte]

They say the hat will melt, so don't. Yeah, we're gonna have to take.

[Paris Martineau]

You're saying my three dollar cat sized cowboy hat might not be made of quality material.

[Leo Laporte]

That there in Laughing Away is Jeff Jarvis. He is the professor emeritus of journalistic innovation at the Craig Newmar Graduate School of Journalism at the City University of New York, now at Montclair State University in New Jersey, and also SUNY Stony Brook. And of course, the author of many books, including the Web We Weave, his latest, and the Gutenberg Parenthesis is now in paperback. We pivoted this show in case you're just tuning in. We thought this week in Google wasn't giving us enough good stuff to talk about. Let's talk about AI. And that's been a wonderful opportunity for us to interview some of the people who are working in AI in interesting ways. I want to introduce an old friend, Harper Reed. He's a technologist, entrepreneur, and hacker. Hello, Harper.

[Jeff Jarvis]

Hello.

[Harper Reed]

Hello. Happy to be back.

[Leo Laporte]

It's great to have you. I've known Harper for ages. I think we first met in Canada, actually, when I was doing a show up there.

[Harper Reed]

Yeah, I think so.

[Leo Laporte]

Yeah. You may know the name. He was the CTO of the Obama campaign. In 2012, he sold his e commerce startup to PayPal. Well, was it Braintree first and then Braintree sold to PayPal?

[Harper Reed]

We were after Braintree sold to PayPal. So we sold to PayPal and then lived inside of Braintree, which is a.

[Leo Laporte]

Good place to live.

[Jeff Jarvis]

Was the Mafia all gone by the time you got there?

[Harper Reed]

Yes, thankfully.

[Leo Laporte]

This was years. This wasn't so long ago. This was long gone. Yeah, yeah. Harper, what are you doing these days? It looks like a little code workshop you got going on up there.

[Harper Reed]

Well, we recently founded a new company, raised a little bit of money, and we're building what I like to kind of describe as super Clippy or Mega Clippy.

[Leo Laporte]

AI based.

[Harper Reed]

AI based. We. We are building, you know, agents that are personified and a little bit unhinged. And we'll do some work for you and help you out and live on your computer and maybe talk shit we don't know yet.

[Leo Laporte]

That's interesting that, that. See, a lot of people think that that is a kind of bad use of AI. Yeah, it's interesting, but I, you know, I trust your judgment. I mean, I don't think you would waste your time on something that was a bad use of AI.

[Harper Reed]

I don't know if we know what the right uses of AI are. I just read a really great blog post. I just found this out of kind of nowhere from Hacker News, which is called the AI Horseless Carriages, which is just about how bad the Gmail AI is. And I think there's a real opportunity for people to, instead of taking AI and shoehorning it into a service that already exists or a product that already exists, is to think, what should this product look like with these new automation? And the author has this perfect quotation. Let me see if I can quickly find it. Where he was like, it's a word. They were like, I don't know their gender. But it's a world where I don't have to spend time doing Monday work because everything. Agents do it for me. Like this idea of like, why have the agents, why have the computers? Like, I guess I totally messed up the quote. I just. No idea what it is. Anyway, read it. It's great.

[Leo Laporte]

I am. I have it right here. We'll put a link in the show notes. It's Pete Kumins. K O O M E N S Blog. AI Horseless carriages. Yeah, we're in that early stage, aren't we, where we don't know what it's going to be. You came to my attention because of course I follow you in many different places. But I saw your blog post about using AI to code. Now, you're not a vibe coder, are you?

[Harper Reed]

I mean, I've been a vibe coder since I was about five years old and I learned basic, man. Like there's. I definitely am the. I'm the type of guy where I'm like, man, that was a stressful day at work. Let me pull up some python to relax.

[Leo Laporte]

Yeah.

[Harper Reed]

Coding has been about vibes since I was a child. But I do think there is a. There is both a positive aspect to the kind of quote unquote, vibe coding that we talk about today and some negative aspects. But I do think that the concept really brings up some complicated points of view. One of them is just simply that maybe code quality doesn't matter as much as it used to. Which is a very bizarre point of view for someone who has, you know, made their career as having good code quality. You know, that's kind of the.

[Jeff Jarvis]

The import of AI entirely is. No, nothing needs to be perfect anymore.

[Harper Reed]

Well, it's. It's even more. It's. I. I stumbled across this kind of nuance that I was acting upon, which was I would never have AI write a blog post, but I have no issue with AI writing my code. I want to be very careful in my language that I'm putting out to the world, but I'm not careful in the code of the apps that I'm putting out in the world. What does that mean? I think it's because with the code, I'm able to set up testing infrastructure, verification tools, etc. To test to make sure that it works and to make sure that it's quality and I can scan it and do analysis. And there's all these tools that have been around for dozens of years, twenties of years that, that are. That will solve that problem. Whereas writing, all the AI writing seems to sound a little bit the same to me, much like all the images seem to look the same. And there's obviously people who are very good at prompting these things, so it sounds much better. But I've just found that when I read. If I'm like, I'm going to do a blog post about AI and I'm going to use AI to write it, I'm like, this doesn't sound like me. First of all, it has capitals and it's misspelled or it's spelled correct.

[Leo Laporte]

And it uses real grammar, punctuation and things.

[Harper Reed]

Punctuation, you know, I mean, it's like, what is a period?

[Leo Laporte]

EM dash. Oh, my God.

[Paris Martineau]

Hey, don't come for the EM dash.

[Harper Reed]

I'm an EM dash machine, man. I put EM dashes, like between every word.

[Leo Laporte]

We all agreed that that's a bad test of AI.

[Paris Martineau]

I use EM dashes so often. I often have to rewrite sentences because I'm like, there's a three EM dashes, like sentences right after each other. I gotta, I gotta throw a parenthetical.

[Leo Laporte]

I gotta put a semicolon in there somewhere. So you write. In fact, in your most recent blog post at Harper Blog, I find that I. My voice to be heard and not normalized, whereas my code needs to be normalized. In this context, we mean normalized.

[Harper Reed]

Well, I think, like, if you're building, if you're working with a giant team of engineers, the goal is to have a consistent quality output. Like, that's just. You don't want someone who is really good and someone is really bad. You want to like, normalize that much, or like a compression algorithm, so to speak. You want to.

[Leo Laporte]

Yeah, sometimes that 10% coder is not the ideal person on your team because he's so far ahead of everybody else. Nobody understands what the code's doing. And he can document it, but that doesn't mean that it, you know, it's just. It's a different level. Yeah, that makes sense.

[Harper Reed]

I think it's. It's like, it's such a. It's such an interesting problem because we all think about that. We want to find the 10x person or we want to find the person that's going to focus on this 100%. And what I've noticed is oftentimes they're just not great to work with. But there's a thing is I remember very specifically on the Obama campaign, there were a few things we had to do that required that person. And so I think those people end up being much more mercenary than they actually want to be because they're just such a. That they're not working with a team. With that said, you know, working with a team that's really running everyone will make everyone go faster, but only if you can achieve that consistency. Whereas writing for me is very much about my individual expression. Like, I would rather tweet stupid than have an AI tweet something that's very consistent. Like, I don't think the consistency in my personal expression is really what I'm looking for other than I want to wear black every day. But that's a simple. That's simple. And I don't have AI doing it.

[Leo Laporte]

Makes sense because when you're coding, what you're doing is you're ta. You're kind of translating human requests into something and the machine will understand. Well, what better to do that than a machine, right? I mean, I can understand that. Machines should be able to code. That's their language.

[Harper Reed]

I think that's true, but I actually think it's much more confusing because we're not. I think that's actually how it used to be, is we used to tell the machine in its language how to do it, which is like a python to, you know, like assembly language. Yeah, eventually, or whatever. Or like you take, you know, you write rust or go, and it compiles to a binary, whatever, all that stuff. What's happening now is we have this layer in between it that is the language model, which is like taking English. That's weird. Like, it's weird. Like we were doing this thing earlier on and one of, one of my, one of the people I work with was yelling at the computer, no, no, no, no, no, don't go to Twitter. Like, like we're yelling. We're like, what are you doing, computer? Like, I'm walking here. It's like this funny thing of like, like, why is it doing this? And you really are like, why would you do that? Like, we talk all the time about like, yeah, I tried to get it to do this and it wouldn't do it. It's like we're talking about a teenager. But, but it is like that because if you think about, you look at the good prompting guides, they have this funny thing. It's like Talking to a 12 year old about taking responsibility for homework or Solving a problem, you're like, well let's make a plan, think step by step, take a deep breath. Like all these things, we add them to the prompt, make the prompts better. And it really cracks me up that we're at that point in computing and I would never have predicted this in all my years as like a Java programmer and like a CTO and all that shit, I would never have predicted.

[Leo Laporte]

That I would be coercing, coercing them machine into cooperating.

[Harper Reed]

It's like a really angry unfed 12 year old who's just like, I don't want to do my homework. And you're like, okay bro, okay.

[Jeff Jarvis]

What would you, what would you advise 12 year old Harper, given the current reality?

[Harper Reed]

Probably move out of the U.S. well there's that.

[Leo Laporte]

Not the current current.

[Harper Reed]

Maybe learn how to shoot a gun. I don't know.

[Leo Laporte]

We're talking, no, we're talking about the coding reality.

[Jeff Jarvis]

Reality.

[Harper Reed]

I think that's actually a really complicated problem because what I'm seeing is I'm seeing a lot of folks who are in engineering, management. So people who are like me in my 40s, had a career programming and the people we're potentially hiring are cutting their teeth on this stuff. This is where they're learning. And those two things is, I wouldn't say it's so far as a cunium paradigm shift, but it's close to that kind of idea where we're using language that is not compatible and not collaborative with one another. So when I say yeah, just go ahead and code that up, first thing they might do is go to ChatGPT, go to Claude, start interacting with it, start doing some planning. Whereas if you told someone who had been programming for 30 years, 20 years, 10 years or whatever, that's not the first thing they do. I'm not mad about that, I think it's really hard. But for those of us who are around in the early 2000s, the late 90s, I think this is going to start to sound very comfortable and horrifying, which is. I think we're back to waterfall.

[Leo Laporte]

Waterfall is what Agile replaced. How would you describe it? You start at the beginning and you get to the end and it's a waterfall of code.

[Harper Reed]

Well, it's a funny thing, right, because I remember very clearly I got a job in 2001 as a Java programmer. I worked with this great team that was at Tata Infotech at the time. We had taken the project from IBM Global Services. These are all dinosaurs from the past. I had this team. I was the only Engineer in the U.S. i had a team of, like, four or five people around Chicago. We were all in our 20s. We all went to bars and all that stuff and had a lot of fun, and it was great. But when we weren't doing that, we were doing. We were coding this app. And so we had a guy named Vinay who had built the spec, and he was the spec guy. He knew everything about the product. He talked to all the business owners, he talked to all the stakeholders, meeting after meeting after meeting after meeting to get the spec. And then after, like, six months, he would be like, the spec's done. And it was like this big deal. Like, we did it. Everyone, let's have a party. The spec's done. And so you'd have the spec and it'd be like, thousands of pages, not really. Probably hundreds of pages. And we would all look through it and we'd make comments and, like, the spec's done. So then he'd pass it, like, over this kind of firewall to the engineers, where we had. No, we just had to do what was said in the spec. There was no, like, choice. It was like, oh, this is what you have to do. You have to achieve this thing. And then once you were done, you know, eight months later, you'd be like, we're done. We finished the product forever.

[Jeff Jarvis]

It's complete forever.

[Leo Laporte]

Yeah.

[Harper Reed]

Then we bring it back to the stakeholders, and they'd be like, this isn't what we wanted at all.

[Leo Laporte]

It's a game of telephone.

[Harper Reed]

It's a game of telephone. But that process, and this is the thing that I'm really reflecting on right now, is to do well with LLM CodeGen, and to really work well in this space, you have to think about what the spec is for such a long time and so carefully. And so strategically, you spend a lot of time working on requirements, trying to define what are you trying to build. Unlike in 2001, it takes 15 minutes, right? Like, it's not like, this is a long time. And then you pop it over to, like, you know, Claude or I use cloud code, for instance. Cloud code. You just dump it in there. You're like, yo, cloud code. Get this done for me. You know, and it's just like, does its little magic and says some weird puns, and then it's like, I'm done. And you're like. Then you check it and you're like, the other. Same result. That's not what I wanted at all. And then you just go back and do it. And so it's the same process in many ways. I wish I was still working with Vinay. It was really fun to work with Vinay. He would be great at this. But it was like this thing of, like, defining that spec, moving forward, building and building and building, and then going back and make sure it matches the spec and so on and so forth.

[Leo Laporte]

It's iterative, but the iterations are much faster now. Is that right?

[Harper Reed]

I think so. And I think if you're very successful, most people I've talked to that are doing well with this seem to be adding a lot more context, a lot more writing to their product, just to everything. So you'll have a doc in your product that might just be like, features for today. And you just go through and define all of these features for today in, like, very great detail. And then the LLM will go through. And because it has so much context, it can actually successfully do that. I think one of the missteps with people who are just starting out is they don't add enough context. And so, like, I remember when I started out, it was just like, code. You know, you're just shouting at the computer. You're like, I can shout at the computer. This is great. You're just like, code. Make the feature go. And then it's just like, okay. And it just does something. And you're like, this is terrible. That's because you didn't give it enough information. But once you just slam it full of information, it does a pretty good job.

[Jeff Jarvis]

Or once you see the Internet. So I was there to start at the start of Epicurious for Conde Nast, though, many years ago, and I watched this cultural conflict. The engineers, we hired an engineering company, and they said, okay, give us a spec for every single page you're ever going to have. And the editors said, no, just start making pages. We'll know when we see it.

[Harper Reed]

Yeah, yeah, yeah.

[Jeff Jarvis]

And now you can. And each was right and wrong in its own way. But now you can do that. You can know it when you see it.

[Harper Reed]

I do a lot of that with V0 from Vercel, which is a really fun product where I just go to it and I'll just, like, type. This will be like, if I'm. You know, you're on that call where you're kind of zoning out. You're just kind of paying attention. You have. You're just a. It's just a checkbox. Your. Your presence. You're not participating there. I'll just be like, fuck it. Let's. Let's make a video Game that's like our type from the 90s, Neo Geo. But I want it to be frogs fighting pizzas. And it's just some like, asinine prompt. And then I just sit there and I wait three minutes and I'm like, oh, look at that. There's that game I wanted. I always wanted the, you know, fight to fight a pizza or whatever. And like, I just do this over and over again.

[Leo Laporte]

It's kind of trivial, right? I mean, by the way, here's V0 what can I help you ship? And then you write in the prompt and they have some examples of what people have done. But Harper, one of the criticisms of this is, well, it can do that because it's writing like Flappy Bird's been written a thousand times. It's seen all that code. So it's writing something it already, you know, has in the LLM. It knows. So can it. But can it do original things?

[Harper Reed]

I'll quote a movie, Life of Brian. We are all individuals. I actually don't think there's that much unique stuff out there. Like, if you're thinking about what are we gonna solve, like, let's build social software. There's so much social software built, successful and otherwise, documented and otherwise, that I think it's gonna do a pretty good job at a lot of that stuff, right? That doesn't mean that. That a team of very experienced people with a lot of engineering experience aren't going to do better than a team of just neophytes who've never done anything. But there's still a lot of luck. You could build most of the things that my career has been defined by. I think you could build with one or two people in like a couple days using AI. Using AI. That doesn't mean it has the customers. It doesn't mean it has the growth. It doesn't mean it has the vibe or the, or the experience or the feel of it. You still have to be good. One of my friends, an investor who's backed us a couple times, James Cham from Bloomberg Beta, said early on, like, this is a time of curation. This is a time of having, like. Just because you can build an Instagram clone doesn't mean that you're going to build the next Instagram, right? Just because you can build something doesn't mean it's going to be cool, feel good, people are going to love it, et cetera. You still have to be good at that. Which means just because you can prompt it and say, I want to make a video game that is A combination of Teenage Mutant Ninja Turtles and, you know, some Chicago pizza company like. Like that doesn't mean that anyone is going to play it or care or what have you. But this is very similar to what it was like before. And I think it's okay.

[Leo Laporte]

Okay? Yeah. I mean, it's. But it's so. It's good for an mvp. A minimal viable product. But you still have. But I've always said this. With anything the AI does, human input is critical to making it. It useful.

[Harper Reed]

I also think that we could take some AI, we could go to Descript or whatever, pop it into Descript, and we could replace AI with human engineers, and the same conversation applies.

[Leo Laporte]

It would be the same thing.

[Harper Reed]

It's the same thing, which is like human.

[Leo Laporte]

So is the AI the equivalent of a junior engineer? Then what is the AI?

[Harper Reed]

I just treat it like a calibrator. I don't how to say that word. I was trying to say that word the other day and I listened to an equal.

[Leo Laporte]

A collaborator.

[Harper Reed]

Yeah, collaborator. That's how you say it.

[Leo Laporte]

Thank you.

[Harper Reed]

I've wondered how to say that for years. But that's.

[Leo Laporte]

By the way, what I do is tell people how to say things. That's my entire. That's my whole thing.

[Harper Reed]

That's the gig. I used to think it was like an intern, where you'd have this intern from some very fancy school and you'd bring them on and they knew everything. They were obnoxious, they knew everything, but they literally knew nothing. You know what I mean? We've all worked with those people where you can ask them any fact and they're going to tell you with authority what it is. But then if you ask them, like, you know, where do. Like, how much does a banana cost?

[Leo Laporte]

No idea.

[Harper Reed]

They're going to be like, I don't know. What do you mean?

[Jeff Jarvis]

What's a banana?

[Harper Reed]

You buy a banana. So, like, these type of things? Yeah, money, money.

[Leo Laporte]

They're savants, in a way. They're idiot savants, kind of these AIs.

[Harper Reed]

Well, yes, but one thing that I learned from Ethan Moloch, who I highly recommend reading, was oftentimes I'm paraphrasing here, I don't think he actually said this, but oftentimes when you find yourself stuck or the AI isn't doing this well, whatever it might be, that's a prompt problem. That's not a capabilities problem. I don't know if that's true. I don't actually care if it's true. But I practice that where if I'M talking to Cloud or ChatGPT, and it's not giving me the example or the results I want. Then I'll be like, okay, this is a me issue. I'll iterate much like I would on code or what have you to try and get to the right end. And usually two things will happen. One, I'll get there. Or I'll just get distracted and start generating images of my friends doing ass and I things.

[Leo Laporte]

Which is basically time wasted. Or maybe not. Maybe that's a good thing.

[Harper Reed]

I mean, Leo, I think that life is too short to be worried about that. I mean, it's fun and I laugh and like, you know, like, I'm gonna. I'm gonna walk outside and he'd get hit by a meteorite or something, and I'm gonna be very happy that I generated a pancake logo for my friend Clint.

[Leo Laporte]

One of the things Ethan Mollock's doing that's interesting, Jeff, you're probably familiar with it, is Wharton Interactive. He teaches at Wharton. And the idea of using games and AI as a kind of a useful tool in education, I think is very interesting. Yeah. All right, Well, I kind of want some best practices from you then, if we can, Harper, because you've written two really good pieces on how you use AI as a kind of partner coding. Yeah. Is that a fair way to talk about it? Yeah.

[Harper Reed]

Yeah. I think this is a really good and important topic because what I'm finding is a lot of folks are just trying to figure out how to effectively start and, well, you say step one.

[Leo Laporte]

Get out of bed with wonder and optimism.

[Harper Reed]

Yeah. Which I think is bullshit. I don't think anyone actually does that. Let's be real. It's hard out here. It's hard out here for us AI programmers. I even say that. I even say, like, who am I kidding? But I think the first thing is, this is the reason I put that in there is because the first thing is if you're starting out with the assumption that it's not going to work, it's not going to work.

[Leo Laporte]

That's a good point.

[Harper Reed]

Like, if you just start out and say, this is a bunch of shit, and we're all. It's only generates shit and it's all terrible, et cetera, et cetera, et cetera, which I hear all the time from friends, then it's not. You're never going to have any success. Like, that's just. You're starting out with. From a point you're not. Now, it's not about good intentions or Not. I don't really care. It's mostly just about, like, you have to assume that something good will happen for a good thing to happen, and otherwise you're just going to be. Because you have to prompt it, like it's going to work.

[Leo Laporte]

And like, so the AI knows whether you believe in it or not. You gotta believe.

[Harper Reed]

That's the vibe, Leo. That's the vibe with the vibe. Yeah. Yeah. It also really starts to get religious at some point. And then we've all moved to San Francisco and talked about AGI. But I think the key that I have found is that I think there is a continuum and on one side it starts with like code gen or Sorry, like autocomplete and Intellisense has been around for forever. You have all these really cool coders.

[Leo Laporte]

Are used to that notion.

[Harper Reed]

Yeah. And that's a normal thing. And so you start there and then you're like. Then you just get asking more and more and then you start pasting code in and out of ChatGPT and Claude. And I think it's actually important to follow these steps because I think what happens if you go to the end, you don't. You don't have the nuance and the practice to prompt it appropriately. And so you end up making mistakes that are just incredibly frustrating. And then your response will be, this doesn't work. It's all bull.

[Leo Laporte]

So the first step is really dialoguing with the AI to understand better what it can do, what it needs to do, what you want it to do. There's a certain amount of training that's involved of you as much as the AI. Yeah, yeah.

[Harper Reed]

And one of the things I think is really hard about this is it's just weird. Excuse me?

[Leo Laporte]

It is. It is weird. It's alien weird and it's alien.

[Harper Reed]

Any sense. And you're also like, I am a very good programmer of Python. Why am I literally yelling at a computer? You know, like, why am I yet, like, why am I talking to a computer? And why is this happening?

[Leo Laporte]

You have to get over that, don't you? You really do. I've noticed that too. You have to over that embarrassment of, yeah, I'm talking to a computer.

[Harper Reed]

Yeah. I don't. I don't know if you guys have taken Waymos, but, like, I have this problem with Waymos, which is very similar, which I'm done with the ride and I get out and I'm like, thanks.

[Leo Laporte]

Do I tip?

[Harper Reed]

What do I like? It's this thing of like, we. Our whole lives have been spent to like, find this way to, like, coerce the network around us of people. And now we're trying to find the delta between how to do that with a robot and not. And it's. I think it's very, very difficult. But where the vibes coding kind of comes in is as you start upgrading down this path and you start doing more automation of your code base, there's a certain point where you're just letting go and just seeing what happens. I think that is very uncomfortable for a very specific type of person who likes control.

[Leo Laporte]

But you're not going to ship this code.

[Harper Reed]

Why not?

[Leo Laporte]

Oh, you are?

[Harper Reed]

Yeah. Every day I ship this code. Well, I was talking to a friend of mine, he was a very talented programmer and he was like, yeah, but it's not production quality. And I'm like, like, what is production quality? My entire life I've programmed computers. A lot of code has gotten to production. There's a certain point where I kind of graduated and I wasn't allowed to put code into production. But there is a thing of like, what does it look like? Every time I've entered a company and looked at the code, the first reaction is you're just like, oh, God, this is horrible. Like, no one has looked at production code and been like, this is perfect. This is wonderful. There's always this visceral reaction of there's a lot of bubble gum and a lot of bailing wire. And so, like, I don't think that human production code is much different than AI. AI production code. Except we can promise that AI production code is going to be normalized so much. Now, the best practice may not be your best practice, but it will be a best practice.

[Jeff Jarvis]

So let me ask you a question there, because I'm not, I'm not a programmer, but I've worked with many. And whenever you try to say, you get somebody new in and you say, oh, this was already done, but can you revise it? They want to just throw it all out and redo it completely. Yeah, right. Law of nature. How is AI at revising code that somebody else did?

[Harper Reed]

It's actually really funny. There's two answers that, one, we don't know. And that's the, like, just adding and editing to like, big code bases. We just don't know yet. Like, we don't know. Most of the people who are doing code gen stuff right now are doing greenfield work.

[Jeff Jarvis]

Right.

[Harper Reed]

On greenfield things. And they're, they're having a lot of success on that. But the question I get most from my friends who Are in engineering managing positions is I have a code base with hundreds of thousands of lines of code, if not millions of lines of code. I need to add this feature or fix this feature. How is it going to do with that? And some people are doing it. There just isn't a lot of people testing it. I'm on my way to try and generate a lot of code, so I'll let you know in a couple of years or whatever. But we're not there yet. But the second one is. In some cases it's easier to just have the AI redo the whole thing.

[Leo Laporte]

It's so fast.

[Jeff Jarvis]

Just like coders have always said too.

[Leo Laporte]

But it's faster than coders.

[Harper Reed]

Here's an example that blew my mind. I was talking to this good friend of mine who has this wonderful company and he was showing me this code base he had. And he had 250 pull requests. And I was like, this is a mess. Why do you have so many pull requests? And he was just like, oh, it's cool. So he showed me each pull request and AI had done. If the tests passed, he would examine it and just take a look. High level look. And probably merge it. If the test failed, he closed the pull request without even taking a look at it.

[Leo Laporte]

It's cheap.

[Harper Reed]

Was cheap. It's cheap. Cheap time.

[Leo Laporte]

Throw it away.

[Harper Reed]

And from a cost, he's like, why? We'll just redo it.

[Leo Laporte]

Why wouldn't we just throw it away?

[Harper Reed]

The whole thing? And so, Jeff, to answer your question, there's a chance that this is why I said, like, code quality may not matter. There's a chance that the fact that we worshiped this code, we were like, it's done. Remember back in CD time?

[Leo Laporte]

Goldmaster.

[Harper Reed]

We did it. Let's ship it.

[Jeff Jarvis]

Shriek Wrap.

[Harper Reed]

We don't need that anymore. If we build an iOS app and it sucks, we just build another one tomorrow. If that sucked, we build another one tomorrow.

[Leo Laporte]

It really is a new way of. It's a totally new way of thinking about this stuff.

[Harper Reed]

It's very strange.

[Leo Laporte]

Okay, so we've gone through. I've used IntelliSense. Now I'm using Copilot. It's basically kind of of auto completion on steroids.

[Harper Reed]

Super, super Intellisense.

[Leo Laporte]

And then. And then I'm pasting code into Claude or ChatGPT.

[Harper Reed]

And yeah, you're still using your brain here. Like so here you're still kind of thinking through your brain trying to figure out what's going on. You're doing the planning yourself and what you're finding is you're like, this is really awesome, but I wish I could do more. And you look over at your friend, your friends like playing video games while his computer's coding for them, and they're just like, wow, how did you do that? And so then you start to think about Ator or Klein or OpenAI. Just release Codex or whatever these things are, which are these terminal applications that you can run, hook them to an LLM and they just kind of do stuff in the background. And it's fun to have them do stuff in the background because you really can just do other things.

[Leo Laporte]

What are they doing? They're writing. Are you giving them tasks? And they're doing that in the background?

[Harper Reed]

Yeah. So I'll just tell you my workflow just because at this point, if you're in the agentic workflow part, this is like where I've been playing along with my team has been doing the same workflow. So kind of what you do is you go to like a ChatGPT or some place where you. Your your favorite LLM. I use ChatGPT for this because I really like their O1 Pro model, which is their expensive model.

[Leo Laporte]

You point out in your blog post though, that it's kind of a person, that's kind of a personal thing. You might like one better than the.

[Harper Reed]

Other and you might like llama the run locally. Like whatever. Like whatever. You know, it doesn't really matter because this process, what you're doing is building context. So I'll go there. I did that earlier today. I got into the office around 9, 8:45. 9:00. I spent 8:45 to about 9 talking to the LLM, being like, Hey, I want to build this thing. Here's my idea. And this is my favorite prompt. And I do this for all my prompts. Now I say ask me one question at a time. I prefer yes or no questions.

[Leo Laporte]

Okay.

[Harper Reed]

And so. Oh, look at that. And so what it will do. I love that. This is my favorite. Everyone hates this thing, but I think it's the best in the world. I think it's so good I will never turn it off.

[Leo Laporte]

That's AI.

[Harper Reed]

But what I'll do is I'll just. The yes or no question thing is really great because it allows you to dive deep into these things without an overwhelming experience with the LLM. So it'll just be like, hey, do you want people to log in with Google? Yes. They'll be like, oh, do you want people to log in with an email?

[Leo Laporte]

Yes.

[Harper Reed]

So it goes through. And as you're doing this, it's just building up more context about what you want to build. Then once you feel like you have enough context injected into the machine, then you say, write a spec for me and that's when I'll switch it to 01Pro and it will generate this very well thought out spec that then I can tweak and edit and make whatever I want into it. Then I'll say to that specific, I'll say, hey, can you go ahead and turn this into prompts? Once again, I'm still within the 01 Pro context. We turn this into prompts that I can pass to a code gen. Then I'll save the spec and then this new prompt plan, so to speak, into a directory. And then I will go to Claude code in a terminal and I will say, claude, look at the spec. Here's the plan, just do it. And I'll go and leave and come back and it'll be done. There's a little bit of poking and prodding and it's asking questions, but that's largely what I do. And with this, you can ship, I think you can ship kind of anything. I think the limit is your patience for dealing with a bunch of nonsense that this thing is going to pop up. But having worked with software teams for most of my life, it's the same nonsense. People are more fun. This is boring. But that's why you just surround yourself with people doing the same work. So you still get the fun, but then you don't have to talk about code.

[Leo Laporte]

All right, I'm going to be the old timer, the gruff old timer who thinks this is all all hogwash and ask you a few questions. How do I debug something that some machine wrote that I don't even maybe understand how it works?

[Harper Reed]

Someone told me a way to do this that I really, really like, which is you do what they call checkpointing. And so this is like in a video game where you have a save point, right? You're going through the video game, you pass that little flag and now it's saved. And if you mess up your game, you just go back to that point. So that's kind of what you do. So you just don't depart bug. You just redo it until it works. So it's this funny thing kind of back to the code is really cheap. Now this is tests to do this. Oh, man, so many tests. So many tests. I've had more tests in the last year than I've had my entire life.

[Leo Laporte]

Does the AI write the tests or do you?

[Harper Reed]

Yeah, I write the tests, but I'll do a lot of passes where I'll take the entire code base, jam it into Claude or ChatGPT or whatever and I'll say, where are my missing tests?

[Leo Laporte]

Oh, I love that.

[Harper Reed]

And then I'll have it build those tests and I'll.

[Leo Laporte]

Is it pretty reliable when you do that?

[Harper Reed]

Yeah, of course it's Missing tests are horrible because there's always going to be.

[Leo Laporte]

One that you think humans are very bad at tests.

[Harper Reed]

Terrible. Humans are very bad at tests. The machines are very good at it. I have a couple projects that we've been working on that are rust that have 100% test coverage. Every line of code is tested. Now some of this is really stupid because you're like, why do I need to test this line of code?

[Leo Laporte]

But it's just cheap.

[Harper Reed]

It's cheap. I mean it's not. The time of running the test isn't cheap. That's forever, right? But it is. But it is cheap to get there. And then once you get there you're like, oh yeah. So then you save it and you basically save it and then you have it go do stuff and if it messes up you're like, ah, try again. Clear the context, reset back to the test to checkpoint and do it again. The once. Once the agent started running the tests. That's really when everything checked.

[Leo Laporte]

Wow, now you're using eight or you like to do this on the command line as opposed to in VS code or something. I mean.

[Harper Reed]

I've always been a terminal person. Yeah, I mean I started running Linux in 93, you know, back when it was you still had telephones. And I think that there's a thing about that, that I will always be most comfortable in that space. But I have many friends who've used Windsurf and Cursor and all these other things that really like that Vibe. I just, that's just not my happy space. And I like when someone was like, this is Vibe coding. I was just like, I'm in, man. Let's burn some incense, let's dress comfortably, use what tools are comfortable for you. I am not worried about what I used to be worried about. It used to be like, let's make sure the whole team is using a consistent tool set. Let's make sure everyone is programming in the same language, so on and so forth. But if we hired someone to do a project and they're like, we're going to do it in PHP or Haskell or whatever and we thought we could hook host it and we could do it. Well then I'm not worried about that because honestly they're just being Vinay, making sure the spec works, making sure that everything is tight, making sure the requirements document is up to speed and then the computer's doing the code. So I think this is a very, it's very bizarre. I'm very uncomfortable with this to be very clear and I'm not sure, quite sure where it falls out. But I would say right now 80% of our code is being generated by 8AI.

[Leo Laporte]

Who's us.

[Harper Reed]

Who? I have a small company called 2389, which is 2389 AI. And yeah, we're building some very strange things and it's all AI generated.

[Leo Laporte]

I love it. So there's a famous comment in some old code that says, you are not expected to understand this.

[Harper Reed]

Yes, yes, that's how I feel every day.

[Leo Laporte]

Yeah, don't change it. I understand it, I wrote it. But you're not expecting to understand it. Don't change it. It's a very famous comment. But that bothers people when at least it used to. I don't understand this code. Do you always understand what the AI code is doing?

[Harper Reed]

I never understand. I barely understand it right now. Like I like here's, here's what happens and it's very funny. So I said this before and someone got kind of mad at me. I don't think that humans context window is as big as some of these AIs.

[Leo Laporte]

Yeah, I think you're right.

[Harper Reed]

I just don't care.

[Leo Laporte]

I think that's exactly right.

[Harper Reed]

I just don't have as much space in my head as I thought I had. Like, I used to think I could hold the entire idea of it. I can think of the spec and all that stuff. But when you start looking at all this code, there's a point where I'm like. And I can feel it, I'm like, I've lost control. Like I was going, I totally under control and then I lost control. And I'll just be like, well, there it goes. I no longer am able to understand it. But there's another thing that I always remember which is I can just ask, I can just be like, hey, Claude, what's going on? Why, what's happening? Why is this, what does this file do? What does this part do? Because unlike that comment of the past, the person that wrote that probably was not very fun to work with.

[Leo Laporte]

No, it's kind of a commercial snotty.

[Harper Reed]

Didn'T trust you wouldn't want to work with you probably thought you weren't good enough. AI doesn't care. It'll be like, oh, that's a great question, Harper. I was waiting for you to ask that. Here's what that, what that does.

[Leo Laporte]

I have found AI to be very good at explaining code, actually.

[Harper Reed]

You're right, I'm not worried about that. What I am worried about though is getting to the point where you are supporting a product that you truly have no idea how it works.

[Leo Laporte]

That's a little scary.

[Harper Reed]

I think what's going to happen is much like when Agile came out, much like when any kind of when mobile came out, you start to get new tooling. And so what we haven't seen yet is what is the tooling look like. I mean, we're starting to see some of the dev tools, et cetera, but I mean like what is the tooling that is going to support the engineers, the engineering teams, the founders, etc. I think there's an important example. A good example would be formal verification. So if you're using like a Rust or some of these compiled languages, you can formally verify that it does what it does. And I think that's something that people are going to start relying on so that they can prove that the code that the AI wrote, even though it hasn't had human eyes on it, kind of does what it says it does. Now I know that as saying this, this is a little sacrilegious, religious to sacrilegious. I don't know how to say that word either. But this is, I guess this is bad. Like most of my engineering manager friends are going to die. Like they're all going to say, harper, what are you doing? You're no longer allowed to come to the party. And I don't know what to say. Like, I think this is the future and I think it's unfortunate that this is where we're at because it's just, I don't think we're going back and I think that's scary.

[Leo Laporte]

I thought maybe I had decoded the name of 2389 Research.

[Harper Reed]

You'd never guess what it is.

[Leo Laporte]

Well, I thought I had because thought.

[Paris Martineau]

Is that you said 30% of your company is now AI. That's 716 of 20.

[Harper Reed]

Well, it's actually really boring. It's just our birthday months. When you start a company, you have no one.

[Paris Martineau]

You think it was late.

[Leo Laporte]

Well, the line number of the code comment you are not expected to understand. This is something like 2389.

[Harper Reed]

If that's it. I'll just change it. Well, we were telling people is our favorite parts of PI.

[Leo Laporte]

Oh, that's good too. And then now you just have to figure out where that is.

[Harper Reed]

Yeah, it's a really good part. It's right around.

[Leo Laporte]

It's the middle of PI. Halfway to the end.

[Harper Reed]

Halfway to the end.

[Leo Laporte]

Yeah, exactly. The blog posts at Harper blog. There's a new one, the original one. This is the new one. LLM codegen's Hero's Journey. The one I read that actually prompted me to call you was my LLM CodeGen workflow at the moment. And it's really good and it really is a great place to start if you're thinking about playing around in this. And I think the fact that you've done this so much now, Hark Harper, is actually encouraging because you're a smart guy and you're not. I don't think you would be advocating this if there were serious issues with it.

[Harper Reed]

Oh, I think I would. I think I would.

[Leo Laporte]

You would? Okay, nevermind.

[Paris Martineau]

He wants to clarify that he's.

[Leo Laporte]

I'm sorry I trusted you. Forget it.

[Harper Reed]

No, no, I think. I think the problem here is that we really don't know how this is going to end.

[Leo Laporte]

That's true.

[Harper Reed]

And what I mean by that is, I mean this very specific thing that I'm sure, like y'all. I have a lot of friends who've been laid off from tech jobs over the last two years a lot recently, and I don't think they're getting jobs again. Like, I don't think. I think this is very similar to 2003 to 2000 to 2003 with print designers and probably very similar to accountants, but I wouldn't know about that. That's more back in the early 1800s for Jeff Jarvis. But the thing is, I think there's. Okay, 1500s. Yes, we need to be accurate. But I think there is a reality here that as much as we are benefiting from this, there is a downside, which is I'm personally not hiring as many people. I have a team of people I've worked with for the last 20 years that I really like working with, and we just don't need as many people anymore. And there's some really scary answers.

[Jeff Jarvis]

So I've got a question for you. Go ahead. You go up.

[Paris Martineau]

I assume it's going to be the same question. What are you going to do when you're the person whose job is eliminated? If this.

[Harper Reed]

I'm going to be. I'm going to be so Happy? I'm going to be serious. I've been trying to force myself out of a job for years. That's my dream. I just want to go and lay in a park and read books. Like, that's, that's really what I want to do.

[Leo Laporte]

So do you believe in UBI? Is that.

[Harper Reed]

Oh yeah, 100%. 100%, yeah. And all those people who are like, what will humans do? It's like, bro, what do you mean? Have you been.

[Leo Laporte]

Do what we want to do?

[Harper Reed]

Like, it's great. Like it's wonderful.

[Leo Laporte]

Nobody likes their job.

[Jeff Jarvis]

Yeah.

[Harper Reed]

Someone who recently said, I spent a lot of my time without work. Without working. I'm very lucky in that regard. And someone said, when are you planning to reenter the workforce? And I was like, what are you talking about? I would never plan that. But I do think that there are some really scary aspects to this. And I do think we are underestimating the impacts of how this will work. And I think we're seeing a lot of founders like myself, big company CEOs, etc. Who I was just talking to someone who is like, I want to know how to use this in my company and I want you to work with my tech team to like help the tech team use this more. And I was like, yeah, but we're kind of having them dig their own grave. Like it's kind of messed up. And so like we, we are not prepared for that. And so I've been thinking a lot about like, what does that mean and how does that work at the same time? Yeah. You know, still gotta, still gotta eat. So I'm over here trying to, you know, hustle in the token mines, trying to build some cool company.

[Leo Laporte]

This is an age old problem though. As productivity improves, people.

[Harper Reed]

I mean, I noticed, I noticed something when I. I've been thinking about this a lot because of the waterfall stuff. When I got it from my first job in 2001, you know, it was like a huge room full of these little. What are the cubicles? You know, just as far as I can see with like tech pieces. People.

[Leo Laporte]

Yeah. Engineers. Design skinny black ties and pockets.

[Harper Reed]

Right? Exactly. I had to dress up every day. Exactly. And then that shrunk. By 2005, I was on a team of like 20 people. And then by 2007 it was 10. So on and so forth. So it's not like this. This started with AI. No.

[Leo Laporte]

This is what computers do in general. Right.

[Jeff Jarvis]

I think democratize it. Yeah. So I'm gonna ask. I have a different question.

[Leo Laporte]

Not that It's a wonderful thing for the people who lose their jobs, but it's kind of, you know. Well, it's like, I don't know how you avoid it.

[Jeff Jarvis]

I teach or try to, and helping start a new program in technology, AI and society, and wondering what young people should do. Advise either a university or its students who are interested in technology, who love it and want to do things with it. What should they be doing?

[Harper Reed]

I think just experimenting at this moment, there's something that I noticed, which is my brother, actually, Dylan. He's Dylan. I think he's Dylan. Not Dylan. I don't remember. So he's a microblog as well. But he texted me one day and was just like, hey, I wrote a video game. And I was like, what do you mean you wrote a video game? You're not a programmer last night? Exactly. He's just like, no, I just wanted. I just wrote a video game. And I was like, oh, that's new and unique information. And. And that is, I think, where we're at. Where someone told me yesterday, this is the age of the idea guy. All of those people who are like, I have an idea and I need a tech person. They're gonna live, they're gonna eat now. But I think the people out there. There's some cool parts about this as well, which is the people out there who have the patience to dig in and experience it. You actually see one right there. That's a ment. My friend Mint who comes in and he hacks on this product that he did not have six months ago. He's not a programmer. He's learning how to program. He's using Replit agent, he's using V0. He's using all these things to build this stuff. He has a. You know, he's a. He's. He's doing it. And the point is, is that he's spending the time to do it. So if I was. If I was going to talk to students today, I think I would just suggest just try it out, try it on, see how it feels. If it doesn't work, try it again. Read the Ethan Moloch book. I think his book is really good at kind of summarizing how to think this. If you start out saying it's not going to work, it's of course not going to work. But if you try it, like, build something weird. Build a mobile app. Build an iOS app. You've never heard that iOS is a thing. Great, try it. You can achieve almost anything with enough patience. And that's A limiting factor.

[Leo Laporte]

It's a very exciting and also very terrifying thing about this time that we are in a time of experimentation. Everything is changing so rapidly and we don't know where it's going. You can't see, say, with confidence, don't become a programmer. You can't say anything. You. But I think you're right, Harper. Become an experimenter. By the way, I got out my copy of Lion's commentary on Unix, 6th Edition, and in fact the code comment you are not expected to understand this is on line 2238. So, oh, so close.

[Harper Reed]

So close. Well, I mean, we can name it.

[Paris Martineau]

I think that might be the necessary next step.

[Leo Laporte]

It's a really. It's a really good Easter egg.

[Harper Reed]

You don't have to do any of that.

[Leo Laporte]

It says if the new process paused because it was swapped out, set the stack level to the last call to save you. Ussav. That means the return, which is executed immediately after the call to Ru actually returns from the last routine which did the save. You. You are not expected to understand this. And that's unix, baby. Right? In a nutshell, right there.

[Paris Martineau]

Wait, what's your company's number again?

[Harper Reed]

2389.

[Leo Laporte]

So close.

[Paris Martineau]

What is line 2389, Leo?

[Leo Laporte]

Oh, that's a good question. Line 2389 is blank.

[Harper Reed]

That's actually, that's why we chose it, because it's a white field.

[Leo Laporte]

It's a white field.

[Harper Reed]

Yeah, yeah, we're ready to go. We're filling it in.

[Leo Laporte]

Harper, you are. You are an idea guy. You never fail to inspire. You never fail to excite me with your enthusiasm for life. And I think you've made a damn fine case for getting out there, experimenting. Go to Harper's blog and read these two posts to get started. Harper, blog. I really think there's a lot of meat in here. And there's also some great pictures. And you can use it to start your own journey, your own experimentation. Makes me want to. You know, Harper, I'm so old. I wish I could start over again. I would love to be young right now and playing with this stuff. I think it's very, very exciting.

[Harper Reed]

I like that. That's the thing you would do if you were young again. You know what I was young again is go inside and use a computer. Not enjoy the world. No, that would be.

[Leo Laporte]

I could do that when I'm old.

[Harper Reed]

But I. I do think that it is worthwhile no matter what. Like, I have so many friends that are like, oh, I can never become a programmer. I think the Exciting thing is you can just go to like Bolt or Lovable or any of these apps and just build the thing you've always wanted to build and see how it goes. And that's a big change to what we were doing before. Like we didn't have that opportunity. You could not reasonably tell someone, just learn how to do a mobile, how to build a mobile app. That's an asinine thing to tell someone. Now you can just be like, no, just build the mobile app. Like it's easy to do. So and it really is easy. Now will it scale to a billion users? Well, that seems like a problem that you'll figure out if you have that problem. Like you'll figure out a solution.

[Leo Laporte]

It's an exciting and interesting time. Not necessarily, you know, interesting times are not necessarily the best times. Yes, but, but it is interesting. You can't deny that it's uncomfortable.

[Harper Reed]

But good, I think.

[Leo Laporte]

Yeah. Harper, great to see you again. Thank you so much. Thank you so much on and exciting us with your ideas. I appreciate it. Harper Reed, ladies and gentlemen. All right, we're gonna take a break. We will come back and we'll talk about Cluey or whatever that is. I don't even know what that is.

[Jeff Jarvis]

We will find out.

[Leo Laporte]

We'll find out. I'll get a. I'll get clued in.

[Paris Martineau]

You'll get it Cluey.

[Leo Laporte]

You're watching Intelligent Machines. This episode of Intelligent Machines brought to you by Big id. The next generation AI powered data security compliance and privacy solution. That's amazing mouthful but if you think about it, AI is transforming businesses. You know that. You watch this show but with data risk bias and compliance challenges, are you adopting AI responsibly? I'm sure in the middle of the night you think that and go, hmm. Well, Big ID can help. They deliver end to end AI and data governance to help enterprises manage that risk, enforce policies and ensure responsible AI adoption option. Ensure AI only accesses safe to use relevant data and automatically tags sensitive information by policy and type. You're in control. Suddenly Big ID is the only leading solution to uncover dark data through AI classification, to identify AI risk and to manage the data life cycle. Plus you can use it to scale your AI strategy. Amazingly, it integrates with everything you're already using. Your existing tech stack stack unmatched data source coverage. It allows you to automate privacy and security workflows. You can take action on data risks with automated remediation orchestrations. You can automate privacy management, regulatory compliance, data rights requests and more. Just does it for you, partners include, well, everything in your tech sack. ServiceNow, Palo Alto Networks, Microsoft, Google, AWS and on and on. With big IDs advanced AI models, you gain visibility and control over all your data. It's the platform intuit named number one for data classification and accuracy, speed and scalability. Big ID equipped the US army, you've heard of them, to illuminate dark data and to automate data retention. I mean you're talking about compliance. The Army's got a lot of compliance rules, right? U.S. army Training and Doctrine Command gave him the best quote Here we go Quote this is from US Army Training and Doctrine Command. The first wow moment with Big ID came with being able to have that single interface that inventories a wide variety of data holdings, including structured and unstructured Data across emails, zip files, SharePoint databases and more. To see that mass and to be able to correlate across those. Completely novel. Again, this is a direct quote. I've never seen a capability that brings this together like Bigid does. Cnbc recognized Big ID as one of the top 25 startups for the enterprise. They were named to the Inc 5000, the Deloitte 5500, not just once, but for four years in a row. The publisher of Cyber Defense magazine says quote Big ID embodies the three major features we judges look for to become winners. Understanding tomorrow's threats today, providing a cost effective solution and innovating in unexpected ways that can help mitigate cyber risk and get one step ahead of the next breach, end quote. Start protecting your sensitive data wherever that data lives. Big ID document Get a free demo to see how Big ID can help your organization reduce data risk and accelerate safely the adoption of generative AI. Again, that's BigID B I G I D bigid.com Im by the way, there's a free guide that can help you understand the risks of generative AI and data driven strategies to ensure responsible and compliant AI adoption. Something everybody needs to read that and Bigid can help of course bigid.com im it's the next era of data security compliance and AI. Bigid.com im we thank him so much for supporting us at this show Intelligent Machines. We invite you to support us too by going there and using that address so they know you saw it here. Bigid.com Im all right I want to know about Cluey now you got me all excited Cluly Chloe is somebody is like a children's show character.

[Paris Martineau]

Close that bluey.

[Leo Laporte]

Oh yeah.

[Paris Martineau]

So this all started. You may have recall I think In February or March at least. I noticed on Twitter there was this kid who kept going viral because he was talking about how he's a sophomore at Columbia and he used some app he created in order to get job offers from Amazon, TikTok, a bunch of other tech companies by cheating. Basically it was a app that would kind of run undetected in the background. Whenever they would ask him like leetcode style questions, it would answer it for him. And he claims that he got job offers through this and then documented it in this thread that went viral, resulting in Columbia basically reaching out to discipline him because I guess he had had, you know, done something kind of strange. Then he recorded the discipline hearings using this tool. Oh my God, that online. Then Columbia found him guilty of blasting their employees all over social media. So they expelled him instead of just suspending him. And today, or at least this week, he announced that he's. His new product, called Clulee, which isn't just for cheating on interviews, it's for cheating on everything. And there's a video to describe this if you'd like to play it.

[Leo Laporte]

So. And by the way, this is all part of a. A thread, a Blue sky thread about how awful this is. Everything about the tech and AI world makes me sick to my stomach. It's deeply antisocial.

[Jeff Jarvis]

The big question is, what's cheating?

[Leo Laporte]

Right. Thank you. Thank you.

[Paris Martineau]

It's a guy on a date.

[Leo Laporte]

His wife. So how old did you say you were?

[Harper Reed]

I'm 30. 30 years old?

[Leo Laporte]

Is it in his profile?

[Harper Reed]

Your profile says you worked at bananazon. Yeah, yeah, I've been there for the past eight years doing like senior software engineering. So, you know, like coding.

[Leo Laporte]

Really?

[Harper Reed]

Look, I. I know I look a little young for my age, but I promise you, I'm. I'm 30 years old.

[Paris Martineau]

Okay, well, maybe you are just way.

[Harper Reed]

Older than you look. I saw on your profile that you're. You're like.

[Leo Laporte]

So how's he seeing this? Or this is just a Mac.

[Paris Martineau]

This doesn't exist.

[Jeff Jarvis]

Essentially something to help you cheat. If you had Serotona Bergerac on an.

[Leo Laporte]

App, and then Onion person, Juniper Beer. Who's. Who's the one who posted this? Finds this deeply disturbing, deeply awful anti human garbage. We want you to cheat on everything. This is from Cluley. Yep. You heard that, right? Sales calls, meetings, negotiations. If there's a faster way to win, we'll take it. We built Cluly so you never have to think alone again. It sees your screen. Here's your audio feeds. You answers in Real time while others guess you're already right. Have you ever heard of this thing? Cluly B Because it kind of sounds like you. I'm thinking maybe I already have this. And yes, it does sound similar. Cluey is an AI designed to assist with various tasks by interacting with users in real time, much like I do. However, the context of its functions, especially concerning sales calls and negotiations, can be pretty controversial. Are you curious about how Cluey works or its implications?

[Paris Martineau]

The voice just really makes me laugh.

[Leo Laporte]

That's why I do it.

[Paris Martineau]

Serious.

[Leo Laporte]

If it sounded like this, it would just not be the same. And yes, they say the world will call it cheating, but so was the calculator. So was spell check. So was Google. Every time technology makes us smarter, the world panics and then it adapts and it forgets and suddenly it's normal. AI is not just another tool that will redefine how our world works. Why memorize facts, write code, research anything when a model can do it in seconds?

[Jeff Jarvis]

So define cheating.

[Leo Laporte]

Well, lying is not good. So I think that that video is unfortunate because it was a lot of lying, right? Yeah.

[Paris Martineau]

The video ends by him trying to order a beverage and the waiter or wine, and the waiter realizing he has a fake ID and giving him grape juice instead. And the girl storms out and then the AI helps him write an apology in real time to try and win her back.

[Leo Laporte]

So that's a very poor example, but I do agree that the calculator would be so. Hey, do you have my new moral panic video? Because I think this is cheating.

[Jeff Jarvis]

Hey, that's my franchise.

[Paris Martineau]

That's going to take a minute.

[Leo Laporte]

For those who are only listening, there is a Leo looks a lot like me, but as you peel the skin back, underneath is a robot head who laser eyes. Laser eyes.

[Paris Martineau]

Skin does not peel back in the way that you'd think, well, it's an AI.

[Leo Laporte]

What do you want?

[Paris Martineau]

I mean, my problem with this is.

[Leo Laporte]

Just, well, you wouldn't want to date this guy.

[Paris Martineau]

I wouldn't want to beat this guy. But two like it perhaps? Yes. One, I don't think we even can say whether or not his claim is accurate that he used this. Got him four jobs at these things, but let's pretend that it is. That says more about the ease the thing. The reason why companies like this ask you to solve like questions that are on like Leetcode and whatnot in order to pass some bar of an interview is to understand how your mind works and test your skills. If you're able to bypass all of that with AI, I mean, I think doing everyone a disservice.

[Jeff Jarvis]

Two questions. Sec. Okay. If you, if you use it to get the job, the date or the degree, maybe, but once you have it, once you're doing the job, if you use this to do a better sales.

[Leo Laporte]

Call, what's wrong with that?

[Jeff Jarvis]

What's the harm?

[Leo Laporte]

You're making it up anyway. As we all know, in sales, it's all. It's all made up.

[Jeff Jarvis]

There's a testing phase.

[Leo Laporte]

People should lie. I do have. I do have a kind of a values driven opinion about that. People should. Yeah. You shouldn't say you're 30 when you're 19, obviously. But if. I mean, but I don't think it's cheating to use AI to supplement your knowledge.

[Paris Martineau]

Yes, And I agree. If you're truthful about it. Like, I think it would be a totally different conversation if the pitch for this was tell everybody. Yeah, I'm going to answer these interview questions, but I'll be using an AI assistant. And I'll also be using that while working. Sure, go for that. But people probably aren't going to want to hire you because they want to hire someone who has these skills innately and could be even better with AI.

[Leo Laporte]

Right. I'm trying to.

[Harper Reed]

I'm trying to run.

[Leo Laporte]

Clearly right now the technology isn't bad, but like, the.

[Harper Reed]

The way they're presenting it really is really telling about who they are.

[Jeff Jarvis]

Yeah.

[Leo Laporte]

I think the app lies because. No, you know what? I'll tell you the truth. That's just good marketing. This is. This kid is brilliant. He's a marketer. And what better way to get us and everybody in the world talking about him to saying, hey, you should use AI to help you cheat. That's all. It's just good marketing. And I frankly, I tried to run his app and it doesn't seem to run. So maybe the whole thing is bs, that's all.

[Paris Martineau]

Yeah, I mean, that was.

[Leo Laporte]

He's just a con man. He's the Anna Delvey of 2025. That's all.

[Jeff Jarvis]

He's. If he has a lawyer. Leo's just joking. He's not really saying.

[Leo Laporte]

Is that. Would that be liable to say you're the Annie Del. Well, say you're a con man, I guess.

[Jeff Jarvis]

Yeah, that'd be.

[Leo Laporte]

Yeah.

[Jeff Jarvis]

Yeah.

[Leo Laporte]

You're not a con man. You're just a.

[Paris Martineau]

That we know.

[Leo Laporte]

Prevaricating son of a gun.

[Jeff Jarvis]

You're a clever. Which is. Which is otherwise known as a good marketer.

[Leo Laporte]

Yeah. No, it's brilliant. Marketing isn't it right.

[Paris Martineau]

I mean, I think people upset.

[Leo Laporte]

That's good.

[Paris Martineau]

Marketing threads about getting expelled from Colombia are certainly good.

[Jeff Jarvis]

And maybe Colombia could have had a piece of this ip, but no more because Colombia is just doing everything wrong these days.

[Harper Reed]

And maybe we shouldn't praise just good marketing. Like, why is that good?

[Leo Laporte]

No, no, I agree. Agree. I agree. No, no, no. I'm not praising it in that sense. I'm just saying the whole thing is an exercise in marketing. Don't. Don't take it.

[Paris Martineau]

Rather than anything about the technology.

[Jeff Jarvis]

Yeah, yeah.

[Leo Laporte]

He raised $5.3 million for Cluly.

[Paris Martineau]

Oh, boy.

[Leo Laporte]

Yeah. Columbia maybe should have thought a little harder on this.

[Jeff Jarvis]

He's money these days.

[Leo Laporte]

Yeah. Abstract Ventures and Susive ventures gave him $5.3 million. It's an interesting thought. So what did he do? He was using it.

[Jeff Jarvis]

If you called this agentic AI, maybe he would have gotten a billion dollars, but that's what it is.

[Leo Laporte]

Part of the reason he wrote it, I'm sure, is these ridiculous tech interviews. Right?

[Paris Martineau]

Yeah.

[Leo Laporte]

Where you're asked to perform like a dancing monkey in ways that don't really tell the interviewer or the company anything about you.

[Jeff Jarvis]

Right.

[Leo Laporte]

And to come up with a. If you can cheat on it, then that just shows you how dumb it is. Right. Sales calls. Yeah.

[Jeff Jarvis]

Or also customer service calls. If it. If it actually listens to what I need and pleases me and has the authority.

[Leo Laporte]

Yeah. I'm not sure agentic AI is going to be a bad thing for things like that if it gets the job done. Harper's point, which I think was well taken, is that we have these kind of. They're almost prejudices about how things ought to be and that really we just need to be a little more open about a world that is changing rapidly. Maybe it isn't good, but we gotta. But we gotta play with it to know. We gotta try it to know. I don't know. He's like me. He doesn't know what the end game's gonna look like. Maybe this is all wasted time. But it's. But it's interesting to experiment with it, I would say. Did you see how chat GPT solved 5 years of jaw clicking?

[Paris Martineau]

I did see this.

[Leo Laporte]

This was really interesting. Now this is actually a known fix, but somebody posted on Reddit. I've had jaw clicking on the left side for over five years, probably from a boxing injury, and every time I opened my mouth would pop or shift. Yeah, that's a humble brag.

[Paris Martineau]

No, the ChatGPT subreddit just to be clear.

[Jeff Jarvis]

Oh yeah, I just, I just started getting this two weeks ago, so I'm looking very good.

[Paris Martineau]

Listen to the instructions.

[Jeff Jarvis]

Very eager to hear.

[Leo Laporte]

I figured it was permanent damage. He's been to a lot of doctors asked. You know, yesterday I randomly asked ChatGPT about it. It gave me a detailed explanation, saying the disc in my jaw was probably just slightly displaced but still movable, and suggested a specific way to open my mouth slowly while keeping my tongue on the roof of my my mouth and watching for symmetry. I followed the instructions for maybe a minute max and suddenly, no click. I opened and closed my jaw over and over again. It tracked perfectly. Still no clicking today. So that by itself. Yeah, yeah, fine. But then all the responses. Oh my God, it worked for me too. Mine does this too. Keep tongue on mouth. Holy cow, this worked for me.

[Jeff Jarvis]

Well, maybe it learned it from a, from a doctor. Oh, it did.

[Leo Laporte]

Oh, by the way, if you continue down the thr, this is a known.

[Jeff Jarvis]

So it's not the AI that's smart.

[Leo Laporte]

The AI didn't invent it. But the point is that the AI was able to scan vast amounts of information.

[Jeff Jarvis]

Gee, I went to Google and asked this. Or gee, I went to a library or I went to, I don't know, a dentist.

[Leo Laporte]

Yeah, I'll give you some PDF links on this from, for instance, the TMJ exercises from the National Health Service. It's in there. Explains it in more detail. Navigation. Obviously ChatGPT ingested this, but the point is you could go to 100 doctors and they may not know this or they may have probably forgotten it. But ChatGPT doesn't forget everything and knows everything. And by the way, I mean this Reddit thread. Thank you. This worked for me. I've had this for 10 plus years. It's amazing. Physical therapists respond. The tongue positioning of the roof of your mouth actually increases the activity of the muscles in your jaw called the pterygoids. This stabilizes jaw mechanics as you open. Pretty important. Now, obviously it's not going to work for all kinds of jaw clicking or tmj, but I was blown away by the overflow number of responses. Wtf? It worked for me too. I thought this was a forever thing for me. How many people have this? And the information's out there, but it's just, I mean, it's just out there. How do you get it?

[Jeff Jarvis]

So I was, I was thinking about how do I get it to do it? It doesn't happen every time. How do I get it to do it for the Dentist. So the dentist can figure it out. So this is.

[Leo Laporte]

Yeah.

[Paris Martineau]

Well, why haven't you found this out?

[Jeff Jarvis]

You know, I just had the, I had the, the, the boxing match on, on CNN with Scott Jennings. And then after that.

[Leo Laporte]

Why do you go on with that right hook? Why do you even give that guy.

[Jeff Jarvis]

I know. I was on again last night.

[Leo Laporte]

I have unsubscribed now to Washington Post.

[Jeff Jarvis]

Whoa.

[Leo Laporte]

Well, you did, too.

[Jeff Jarvis]

No, I, I, I still subscribe.

[Leo Laporte]

Oh, no, I unsubscribed. That was it. The information. I unsubscribed. I stopped watching cnn. Especially now I msnbc. The same thing's going to happen. They fired all those people. I'm sick of this. And I'm going to use my you're mad as hell anymore to vote by walking because I have no power at all in any of this. But I'm a pretend. I do. We does feel a little.

[Jeff Jarvis]

You did look like the old man with the balloons right there.

[Paris Martineau]

You've got to get your computer to do that little balloon thing and then it'll really.

[Leo Laporte]

I can turn that on. Sure.

[Paris Martineau]

I don't know what the command is.

[Leo Laporte]

For it, but I don't either. But you can actually do it from the, the command line. It's an, it's an apple. It's an apple.

[Jeff Jarvis]

Oh, it's an apple thing. Not a zoom thing.

[Paris Martineau]

You haven't. You're on an apple right now.

[Jeff Jarvis]

I can't. Why doesn't this. Did you turn off the ability for the zoom to do that? Oh, there we go.

[Paris Martineau]

Can you look sad because your wife just died?

[Jeff Jarvis]

Angry. A little grumpy. Grumpy, too.

[Paris Martineau]

Yeah.

[Jeff Jarvis]

No, no, no. Too much, too much. Back a little bit. There you go.

[Paris Martineau]

Oh, I really like that. That's really good. The slowness of the balloons is really what does it for me. It's not. They're not rising fast. It's just kind of.

[Jeff Jarvis]

See, what's irritating me is now Leo is going to be either from that or from the moral panic. He's going to be alone on the card.

[Leo Laporte]

Oh, my God. This is good, too. You're kicking this off with the balloons. Little rain and balloons.

[Paris Martineau]

Yeah. How do I turn these on?

[Leo Laporte]

You want to make.

[Paris Martineau]

I am.

[Jeff Jarvis]

Yeah, I am, too.

[Leo Laporte]

Okay. There's a little thing in your menu bar that is the. Maybe it's a zoom or it's on the camera.

[Harper Reed]

Times have changed. Everyone used to be like, how do I turn this off?

[Leo Laporte]

Now everybody wants to turn it on.

[Paris Martineau]

Yeah, Yeah, I know. I turned it off at Some point for this reason, but now this seems fun.

[Jeff Jarvis]

Is it system preferences?

[Leo Laporte]

I can never trigger it. It was funny that Harper had it on and triggered it. Yeah, it probably is. I don't know. It's in my menu bar, but I don't know. It's cuz my camera has a menu bar probably.

[Paris Martineau]

Wait. You know, you guys recent Twit recently turned 20. Do you want to know it also turned 20?

[Leo Laporte]

YouTube. I thought it was for me. I went to YouTube and I saw a cake. I thought that was nice of them. Celebrating our 20th anniversary. So we started at the same time as YouTube.

[Paris Martineau]

Apparently slightly before.

[Leo Laporte]

You didn't know that there's a little, little logo here, a little dancing going on. Party animals join us in celebrating 20 years of creators, artists and transforming video on YouTube. Now I'm going to press this button. I don't know what's going to happen. It says cut the cake.

[Harper Reed]

I wonder if they're going to let us.

[Leo Laporte]

Oh, YouTube. YouTube. Rick rolled me. Wow.

[Paris Martineau]

A rare on screen Rick roll.

[Leo Laporte]

That's not nice. YouTube not okay.

[Harper Reed]

The funny thing originally but made YouTube big is that they were. That's how you got Rickrolled is they sent you a YouTube link.

[Leo Laporte]

Yeah, it probably helped.

[Jeff Jarvis]

That's how YouTube got big.

[Leo Laporte]

I think it was dick in the box, but okay. You know we could fight over that. We had. It's so funny. They were running it still until recently in Canada when I was doing the show. Call for help in Canada. In Toronto, my co host Amber McArthur. This is around 2004, 2005 brought in. She said, oh, you got to see this new site. It's called YouTube. You can share videos. And I'm going, oh, this is wow, what a great idea. This is so cool. But that's fine in 2004. But when you air that in 2021 makes us look like complete morons. Wow. What is this YouTube thing of which you speak? I was glad they finally, I think stopped airing it. China of all places is cracking down on autonomous car claims. We look at how many fatal accidents here in the U.S. yeah.

[Jeff Jarvis]

Jesus.

[Leo Laporte]

Authorities in China now prohibit the use of autonomous and smart driving in car ads. Wonder if they'd allow autopilot. It's funny because the DMV has accused Tesla of falsely portraying its vehicles as autonomous based on the language on its website. But it didn't ban it.

[Paris Martineau]

Hey, but according to Elon Musk on the earnings call this week in June we're gonna have autonomous model Y's driving around Austin.

[Leo Laporte]

So what do you. What do you. What's going on in his brain? Do you think that's.

[Paris Martineau]

That's a big question.

[Jeff Jarvis]

Worst living through chemistry.

[Leo Laporte]

I mean, 71% drop in revenue year over year for Tesla.

[Paris Martineau]

He promises he's going back.

[Leo Laporte]

He says I'm gonna only do Doge two days legitimately.

[Paris Martineau]

Recently, like, minute. Maybe this is just my news alerts, but I got the news alert. 71% drop in revenue at Tesla. And then the news alert. Elon Musk promises he's leaving Doge and going back to Tesla as soon as possible.

[Leo Laporte]

It's like, yeah, but take a look at the Tesla stock price. 71% drop in revenue. The stock price went up.

[Paris Martineau]

Well, it went up because Elon said he's going back to Tesla. It went down and then it went back up.

[Leo Laporte]

It's a meme stock. It's not a real stock.

[Jeff Jarvis]

No, it's not.

[Leo Laporte]

Because honestly, I mean, I don't. I don't own any Tesla stock or anything else, but does seem to me not a great investment. Look, here's the even more clear graph.

[Jeff Jarvis]

There was an actual ETF where you could buy the ETF to short it. So I did, for 200 bucks, for.

[Leo Laporte]

Oh, you're one of those shorters. For the longest time. Because I was a big Tesla fan. I remember I cried when I went to the. When I picked up my Model x back in 2015 and I got the tour of the factory, I was in tears. I was so moved by the mission.

[Jeff Jarvis]

And then your wife cried when the door hit her in the head.

[Leo Laporte]

But there was a. There were. There were many more tears shed for various reasons. She hated that car. She called it Christine. And after three years, I did turn it back in. So I no longer drive a Tesla, Although I've. It's a fun thing. Thing. My son, who, as you know, is moved to New York. He's opening his restaurant, by the way, made Memorial Day weekend, May 31. Oh, grand opening of Salt Hank Sandwiches on Bleecker street, right next to John's. And apparently a mortuary is right there, too.

[Jeff Jarvis]

Yes, there is. That's right.

[Leo Laporte]

I've been watching Henry's vlogs. Anyway, when he left California, he left his Tesla behind, right? He's not going to drive in New York. He didn't want to have a car in New York. So I have a Model Y in my driveway.

[Jeff Jarvis]

Oh, no.

[Leo Laporte]

And I've been watching.

[Paris Martineau]

Ask yourself, Model Y.

[Leo Laporte]

Why? Why? I'm watching the price. You know.

[Jeff Jarvis]

You know what, Leo? Here's a Way to make money with it. You charge $250 each to come with a sledgehammer. Oh, and smash it.

[Leo Laporte]

Oh that's. I'll ask Henry if he'll let me do that. What a good idea.

[Paris Martineau]

That could be one of the hours on our 24 hour live.

[Jeff Jarvis]

Yes, it could be.

[Leo Laporte]

Destroy Henry's Tesla.

[Harper Reed]

Yeah.

[Leo Laporte]

We already know this. Thank you.

[Harper Reed]

Prepares to hit.

[Jeff Jarvis]

Yes, that's perfect.

[Harper Reed]

$10 for me to hit it with a sledgehammer. I would absolutely.

[Jeff Jarvis]

How much might it get now? So how many sledgehammers? Let's figure this out.

[Paris Martineau]

Yeah, let's figure out how many sledgehammer hits it would take to fully destroy.

[Leo Laporte]

And then I got a CarMax for $22,000. So. But I don't know if that's good anymore. I'm gonna. Because it's been a week or two and I think it's. I think it's plummeting. I sent Henry a note saying dude, help me sell this car. The price is plummeting. That's a lot of. I think it's quite a few.

[Jeff Jarvis]

A lot of happy people.

[Leo Laporte]

People though I know suggest it.

[Paris Martineau]

I asked Chad gbd how many sledgehammer hits would it take to total a Tesla model Y And it wrote a novel. It wrote a thousand words.

[Leo Laporte]

But it's thinking, it's reasoning. And the answer is.

[Paris Martineau]

Well, it says there's a. Depends on a few variables. Depends on what counts as totaled. It depends on the answer who's swimming damage by section totals. A strong person could total a Model Y in 50 to 100 well placed sledgehammer hits. If you're only smashing body parts randomly, it might take several hundred do an insurance total level damage. Want a breakdown based on a specific scenario like one person, one hour ten pound sledgehammer and then. And then there's a pop up that chat GPT asks do you like this personally?

[Leo Laporte]

Is this something you want? Are you violent? So here's perplexity. It would take 32 full force sledgehammer hits to cause significant structural damage. It's got steel plates in the cabin and mega castings. Look at this. I got a table. Yeah. See, I couldn't charge enough. I'd have to charge like $500 a hit to really make it.

[Jeff Jarvis]

I think you'd find people who I think it's not.

[Paris Martineau]

I think that they're overestimating the strength of the person using the hedgehammer. I think.

[Leo Laporte]

Think that maybe if I give him just a little hammer. Wait a minute. Why am I giving him something we can do. Let's give him a baseball bat.

[Jeff Jarvis]

Baseball bat.

[Harper Reed]

Wooden baseball bat.

[Jeff Jarvis]

Yeah, yeah, there you go.

[Paris Martineau]

Baseball bat.

[Leo Laporte]

Profit.

[Paris Martineau]

Yeah, that's profit.

[Leo Laporte]

Profit.

[Jeff Jarvis]

Okay.

[Leo Laporte]

I'm gonna message Henry. If he agrees to it, we'll do it. We'll stream the whole thing.

[Jeff Jarvis]

Probably the battery will go on fire and it'll be fun to watch.

[Paris Martineau]

Will be great. They can see us blow up live on air. That'd be great.

[Harper Reed]

And you can charge locals more so they can do it themselves.

[Paris Martineau]

This could be the new version of the twit brick. You could take home a piece of Henry's Tesla that you hit off yourself.

[Leo Laporte]

Oh, that's right.

[Paris Martineau]

Oh, or we could throw bricks.

[Leo Laporte]

If you get it, there you go. And if you get a piece, you get to keep it.

[Paris Martineau]

That's true.

[Harper Reed]

All good ideas.

[Leo Laporte]

I thought it was interesting that Harper said. Yeah, eventually we won't have to worry work and then it'll be a universal basic income. Apparently he's not alone with this idea. A company is being founded called Mechanize. And the premise. This was founded by, according to TechCrunch, famous AI researcher Tame Bessel Roglu. The premise is the full automation of all world work. Not just some, but all work.

[Paris Martineau]

I'm sorry, I laughed at Burke's comment in the discord, which is unrelated but is related to the Tesla, and I have to say it now. He said in all caps. I have tried to destroy a car with a sledgehammer. It is harder than you think.

[Leo Laporte]

Yes.

[Paris Martineau]

Knuckle scars to prove it.

[Leo Laporte]

It's. No, it's just like strangling a person. It's harder than you think. Think very similar. I'm not claiming anything.

[Jeff Jarvis]

All right. Why did you try to destroy the car with a sledge?

[Leo Laporte]

That sounds like Burke. Oh, he used to work. I know why he used to work with Mark. Pauline the world famous. Was it with. When you were doing it with Pauline the world famous?

[Paris Martineau]

What. What line of work lead to Burke destroying a car with a sledgehammer?

[Leo Laporte]

Mark Pauline was the founder of an art project called Survival Research Labs, srl. And they would do scary.

[Paris Martineau]

Oh, this is a sick website.

[Leo Laporte]

No, no, he's great. They would do. No, this guy's an artist.

[Paris Martineau]

Yeah, I'm saying it's awesome.

[Leo Laporte]

Incredible shows underneath the overpass in San Francisco in which fire eating machines would destroy each other.

[Paris Martineau]

And I. Burke said it was completely unrelated to work. It was in high school.

[Leo Laporte]

Oh, Burke, did you used to work for Mark?

[Jeff Jarvis]

Did you get arrested? Burke.

[Paris Martineau]

He had consent.

[Leo Laporte]

Okay, that sounds like a high school.

[Jeff Jarvis]

Well, Joe Esposito used to smash cars in his dad's wrecking yard.

[Leo Laporte]

Oh, this Joe would know about this. Joe, that's great. You can be our local expert.

[Jeff Jarvis]

Yeah.

[Leo Laporte]

Come on down. Help me out. Henry's gonna kill me.

[Harper Reed]

You wanted to move to New York.

[Paris Martineau]

Now we gotta.

[Leo Laporte]

Okay, we got to do something with this model Y.

[Jeff Jarvis]

It's also bad for your reputation to have that in the driveway, Leo, so you got to get it out of there.

[Leo Laporte]

I know. I keep waiting for somebody to spray paint it.

[Jeff Jarvis]

Yeah.

[Paris Martineau]

Just imagining Henry trying to explain this to someone, being like, yeah, you know, I had to move to New York to open up my TikTok sandwich shop. So now my dad is live streaming people on the Internet attacking my Tesla with hammers. They're like, oh, another day.

[Leo Laporte]

All right, let's take a little break. More AI news and other news still to come. Paris Martineau is here. The wonderful Paris Martineau. You. I. Have you done something with your hair?

[Paris Martineau]

I have a different light behind me that is reflecting kind of a red thing.

[Leo Laporte]

It looks good. It really looks good. I like it.

[Jeff Jarvis]

It brings out the auburn.

[Paris Martineau]

It does. It's a plan. It's the light I normally use for my plants, but it is just in a good position.

[Leo Laporte]

You're going to grow good plants.

[Paris Martineau]

I'm getting taller.

[Leo Laporte]

Good for people.

[Jeff Jarvis]

Paris, you know, is taller than you thought.

[Paris Martineau]

Yeah, I'll be 6:1 by the end of this podcast.

[Leo Laporte]

And I'm getting shorter, so there. That is Jeff Jarvis to her right. He is the professor emeritus of journalism, but also the author of fabulous books like the Web We Weave and the Gutenberg Parenthesis and Suni. Stony Brook. Are you going to be teaching? Have you found out yet? In the fall, or is that not.

[Jeff Jarvis]

No else is going to teach the course. I'll probably be lecturing the course, but it's the course I wrote. I was gonna say this with Harper in AI and Creativity. The dean told me that it was the most popular sign up yet in the term.

[Leo Laporte]

I think this is the future, Jeff. This is absolutely the future. And I think I really liked Harper's attitude, which is, hey, we don't know, but. But play with it. Experiment with it. Yeah, there's something going on. I mean, we all. Even you, Paris, Escape Skeptic. I have to say. There's something going on. Right?

[Paris Martineau]

Yeah. No, I will say that that conversation with Harper made me want to design a. See if I can design a video game, see.

[Leo Laporte]

And play around with it. Get to know your local.

[Jeff Jarvis]

Here's the Question I have. Once you. Once it codes it, how do you make it executable and where do you put it?

[Leo Laporte]

It codes an executable. Well, it depends what you have it right in.

[Paris Martineau]

If it's gonna say I understand coding, but I. My knowledge of coding is so little like little that I don't. I can't even comprehend any of it.

[Leo Laporte]

You're better off off.

[Jeff Jarvis]

Where do you host the game is what I'm trying to ask. In par.

[Leo Laporte]

You give it. You people can download it and play it or you. Well, you have to tell it. If you want it to be a web based game, it can do that. You tell it where you want to, how you want to deliver it, how they want to package.

[Jeff Jarvis]

All right.

[Leo Laporte]

Can you help me build a game? He's very slow. Absolutely. What kind of game do you want to build? Do you have a specific concept or idea such as the genre, gameplay mechanics or platform? Let me know what you're thinking and I can.

[Paris Martineau]

Flappy bird with Jeff Jarvis's face.

[Leo Laporte]

I want to do a flappy bird clone. But instead of a bird, I want it to be a flying journalist professor with white hair. That sounds like a fun and creative twist. Here's a basic outline to get you started on your flappy bird clone featuring a flying journalist professor. 1. Game design concept character. Instead of a bird, you'll have a flying journalist professor. Consider what costume or gear like glasses, a notepad or a pen. This character might have to enhance the theme. Obstacles. You could use various obstacles based on journalism like stacks of newspapers, camera tripods or even giant pencils. Two Game mechanics Controls. Players tap the screen or press a key to make the character flap run, rise release to let gravity pull them down. Scoring. Players earn points by passing through gaps between obstacles. Powerups. Consider adding items like scoops that give extra points or temporary invincibility. You got scoop development tools. Game engine. You can use engines like Unity or Godot for this type of 2D game. Both have strong communities and lots of resources for beginners assets. You'll need writing with a backgrounds. You can create your own or find free resources on online. 4 Basic Game Loop. Create the main game screen, implement character controls and just do it. Well, it's giving me an outline of the things I need to solve. I can't figure out how to shut it up.

[Harper Reed]

And then you need to see that.

[Leo Laporte]

Oh no. I have the transcript. That's fine. Thank you very much. Bye.

[Paris Martineau]

You just wasted how many millions of dollars?

[Leo Laporte]

I burned a forest for that baby.

[Jeff Jarvis]

Just saying please and thank you. Costs the millions of dollars.

[Leo Laporte]

Isn't that hysterical? We talked about the. That last week, didn't we?

[Paris Martineau]

Or on Twit.

[Leo Laporte]

On Twit. That's why Paris was there. She remembers it. Yeah. Sam Altman said that every time you say please and thank you, it uses more CPU cycle, obviously, because that's the thing of a response. And he says it's costing us tens of millions of dollars in aggregate. Thank you, Sam. A new way to ruin your life. All right, we're gonna take a break. We gotta wait.

[Jeff Jarvis]

Wait, wait, wait. No, I've got it. It's a game where you're smashing the Tesla.

[Leo Laporte]

Oh, you're going to be rich.

[Jeff Jarvis]

Rich, I tell you.

[Leo Laporte]

Actually, that would probably do very well.

[Harper Reed]

They'll sue you if you use the Tesla IP in the game. You're not going to get over.

[Leo Laporte]

We don't call it a Tesla, we just call it a.

[Harper Reed]

Then it's not going to sell because that's the selling point.

[Jeff Jarvis]

We call it a cyber plunk.

[Leo Laporte]

I'll release it anonymously. Let's erase the last 10 minutes. This episode of Intelligent Machines is brought.

[Paris Martineau]

To you by Silly Vibes today.

[Leo Laporte]

Silly Vibes brought to you by StoryBlock. Let's talk about you've all you both worked, I'm sure with legacy content management systems. Cmss. Ugh, they can be so painful. Yeah, they promise enterprise grade features, but they're slow, clunky and of course the smallest change, smallest update. You got to get a developer involved and when you're trying to move fast, that's the last thing you want to do. Well, StoryBlock is brand new and it changes all of that. Unlike like those kind of monolithic, inflexible CMSs. Storyblock. So cool. I know a little bit about this because we did. This is a headless, a headless cms. It completely decouples the back end from the front end. So developers can build the front end in any framework they want. They can use React, Astro Vue, whatever, marketers can even use it. It's so easy to use. They can use an intuitive visual editor to create and update content without having to get a developer involved. Don't have to file a ticket. StoryBlock scales whether you're a freelancer or part of a global enterprise. And by the way, they are a global CDN content delivery network. They use AWS data centers all over the world, us, Europe, Asia. So it's built for performance and at performance, at scale. Where do you see some of the people who use StoryBlock? StoryBlock really is enterprise ready. With role based access control, of course, enterprise SLAs and top tier security. It's the stuff Fortune 500 companies demand. One global E commerce giant. You would know the name. Switched to Storyblock and cut the content update cycles from weeks to hours. Another major brand, a name you'd know again. Empowered marketing to launch campaigns completely independently, freeing up the developers for bigger projects. StoryBlock is API first and this is such a great way to do it. Your content loads fast anywhere in the world. World. It means better UX design, higher engagement, improved SEO. And with their real time visual editor, marketers see exactly what the content will look like before publishing. No more endless back and forth over minor tweaks. You can move something, change a color, do it all yourself and you'll see immediately what it looks like. Developers get fewer interruptions, marketers get more autonomy. It's a win all route. If you're an agency, StoryBlock offers multi client workspaces, flexible permission missions, seamless collaboration tools. You can manage multiple projects without disrupting development workflows. It's all in there ready for you to use. So whether you're a startup, an enterprise, an agency juggling multiple clients, Storyblock. Remember that name. StoryBlock gives you the power and flexibility you need. Try it today. Storyblock.com TwitTV25 with the offer code TWIT25 to get you 20% off for three months on the Growth and Growth plus plans. Storyblock.com TwitTv-25 offer code TWIT25 for 20% off the first three months on the Growth and Growth plus plans. S-T O R Y B L O K.com TwitTV-25 Again, the offer code is Twitter25. Thank you Storyblock for supporting Intelligent machines.

[Jeff Jarvis]

Will you indulge me a quick CMS story?

[Leo Laporte]

Oh God. Wait.

[Paris Martineau]

Brief aside first, and then this. I asked ChatGPT to make me a flappy bird with Jeff's face. It required some troubleshooting. It ostensibly works, but I don't know what the controls are. So every time I open it, it's just a photo of Jeff's face that immediately falls to the ground and then says game over. And I do think it says something about what's going on.

[Harper Reed]

You need to prompt what you want the control be. Just tell it what the.

[Jeff Jarvis]

Should we have a screenshot?

[Paris Martineau]

I will. Hold on a second.

[Leo Laporte]

What are you saying?

[Paris Martineau]

That story. I just. I just asked Chat GPT generate code for simple flappy bird like game but with this Man's face as the bird.

[Leo Laporte]

It's kind of mindboggling, isn't it on I'll.

[Paris Martineau]

I'll ask it to do the controls if you want to tell your CMS story.

[Leo Laporte]

And then we will put it up on. On our discord for our club members.

[Jeff Jarvis]

Yay. Yeah.

[Leo Laporte]

One of best benefit can make membership does Jeff. I don't know how to publish Jeff's arms flap or whatever.

[Paris Martineau]

I mean we'll find out.

[Leo Laporte]

So. CMSs are notoriously troubling, especially the old school ones.

[Jeff Jarvis]

So I go way, way, way back with this. My first newsroom system was 1974 kids. And when I started Entertainment Weekly I didn't like the computer department and my wife is brilliant. And so we set up the first major all Macintosh Macintosh system.

[Leo Laporte]

Nice.

[Jeff Jarvis]

And I, I. It's part of my liner type book, believe it or not because it's kind of the end of the liner type story is desktop publishing. So this. She went to, to the Seabold conference and presented.

[Leo Laporte]

Oh yeah.

[Jeff Jarvis]

Work. Right. All these things.

[Harper Reed]

Right.

[Jeff Jarvis]

So we had no management system. Oh, that changed the white balance.

[Leo Laporte]

What, what is, what does your wife do? You know, she talked about her.

[Jeff Jarvis]

She trained. So she was a special ed teacher.

[Leo Laporte]

Right.

[Jeff Jarvis]

And then she went to work at Time Inc. That's where we met. And she trained all of the editors and correspondents on computers.

[Leo Laporte]

So this is, this was her thing.

[Jeff Jarvis]

Oh yeah. That's what she did.

[Leo Laporte]

Well that's cool.

[Jeff Jarvis]

We met and her department hated me and and, and so we went to it.

[Leo Laporte]

Why did they hate you? Just had A, because I.

[Jeff Jarvis]

Because I was constantly fighting them for the system they wanted and then, but then I got my room revenge. A, because I had married her and we took her away from that department and B, because when we started EW she and we've been talking about lately because we're writing this up. It was audacious in ways we didn't know. It was the very beginning days Quark was brand new.

[Leo Laporte]

Oh yeah.

[Jeff Jarvis]

That was the page.

[Leo Laporte]

That was the desktop publishing app.

[Jeff Jarvis]

But here's the cms. Part of it is that there was no content management system. Right.

[Leo Laporte]

So no way to interface to it.

[Jeff Jarvis]

And we had servers with content on it, but there's no way to. To manage it. So we invented what we called the Conch system. It was a physical vanilla folder that had the name of the story on it. And unless you had the conch, you couldn't speak, you couldn't edit it.

[Leo Laporte]

Oh that's. That's cool.

[Jeff Jarvis]

That's how we did it. So this has all been memory lane for me. The Computer history museum in 2017 did a phenomenal multi day all recorded pioneers of desk desktop publishing. I watched all 12 hours of it. It. Is.

[Leo Laporte]

It online somewhere?

[Jeff Jarvis]

Yes, it is. It's all online on YouTube. Computer History Museum. Pioneers of desktop publishing. Yeah, desktop publishing. And. And they have all kinds of other material up and it's just wonderful.

[Leo Laporte]

The reason you needed the conch, and this is historically a difficult challenge, is if two people are working on the same phone file, it really becomes a problem because they can make different changes and which one should they keep and so forth. So they. So it was a clever way to say you only. You can touch this file right now, and then you hand the conch the little shell off which people use in real life during round table discussions and things.

[Jeff Jarvis]

Did you know of Jonathan Sebold and Seabolt seminars?

[Leo Laporte]

Oh, yeah. Oh, yeah, absolutely.

[Jeff Jarvis]

So that was. Yeah, I'm talking to Jonathan soon. So she was invited to go present this out everybody because it saved us $3 million a year with the computer department at timing.

[Leo Laporte]

We. We still have that problem. We had an intern design a sales system for us and only one person can be in it at a time. So Lisa has to go out onto our slack saying everybody off the sales system.

[Jeff Jarvis]

She was the conch.

[Leo Laporte]

Only this is the. We got to get a conch. That's what we need, a conch.

[Paris Martineau]

Flappy bird game is a bust, guys.

[Jeff Jarvis]

No, it.

[Paris Martineau]

I can't get it. Part of the. You guys know how flappy bird works? The bird starts. You got to click for it to go through. There's something about the bird is like too heavy. And the amount it jumps when you click is either too little or too much to where I finally gotten it by asking it some very rudimentary questions to where I can kind of click to make the bird flap. But it just shoots off the screen either way and it's too hard to control. It turns out designing a game is harder than writing.

[Leo Laporte]

It's a very accurate flappy bird, actually. Flappy jet True. Ah. And we have. We have Pretty fly for psy. Guy has. Or CIS Guy has created a box cover. Flappy bird. Flappy Jeff, I think we're going to call it. And we're ready. We're ready.

[Jeff Jarvis]

Yep, I'm ready.

[Leo Laporte]

I think that this is a good project for you, Paris, to take what you've learned from Harper Reed and all of these conversations and to work on this through the night until you get a full term.

[Paris Martineau]

I'm going to make. Maybe I will a whole new.

[Harper Reed]

That give you an appreciation of how many details there are actually are to creating a game.

[Jeff Jarvis]

Like there are to think all that.

[Harper Reed]

Yeah.

[Jeff Jarvis]

You have to figure out how.

[Harper Reed]

How much you want it to flap when you press the thing and how. How much or how did all that stuff.

[Leo Laporte]

I've written a few games. I wrote a snake game, a soku snap solver. I've written.

[Harper Reed]

Do you.

[Paris Martineau]

Do you only write names that began with S?

[Leo Laporte]

Oh, I guess S. And then I got to the T in the Alphabet and next I'm like, yeah, something called Ulysses. Okay, I think we're now to the rest of the story.

[Paris Martineau]

And now the rest of the story.

[Leo Laporte]

The rest. Oh, you know that. Do you know. You know. Know him?

[Paris Martineau]

I know him.

[Leo Laporte]

Paul Harvey. And now page two.

[Paris Martineau]

I'll see you on the radio.

[Leo Laporte]

I'll see you on another. I tell you the story about my friend who met Paul Harvey. He was in Chicago. Paul used to work at. I can't remember what was it, WSJ or whatever it was in Chicago. And he was in Chicago. He got in the living elevator and there's Paul Harvey standing there. Everybody knows Paul Harvey, world famous newscaster. They go up in the elevator. He gets to his floor, starts to get out. Paul Harvey says, goodbye, American. It must be nice to know you're. You know, you're. You're. You're recognized. Your catch line is going to be recognized. Welcome to Slop World. How the hostile Internet is driving. Driving us crazy from the Financial Times. The last. This is a good illustration. The last bits of fellowship and ingenuity on the web are being swept away by a tide of so called self. No, not self. Artificial intelligence from Kierkegaard's journals. In 1843, suppose someone invented an instrument, a convenient little talking tube which, say, could be heard all over the land. I wonder if the police would not forbid it filling. The whole country would become mentally deranged if it was used. Kierkegaard's talking tube. It's famous. No, it's not. Well, I presume you put this in Jeff. This feels like a very. Jeff.

[Jeff Jarvis]

Yeah, it's very essay. Ish, but it's. It does speak to the problem of the web now, right? It's being filled up with slop of all sorts. Not just computer generated slop, but slop. Slop.

[Leo Laporte]

I think it's important.

[Paris Martineau]

You're calling Flappy Jeff slop. It's a game that doesn't work at all. It's a perfectly.

[Leo Laporte]

I love to hear you say that, Jeff, because you have so long defended the Internet.

[Jeff Jarvis]

I do defend the Internet. I want to defend it from slop.

[Paris Martineau]

It is full of slop now, that's true.

[Leo Laporte]

But a lot of that slop is output put by plain old human beings. I mean, it's not all AI slop by any means.

[Jeff Jarvis]

So it has this nice analog in here. Metaphor. The hostile Internet has a meat space analog in New York City's Moynihan Train hall, which you'll see when you're in New York. A 1.6 billion 486,000 square foot station unveiled in 2021. The building was supposed to be an homage to the original, much mourned Penn Station. But on a human level, the new facility is a disaster. Like so many other places designed by the principles of hostile architecture. There's almost nowhere to sit, lest a homeless person might find a place to take a nap. There are plenty of places to shop and spend money, along with the requisite phalanx of surveillance cameras. Enormous high resolution screens circle the main atrium, broadcasting constant ads. Train times are displayed on smaller screens strewn around the building. Like the Moynihan Train Hall. Today's Internet isn't really designed for us, but rather to elicit certain responses from us. Responses which, to put it loftily, are hostile to human flourishing. I think that's going a little far, but I think that it has been taken over by, as it says here, a witch's brew.

[Leo Laporte]

I would say, though, that the response to that is it is up to all of us to maintain our mental health by tuning out the stuff that we feel is causing those problems. I, you know, I don't go to x.com anymore. I don't go to Instagram. I mean, I think that that's. If. Are you saying people are unable to tune that stuff out or to turn that stuff off?

[Jeff Jarvis]

Well, it's harder to.

[Leo Laporte]

Newton Minnow said that about tv. It's a vast wasteland. Okay, well, just turn it off and.

[Jeff Jarvis]

I make his daughter mad. I make fun of it.

[Paris Martineau]

Tv. In this case, the stuff is at least being made by humans and pitched to someone who agreed it had some value of some kind. Whether or not that's traditional value or just could be something to consume and brain rot. I think what Jeff is talking about, one example of this, I mean, you just said you're not on Twitter anymore, which is a good choice, but I'm sure this happens on other sites. But I've most recently noticed it in Twitter. You'll go and look at a post and the replies are all clearly like accounts using chat bots to reply. They all use kind of the same phrasing or have the same formatted answers. Then you go to that profiles page and they've produced like 200 of those types of replies on different people's tweets. Like that day, it's. It's just. There's slop everywhere. It's just making the Internet, Internet more noisy in a way that is frustrating.

[Jeff Jarvis]

Reading again, the scholar John Durham Peters made the connection explicitly in his 2010 paper, Broadcasting and Schizophrenia. What was once mad or uncanny is now routine. Hearing disembodied voices and speaking to nobody in particular, he wrote. Prodded by slacking bosses, tempted by Instagram ads, trolled, provoked by inflammatory content, served up by recommendation systems tuned to do just that, we can become. Become our own worst selves online or some other self entirely surrendering to the libidinal forces of algorithmic mass media.

[Leo Laporte]

Yeah, but I mean, weren't news groups pretty full of crap even before?

[Jeff Jarvis]

And the newspapers were.

[Leo Laporte]

Yeah, newspapers were.

[Jeff Jarvis]

That was the whole point of mass media became full of crap. I know, I'm not agreeing with it, but I found it interesting. It was provocative.

[Leo Laporte]

It is provocative. And it feels like the same same old moral panic BS that people have been saying forever. President Moral.

[Jeff Jarvis]

Oh, I like that one.

[Paris Martineau]

Guys, when are we gonna get a sand? A sand reference. A sand.

[Jeff Jarvis]

Yeah, sand reference.

[Leo Laporte]

Yeah, yeah, yeah.

[Paris Martineau]

We need to get a sand button going on.

[Leo Laporte]

Look, this is just some old person who doesn't like the way the modern world has turned out. And you guys ought to turn down your music and cut your hair.

[Paris Martineau]

I think that there are real examples of this. For instance, Dropout tv, the subscription network I often plug because I like their shows. They have one show that I don't watch that often, but I'm aware of it. It's called the Smarty Pants site. It's kind of like a PowerPoint party presentation, but with comedians doing different bits. And there was a controversy recently because people online found that One of the PowerPoint someone presented used a bunch of AI generated art. And typically the company companies against that, and they'd issue a statement being like, listen, we do have a very strong sense against using AI generated art. We bought these photos from like, Getty or some. Some subscription service we use, thinking they were real, but we were tricked into using fake images basically. And that's a problem when people can't tell from what's fake and real. I think that that's an Issue Photoshop.

[Leo Laporte]

You know, so the Pew Research center, which, I mean, we like Pew, right. We trust their surveys, but Pew, yeah, Pew, Pew. They do a lot of surveys trying to understand.

[Jeff Jarvis]

That was a little, little view of 15 year old Paris there.

[Paris Martineau]

I'm. I don't know what the energy is today, but it's kind of, it's changing me somehow.

[Leo Laporte]

They did a survey of teens aged 13 to 17. They found their parents are generally more worried than their children about the mental health of teenagers today. Both groups call out social media impact on young people's well being. Parents are much more likely to make this collection. 48% of teens say social sites have a mostly negative. Oh, now Pew is asking me some questions.

[Jeff Jarvis]

Slop. It's all slop.

[Leo Laporte]

Pew, stop. I'm trying to read your goddamn story out loud. Loud. Thank you credit for it. Roughly half of teens, 48% say these sites have a mostly negative effect on people their age. But and this is Most important, only 14% they think they negatively affect them personally. It's those other people it's affecting very much like this story, by the way. One in five say social media sites hurt their mental health. But most teens say it makes them feel more connected to their friends. Like 80%.

[Jeff Jarvis]

Yeah. Which they've been saying for quite a while now.

[Leo Laporte]

Yeah. More teens say they spend too much time than before. It was 36% in 2022. Now 45% of teens say they spend too much time on social media. Not half, almost half. Teen girls are more likely than boys to say social media hurt their mental health. But it's still 25% versus 14%. Confidence 20% versus 10%. Girls versus boys. 50% of the girls say social media has hurt their sleep. 40% of the boys. I think social media has hurt my sleep. I've made a conscious effort to stop. If I'm tossing and turning at three in the morning, reaching over for my phone, I don't. I just toss and turn. Yeah.

[Jeff Jarvis]

Even I wouldn't do that.

[Leo Laporte]

I did that for a long time. That's why I bought all those stupid things on Instagram like this fine koala screen cleaner and microfiber cloth made in a America. That was an Instagram purchase. A majority of teens see social media as a positive space for friendships and creativity. 74% of teens say social media makes them feel more connected to their friends. 63% say it gives them a place to show off their creative side. It's the parents that are worried. 80% of parents. 80% of parents. Parents say they're extremely or very comfortable talking to their teens about their teens mental health. But only 52% of teens do. 34% of teens say they get at least sometimes get information about mental health on social media. I think these, this is a really useful kind of. It's not as bad as the parents think it is. Kind of a result.

[Jeff Jarvis]

Yeah. Cause the parents are in a moral panic.

[Leo Laporte]

Yeah. And I feel like that financial. Oh, oh, more AI slot ones already.

[Harper Reed]

You made me burn like three new ones already.

[Leo Laporte]

Good to burn them. You can use them again and again.

[Paris Martineau]

They're life changing favorites. Yeah, that one, that one's a top one for me because of the really jarring peel sound of Leo being peeled from the screen.

[Leo Laporte]

It shocks me. This is a show. You better watch the video I guess because we can just describe looks like it's me. But then a hand reaches in and peels.

[Jeff Jarvis]

You have to put that up on social.

[Paris Martineau]

Two big cartoony three dimensional fingers come in and peel Lego from the screen.

[Leo Laporte]

Are those all from Anthony Nielsen with.

[Jeff Jarvis]

A wet schloping sound?

[Leo Laporte]

Yeah. Anthony is our AI genius. In fact, if you want to know how he makes those, you should definitely tune in Friday at 1pm Pacific for our monthly AI User Group meeting. Anthony's going to explain a little bit about how he does these wonderful things. He's very good at this. I will be there unless I'm forced to sell my son's model Y at that time. Pick some. These are good. There's a lot of good stories in here. Thank you for.

[Jeff Jarvis]

Oh, I know the one I, the one I really like is what? Where to go here, where to go, where to go, where to go.

[Leo Laporte]

Oh, Kaggle, Kaggle, Wikipedia. So we've mentioned this. A couple of weeks ago the Wikipedia was getting swamped by AI bots scraping its content.

[Jeff Jarvis]

And if they were like a big old media company, their response would be to complain and sue and get publicity. But instead Wikipedia did what Leo, they.

[Leo Laporte]

Went to Kaggle, which is a very well known data scientist site. Kaggle's used data scientists go there to test their abilities and so forth. And they have many, many data sets, 461,000 freely accessible data sets that AIs among other things can use, researchers use, students use to train, to learn.

[Jeff Jarvis]

So don't scrape our site, just go there and download it.

[Leo Laporte]

Yeah, Kaggle is hosting.

[Jeff Jarvis]

That's what everybody should do.

[Leo Laporte]

A beta release of Wikipedia's entire data in both French and English. Go there, get it from kag, Kaggle. Kaggle can support it. It's formatted for machine learning. Great idea.

[Jeff Jarvis]

Brilliant, isn't it? I love that.

[Leo Laporte]

Yeah. And it looks like Google helped them because this is from the Google keyword blog. So I'm guessing Google's providing some bandwidth or something like that. You know, this is how these AI companies have got to respond. They've got to be, you know, caring participants in this because you can. You're going to end up killing all the data resources that you need to make better AIs. So. So I think that's good. I'm not so happy about the administration's plans to cram AI into K through 12 schools.

[Paris Martineau]

Yeah, sorry, don't you mean A1?

[Jeff Jarvis]

A1?

[Leo Laporte]

Yes, I mean A1. It's under consideration. And, you know, we haven't seen the.

[Jeff Jarvis]

It was called pre. Decisional.

[Leo Laporte]

Pre decisional. In other words, we haven't told Donald about it, and we're hoping he's going to go for. For it and take the Sharpie to it. Under the draft executive order, federal agencies would be instructed to take steps to train students in using AI and incorporate it into teaching related tasks. I'm not against this, but this is exactly what you had a department of education for.

[Jeff Jarvis]

Yeah, exactly. And you don't do it with a Sharpie saying, oh, we poured the sauce over and now it's in the classroom. This is. This requires thought.

[Leo Laporte]

And you need experts. Yeah, you need experts. Experts in this field to help you do this instead of just saying, well, well, David Sacks told us it was a good idea, so we're gonna do it right.

[Jeff Jarvis]

And my district here, you know, well, to do district is. Is firing 30 people because of the budget problems. And you know what's going to happen. They say, well, you got AI now. It's okay. You just use the AI. We'll give you another 10 students. You got the AI. It'll be okay.

[Leo Laporte]

Yeah, I mean, it's probably, you know, a. I've seen some evidence that AI can make a pretty good tutor, that kids can learn from AI tutors. But I think it has to be done intelligently, carefully, and slowly. I don't think you need to rush it in or jam it in. And I have a feeling that's what this will end up doing. So.

[Jeff Jarvis]

Yep.

[Leo Laporte]

Well, welcome to the new world.

[Jeff Jarvis]

So, Leo, I have an idea for a new service that you could provide.

[Leo Laporte]

Yes.

[Jeff Jarvis]

Your hosts and perhaps your guests and others.

[Leo Laporte]

Yes.

[Jeff Jarvis]

Is perhaps you could be a verification.

[Leo Laporte]

So I saw this thing. So this is on the skeet, on the Blue Skee. I want to just Call it the skeet. Can I just call it the skeet? Blue sky is adding blue check marks. But they're doing it, I think, in the right way. They're not. They're not just kind of give us $8. You got a check mark. In fact, I still don't have a check mark.

[Jeff Jarvis]

You never got one?

[Leo Laporte]

One on Blue Sky?

[Jeff Jarvis]

No, on Twitter.

[Leo Laporte]

Oh, oh, no, I've had one on Twitter. I thought, okay, what Cory Doctorow calls a non consensual blue check. They just gave it to me, actually. It's been good because it gives me full access to Grok. So I can play with Grok a little bit. No, I'm saying on Blue Sky, I don't. The thing about Blue sky that I. I think it doesn't really need this because I self verify, which they say.

[Jeff Jarvis]

They make a point of that, saying that's their primary means of verification.

[Leo Laporte]

Okay, good.

[Jeff Jarvis]

Is that you use your domain, which you long ago argued was the benefit.

[Leo Laporte]

A lot of people can't do it. I understand. I mean it takes a little DNS, foo and so forth. Plus you have to own a domain. I have a domain I wasn't using called LeoLAport Me. So when you go to My Blue sky, you'll see it's at leolaport Me. And. And I guess the theory is that people might want to, you know, actually go to that website.

[Jeff Jarvis]

When you, when you added that, when you changed to that, did you lose anything?

[Leo Laporte]

No.

[Jeff Jarvis]

No. Okay.

[Leo Laporte]

No.

[Jeff Jarvis]

I've just got to do it.

[Leo Laporte]

And then what it will take you to is what I consider a proof of identity. First of all, that it's there, it's hosted on fast mail and it's signed by my PGP private or. Yeah, private key which you can validate with my public key which is publicly available. So I, I mean I consider that a pretty rock solid proof. Plus it then links, plus your outpace page, which then tells you more about me and has a QR code you can scan if that points back to Blue sky and so forth.

[Paris Martineau]

So Jeff, the thing that I think is particularly useful for you, so when I set up mine, Mine, my bluesky is Paris nyc. This was before they'd rolled out this particular feature. But now if you set up a custom URL, it will save your current Blue sky handle, Jeff Jarvis bluesky Social so that no one could do that. Like I have to have a. I have a second Blue sky account that's just my old handle that nobody could camp that. But because it's only because I was an early member and set up a custom custom URL in like 2023.

[Leo Laporte]

But you have to know how to do that. And I understand that 90% of users of this kind of thing aren't going to know how to do that. Right.

[Paris Martineau]

That's fine.

[Leo Laporte]

So. So what is the mechanism now? You're saying that I, as an organization, twit.

[Jeff Jarvis]

As an organization could they will approve some approval verse? Yes.

[Leo Laporte]

Oh, well, some I'm doubt I'm gonna get in this. No, I think.

[Jeff Jarvis]

I think it'd be perfect.

[Leo Laporte]

I think it'd be Washington Post might get in the sum.

[Jeff Jarvis]

No, I think Twitch and his whole network.

[Leo Laporte]

Well, I'd be glad. So what do I have to apply to?

[Jeff Jarvis]

Yeah, I think there's something like that there.

[Leo Laporte]

Okay, well, I'll look into it. Yeah, I would. I mean, so then there would be some. Patrick would probably have to do it. Right? Sorry, Patrick could be some sort of method that we would.

[Paris Martineau]

Well, then you could. I guess as part of the whole Tesla scam, you could be like. You could charge people for a.

[Leo Laporte]

So actually, I realized the thing is that you sell the sledgehammers, you sell the ability to hit the thing, and then if you want to pay a little extra, you can get a blue check.

[Paris Martineau]

That's great. Yeah, yeah. Make sure. Let's scrub this part of the video too.

[Leo Laporte]

So. Yes. We don't want to get in trouble.

[Paris Martineau]

We can't have Blue sky hearing about this.

[Leo Laporte]

Anybody?

[Jeff Jarvis]

I didn't know the 4chan was dead.

[Paris Martineau]

Yeah, that's actually crazy.

[Leo Laporte]

Yeah, it's dead, but there are plenty of other replacements.

[Paris Martineau]

There was a time in my life where I'd have to check 4chan every night to see what the people were up to. That's how I discovered Qanon.

[Leo Laporte]

I'm sorry, was that as in your role as a reporter or.

[Paris Martineau]

That was. That was when I was at New York magazine.

[Leo Laporte]

Wow, what a horrible assignment.

[Paris Martineau]

I mean, it wasn't an assignment. I chose to do it because I was like, people are going to want to read about what the freaks on 4chan are doing. I did one good article about how people on 4chan were consuming raw onions because they thought it made them more manly, increase their testosterone.

[Leo Laporte]

Well, it guarantees they're not going to get a date.

[Paris Martineau]

That's. That's the way to do it.

[Jeff Jarvis]

Well, that's maybe their goal too, actually.

[Paris Martineau]

But now I'm Excuse even do that.

[Leo Laporte]

Oh, they hate me because I eat onions. Wow. Okay, okay, okay. What was I. I had some thought about Blue Sky. And, oh, this is my biggest kind of thing that I would love to see solved. There isn't really anywhere a centralized way to say I am who I say say I am. There used to be a wonderful site called keybase. Keybase.org it's still around, but they were bought by Zoom, and I'm not sure it really does what it used to do.

[Paris Martineau]

Oh, is that why Keybase stopped working for me? That would make sense.

[Leo Laporte]

We need something like that. Don't you agree? So what I did with Keybase is you would go there and isn't this the orb?

[Paris Martineau]

Isn't this what the orb is?

[Leo Laporte]

I guess it is what the orb is. I mean, he's trying to solve that problem and it does. I don't know if it should be with Biometrics, though. The idea of Keybase was you could use DNS to prove you are. Because you'd have a website that you controlled, maybe your Paris NYC site, and you could use DNS to post a little text string in there that would say, see, I am Paris, I control this site. And then Keybase would say, yep, that's her site. And then you could continue to do that. You may have seen in the back in the day, people posting tweets or messages on Facebook that said, this is my Keybase verification. And there's a long string. And the fact that you could post that under your name would then associate it with your Keybase account. We need some way that's easier than that of people saying, yeah, I'm Leo and these are my sites and you can trust that. And then we wouldn't have to have blue skies check marks or Xbox boxes. X.com check. Check marks. Right. Do we. Do you think we need something like that or is that.

[Paris Martineau]

I think the use of a check mark is just. It gives people, casual users of a social network, a easy.

[Leo Laporte]

Right.

[Paris Martineau]

Like easily look. It's something you just look at at a glance and you're like, yeah, this person has some credibility. And I do think that that serves a purpose. I think that it has. I'm not sure that checkmarks are the way to do that anymore. I'm un. Well, I'm genuinely unsure whether it's useful or not useful because the idea of the check mark has been so diluted over the last couple of years. And the fact that on products, but also on X, you can just pay for.

[Leo Laporte]

Ruined it.

[Paris Martineau]

Yeah, yeah.

[Leo Laporte]

And we knew there was a time where it meant anything. Yeah. Well, I think this would be a very good blue sky. Could even be that validator, Right. You see the blue check. And then in the sidebar, here's all the other sites that belong to me. And Blue sky has made the effort to validate that. That's all real. Mastodon does that with the, I think a fairly effective mechanism, but it's not easy to use. And that's the real problem. We need an easier way to do that. The problem also is that people would then say, well, that's kind of a national identity database. Do you really want the government to have something like that? Which I, I don't, actually. So I've just learned, by the way, I'm sorry, a little distracted, that Henry is going to a chef's dinner with Martha Stewart tonight. I don't know if it's a date. I hope it's not.

[Jeff Jarvis]

Well, you know, the next thing that my wife did after we left Entertainment Weekly because I walked out in the huff and she kind of had to go too, is she took the Mac network and set it up for Martha's Stuart Living Nice. She got a jar of jam.

[Leo Laporte]

It's a good thing. Did she really get a jar of jam? Yeah.

[Jeff Jarvis]

Oh, yeah.

[Leo Laporte]

I hope they paid her.

[Jeff Jarvis]

Yeah, they did okay. I think it might have been strawberry.

[Paris Martineau]

Good jam flavor.

[Leo Laporte]

Did Martha make it?

[Jeff Jarvis]

Oh, I'm sure.

[Leo Laporte]

I watch her things and I wonder, did she really get down on her knees in the garden, or did she do that for the camera, then get up and get a guy named Jesuit to do it? I feel like it's the latter, I'm just saying. But I don't know.

[Paris Martineau]

Probably.

[Leo Laporte]

Yeah, she loves to Garden. Yeah. 4chan is dead. Americans. This is depressing. This adds to your, your Financial Times article. Americans averaged over $700 in TikTok purchases in the last year.

[Jeff Jarvis]

How much did you, you alone added that up to?

[Leo Laporte]

I think I, I, I easily.

[Paris Martineau]

How much do you think? Can you show us? Go through Tik Tok Shop. What have you bought from there? I've never purchased anything on Tik Tok.

[Leo Laporte]

So I think I got the Plaud note there. That was the AI recording.

[Jeff Jarvis]

I, I, I bought momofuku noodles, which was a mistake.

[Leo Laporte]

I bought Momofuku noodles. But see, the problem is I don't know if it was their Tik Tok Shop.

[Jeff Jarvis]

That's the only thing I bought.

[Leo Laporte]

But I don't know if it was Instagram or Tik Tok. Wow.

[Paris Martineau]

Maybe I'm a boomer.

[Harper Reed]

I don't, don't.

[Leo Laporte]

You're the boomer.

[Paris Martineau]

I'm the boomer. Now I'm the one being like, it seems sketchy to order stuff off Tik Tok. Whenever I accidentally swipe over to that Tik Tok shop thing, I, like, throw my phone away. I'm like, oh, it's just such an aggressive part of the app to be on.

[Leo Laporte]

It's way too much.

[Harper Reed]

I.

[Leo Laporte]

And I got rid of my Tik Tok. By the way, I don't do Tik Tok anymore either.

[Paris Martineau]

Whoa.

[Leo Laporte]

Yeah.

[Jeff Jarvis]

Oh, man. Cur.

[Leo Laporte]

So I, I've been watching a, I thought a very interesting documentary on Netflix called Buy now, which came out last year, but I didn't see it. It is very depressing. It is. It's people who worked at Amazon in the UX design, people who worked at Adidas. In fact, the chief marketer Adidas confessing all the things they did to get you to buy more crap. And really, they're saying, it's not your fault that there's something like 18 billion running shoes made every year. It's not your fault that 13 million phones are thrown out every day. 13 million phones are thrown out every day. We have created this monster of E commerce and made it so easy to buy, but nobody considers what's at the other end of all these purchases. It is a little bit depressing, but it's maybe going to help me kind of cut back on my per. Oh, wait a minute. My new mixer just came. Hold on. I got, I don't know, I shouldn't. $700 in Tik Toop shop purchases.

[Paris Martineau]

That's crazy.

[Leo Laporte]

So how does nearly half of them.

[Harper Reed]

How does that mechanism. Is your credit card saved with them or do you have to pull that out every time or what?

[Leo Laporte]

Oh, I don't know. I don't remember how it works. I think they know who you are.

[Jeff Jarvis]

Well, I just, When I start filling one in, Google fills it in for me.

[Leo Laporte]

Ah. Nearly half of the purchases on TikTok shops are due to influencers, of course, because that's what you see. You see a video that's our new economy guy saying, hey, I can't believe this thing. I just found it and it's amazing. It's transformed my life. And I, I, I don't know how I live with that. Oh, I did buy gum on TikTok.

[Paris Martineau]

What?

[Leo Laporte]

That's where I got that gum.

[Jeff Jarvis]

Oh, here's solar lights for your patio. You can put in. That's nice.

[Paris Martineau]

That's the most dad thing I've ever seen. You're getting a tick tock? No, him buying solar lights for his patio.

[Jeff Jarvis]

I'm not buying them.

[Leo Laporte]

I'm just saying that's 37% of Americans under 60. Which means you, Paris, have purchased something on TikTok shop. Averaging $59 a purchase. Only one in three Americans under the age of 60. But those one in three are working hard. They average 12 purchases in the last year. That's amazing.

[Harper Reed]

Yeah, I feel like this is more of a Wales problem than an average American.

[Leo Laporte]

It's whales. The biggest spenders. Yeah.

[Jeff Jarvis]

The average QVC for the new age. Yeah.

[Leo Laporte]

Most popular purchases. This is from partner Centric affiliate marketing resources. 46% personal accessories. 43% patio solar lights. 40% fashion items.

[Harper Reed]

Patio solar lights. That's old people for sure.

[Paris Martineau]

No, that was. That was a joke.

[Leo Laporte]

No, well, household items.

[Harper Reed]

No, no, absolutely. My mom buys those all day.

[Leo Laporte]

Yeah. Makeup, cosmetics. 33, 27. Technology. Now this is interesting. Only less than a quarter found the item by browsing the TikTok shop. Most of the time I think they saw a video as they were scrolling through their TikTok and they went, oh, I gotta buy that. Why do consumers use TikTok shop? 50% say easy to purchase. 50% say low price. 39% say the item was on sale. They're always on sale sale, by the way. Nearly one in four say they make impulse purchases. Top regrets, 17%. The item was cheap quality. Okay. There was this wine cork remover. This guy was really. He made it look good. He made it look really good.

[Paris Martineau]

How many wine cork removers do you own, Leo?

[Leo Laporte]

A few. Most of them are standard. You know, they're ahsos or the corkscrews, that kind of thing. But this guy, he was Australian. He came up, you could find a on tick tock. And he comes up, he's very suave, comes up at the table, says, let me open that for you. Goes. It's like pumping it up. Thing comes out, there you go. And he poison one. It's just that easy. So I bought it. It was the crappiest. I thought I was going to give it to Lisa for Christmas. I thought, that'll be a great Christmas gift. I couldn't. I threw it out.

[Jeff Jarvis]

Have you bought any Bell and Howell laser lights?

[Leo Laporte]

Of course they've licensed the name Bell and Howell because there is no Bell and How anymore.

[Paris Martineau]

Who's that? What's that?

[Leo Laporte]

At least 20. Yes, it was an old.

[Jeff Jarvis]

It was an old camera brand.

[Leo Laporte]

You ever heard of Packard Bell Paris?

[Jeff Jarvis]

Long defunct and it's a resurrected brand. And then they put this crap out on TV. It's mainly TV.

[Leo Laporte]

23% regret at least one purchase from the TikTok shop.

[Harper Reed]

Wow.

[Leo Laporte]

Really? That was bad. That was bad. 63% of Americans take advice or suggestions from influence influencers.

[Harper Reed]

And there's your problem.

[Leo Laporte]

Yeah, that's a.

[Paris Martineau]

Only take suggestions or advice from Leo.

[Leo Laporte]

Yes. I'm not an influencer. I just play one on tv.

[Jeff Jarvis]

Where.

[Leo Laporte]

Who did I, who did I. I was talking to somebody who said. Did you see this study that said something like 57 of teenagers want to be an influencer when they grow up.

[Paris Martineau]

I'm shocked that isn't higher.

[Leo Laporte]

There's no other jobs.

[Jeff Jarvis]

That's all you got?

[Harper Reed]

No, it's just the new I want to be famous.

[Jeff Jarvis]

Yeah, yeah, yeah it is. But it is univers and rich at the same time.

[Leo Laporte]

Rich and famous. Yeah.

[Paris Martineau]

And get a bunch of free stuff.

[Harper Reed]

No, they assume rich comes with famous.

[Jeff Jarvis]

That's the problem.

[Harper Reed]

They assume rich comes with famous.

[Jeff Jarvis]

Yeah, you're right.

[Leo Laporte]

This is. I don't know if this is valuable. It was research done by the influence index. The influence index. This is from Australia, presented by the odds. I think it's Australia. Mysterious news.com half of all Australian Gen Z girls would quit school or their job to become an influencer if they could. 73%. Yeah. Who wouldn't?

[Paris Martineau]

Would you quit your job if you. I tell you free stuff and get paid for making little short term.

[Leo Laporte]

My son who is absolutely in that category of influencer. Right. You would if he's having dinner with Martha Stewart. I think so. It's killing him. It's hard damn work.

[Paris Martineau]

Well, I mean it's hard because you can't really take a break. Like you don't.

[Leo Laporte]

Right. People don't understand that. They think oh it's fun, you're just making little videos, you get to try and close. Lot of work. Also there's the parents who are ripping off their influencer kids. Kids, which is another problem.

[Jeff Jarvis]

Yeah. Well it's like child stars in the day.

[Leo Laporte]

Yeah.

[Jeff Jarvis]

There's California's new law.

[Paris Martineau]

Interesting laws in states, a couple in the Midwest actually if I'm recalling correctly, that require that parents don't just spend all of the money their kids earn that if their kid is being an influencer they have to have control of at least some of the money.

[Leo Laporte]

California as of January 1st past 8, child influence influencer law. Content creators who feature minors in at least 30% of their content must deposit 65% of the miners gross earnings into a trust account. But this was a law, the Jackie Coogan law. Remember back in the day. Remember Jackie Coogan, Jeff? He was a hoot. He was a child star. You might not remember him in Paris because it was in the 30s.

[Paris Martineau]

I was about to say that you could be taller. Talking about brontosaurus. For all I know.

[Jeff Jarvis]

She'S calling us dinosaurs.

[Leo Laporte]

Jackie Coogan. But he was. This is in the 60s.

[Paris Martineau]

Pretty old child.

[Leo Laporte]

Yeah. No, he was married to Betty Grable, believe it or not. You know who Betty Grable is? No.

[Jeff Jarvis]

Or she doesn't.

[Leo Laporte]

He was in the Charlie Chaplin movie the Kid, which was. Made him one of the first child stars in history.

[Paris Martineau]

Every once in a while, when you guys say one of these names, I'm like, they're talking about some Charlie Chaplin ass guy. And this time it is actually true.

[Leo Laporte]

Look at him. Look at him. Isn't he Dorbs? Look at him. Who wouldn't fall in love with that Put him in trouble. But that's in 1921. He got ripped off by his folks. California enacted the first legal protection for the earnings of child performers, widely known as the Coogan Act. California's child actors. Bill. Be fun to drop that in you. I know. You know Jackie Kan. You know why? He was Uncle Fester in the Adams family.

[Paris Martineau]

Which one was that again?

[Leo Laporte]

The old Adam.

[Jeff Jarvis]

Adam's family.

[Paris Martineau]

It's. Yeah, they kind of blend together to me, I mean, I don't.

[Harper Reed]

I'm pretty sure Paris's reference to that, though, is the modern Adams family.

[Leo Laporte]

I know.

[Paris Martineau]

No, I. I've watched the old Adams family. I just don't really recall Uncle Fester.

[Leo Laporte]

Uncle Fester is a guy like a monk's outfit to me.

[Harper Reed]

It's for Lloyd's Uncle Fester.

[Leo Laporte]

Oh, yeah. That's not the same. Coogan as a kid earned $3 million, which is equivalent to $68 million today. When he turned 21, his fortune believed to be well intact. But his assets were conservatively managed by his father, who died in a car accident five months earlier, who soon discovered nearly the entire amount had been squandered by his mother and stepfather, Arthur Bernstein. On fur coats, diamonds, other jewelry, and expensive cars.

[Jeff Jarvis]

Mom.

[Leo Laporte]

Mom. Mom said no promises were ever made to give Jackie anything. He was a bad boy. Seriously. She said that he sued them, but after legal expenses, he got just $126,000 thousand dollars. When he went broke. During the litigation, he asked Charlie Chaplin for assistance. Charlie gave him $1,000 in cash without hesitation.

[Paris Martineau]

I'm going to be honest. Uncle Fester looks exactly like Mark Andreessen.

[Leo Laporte]

He does, actually. Mark Andreessen is More of a bullet head, even.

[Paris Martineau]

But, I mean. Yeah, but yes.

[Leo Laporte]

Anyway, California does have a successor to the Coogan Law, the Coogan act, which is for child influencers. Which it should. It should. Washington Post has given all of its content to open AI. I thought it had done that a while ago.

[Jeff Jarvis]

This is a whole new deal where they're going to get promoted there. I don't know what they. I don't know who paid what for Powell on this.

[Leo Laporte]

The Post's essential journalism is now featured in ChatGPT. Search responses. By the way, it is the Washington Post that called its journalism essential. This is in the article from the Washington Post to which I no longer subscribe.

[Harper Reed]

Yeah.

[Leo Laporte]

Note the big resubscribe button on the top of my page.

[Jeff Jarvis]

Have you seen the sink twins?

[Leo Laporte]

No. Who are they when they're at home?

[Jeff Jarvis]

I hope we get line 117. I don't know. Can we play this or not?

[Leo Laporte]

I'll be the judge of that. I have a strong instinct for content id.

[Paris Martineau]

Yes.

[Leo Laporte]

Yes.

[Paris Martineau]

We never start play something and then jumps in and it's like, stop, stop, stop, stop.

[Leo Laporte]

Okay. You're the leaker. It's called.

[Harper Reed]

Yeah, no, it's Kimmel.

[Leo Laporte]

Can't play it. Nope. It's Jimmy Kimmel. He's notoriously litigious.

[Jeff Jarvis]

So there's these two. There's these twins. I would say two twins. There's these. These twins who. And they just say. They talk simultaneously. Maybe you can find a news clip. Can you find a news clip with that?

[Leo Laporte]

So Pete Heth slams leakers after text fallout. Oh, no, that was a separate story. Yeah. So these. These two. These two. Oh, they're adults.

[Jeff Jarvis]

Yes, yes, yes.

[Leo Laporte]

And they just say the same thing all the same time.

[Jeff Jarvis]

And it's not an act.

[Leo Laporte]

It's. It's.

[Jeff Jarvis]

That's them. They. They. There was a.

[Leo Laporte]

And you complain about AI filling the Internet with slop.

[Jeff Jarvis]

I know. This is amazing. It is absolutely amazing. I. I watched it before Kimmel and I said, oh, my God, what's happening here? And that was Kimmel's reaction as well.

[Leo Laporte]

If you.

[Jeff Jarvis]

If you search on the. What are they called? The pelican twins. Search on. They're known in Australia. They.

[Leo Laporte]

They.

[Jeff Jarvis]

They rescue pelicans. Pelicans.

[Leo Laporte]

Oh, okay. All right.

[Jeff Jarvis]

Daily Mail. Can you play Daily Mail?

[Leo Laporte]

I could probably play the Daily Mail. Let's see.

[Jeff Jarvis]

So just pelican twins.

[Leo Laporte]

Here they are. Well, it just happens. And we don't know why, but we have tried to talk separately. Oh, this is ourselves. It's not us.

[Harper Reed]

And it's very hard.

[Paris Martineau]

They're barely speaking.

[Leo Laporte]

I could do that, Paris. Go ahead. Let's you and me do that. That Paris and I can talk. Say the same thing at same time. Especially if we pause carefully, and we've kind of learned how we talk. Can we get on camel with this, do you think?

[Paris Martineau]

I think we could. I think if we practice. Okay. You know what could get so phony?

[Leo Laporte]

I can't believe how phony that is, Jeff. And you thought that was cool.

[Jeff Jarvis]

It's. I'm telling you. Wait.

[Leo Laporte]

And he was coming back down towards it.

[Jeff Jarvis]

There you go.

[Paris Martineau]

And he goes.

[Harper Reed]

Run.

[Leo Laporte]

He's got a gun.

[Jeff Jarvis]

All I was thinking about when. When we're.

[Leo Laporte]

No, it's so obvious. They're taking cues from each other, Jeff.

[Jeff Jarvis]

But. But that. That's what. That. That's what they do, is they live.

[Leo Laporte]

Yeah, because they got nothing else.

[Paris Martineau]

Wait. Pelicans.

[Jeff Jarvis]

They got pelicans.

[Paris Martineau]

I have to interrupt this. Darren Oakley in our Discord made Flappy Jeff.

[Leo Laporte]

Yay. I figured something. Oh, Darren is very accomplished, I must say.

[Paris Martineau]

Flappy-jeff.ro fleet app. No, it's in.

[Leo Laporte]

Oh, I gotta play it.

[Paris Martineau]

You can actually play it in there. Wait, no, no, no.

[Leo Laporte]

You're. You're.

[Paris Martineau]

You're watching a video Leo, I believe.

[Leo Laporte]

How does the computer work? Wow, Darren did that. He's very adept with this stuff. Darren, where's the game, Darren? Is there a way I can play it?

[Paris Martineau]

It's in the link. If you look up in the discord.

[Leo Laporte]

Wow.

[Paris Martineau]

I'm not good at it.

[Leo Laporte]

Oh, there's a download.

[Harper Reed]

Yeah, no, that's the video. You got to keep going up.

[Paris Martineau]

No, it's further up. Hold on.

[Leo Laporte]

Oh, it's up above. Okay, here, I'll tag you.

[Paris Martineau]

I just tagged you. Click the link that I just did, Grandpa.

[Leo Laporte]

What? What did you do? What? Flappy Jeff. Okay, Navigate through stacks of books. Now, how did you do this, Darren? Did you do it in Python? What did you. Whoops. Game over, man. Game over.

[Paris Martineau]

See, that's what was happening to my flappy Jeff, you've got a. Oh, it's hard.

[Leo Laporte]

But this was always hard. Did you play the original?

[Paris Martineau]

Yeah, I did. It's hard, not easy. He said pure vibe coding. No actual coding. Oh, I really like this. This is fun. I really like the way he's got.

[Leo Laporte]

Jeff's head and everything.

[Paris Martineau]

Yeah, that's great. You'll forever be remembered, Jeff.

[Jeff Jarvis]

Yeah, I'm very, very proud. I'm very proud.

[Paris Martineau]

Oh, three. All right, I want to see who can get the highest Score in Flappy Jeff by next week, post it in the discord or add us on Darren.

[Leo Laporte]

I got. I gotta know more. So you. You did it? What did you do? Tell us all the. Tell us all the details. He says there was no actual coding. It was vibe coding. So it was just a prompt. Did the prompt. Who was the prompt to? Was it a Claude or. Darren uses local LLM as well. He did. He used Replit to do it. Made the image in chat GPT so that even the images were generated. Oh, wow. Out of sync. Got 72 seconds, a high score of 13 points. Nice job. Out of sync. And Replit hosts it as well. Yeah, Replit used to be the place you go to learn how to code. It was REPL's for all the different code languages. But they saw the light early on and said, you know, we should just make this a place you can do AI coding. And so, in fact, the CEO of Replit has very famously said, there's no, don't study coding. There's not going to be any more jobs in coding. Which is a shame because Replit was a great place to learn how to code. I had a high school mentee I set up at Replit to learn Python at the time, but because you didn't have to set it up on your own computer, you could do it in the web and so forth.

[Paris Martineau]

I'm on a good run right now. I'm on 16. I'm keep. I'm going, I'm going. I'm going completely.

[Leo Laporte]

Well, let's take a break while Paris masters flappy jeff.

[Paris Martineau]

19. I'm at 19.

[Leo Laporte]

You were getting out of sync.

[Harper Reed]

I know.

[Jeff Jarvis]

High, high score.

[Leo Laporte]

Let's take a word from our sponsor. More to come.

[Paris Martineau]

Our sponsor, Flappy Jeff, brought to you by Flappy Jeff.

[Leo Laporte]

No, it's actually brought to you by a perfectly appropriate sponsor for this particular segment. Our friends at OutSystems. What is that? It is the leading AI powered application and agent development platform. Now, they've been doing this for a long time, for more than 20 years. The mission of OutSystems is to give every company the power to innovate through software. You know, it's always that build or buy decision. In fact, we. This is where we made a mistake. I was talking, talking about that sales system that we had an intern, right. We probably shouldn't have built it ourselves. It's kind of janky. But on the other hand, we didn't want to buy it. They're very expensive and they wouldn't do exactly what they want. In the past that was your choice. Buy off the shelf SaaS products, you'll get them up and running, but you lose flexibility and differentiation and frankly they cost a bundle. Or build custom software, but that's going to cost even more in time and resource sources. But AI has changed everything and thanks to OutSystems, there's a third way. The fusion of AI, low code and devsecops automation into one development platform. And as I said, they've been doing this for 20 years, you know, for a long time with low code or no code. But now with the addition of AI, you've really got something. Your teams are going to build custom applications just as easily as Darren just built that Flappy Jeff With AI agents it's as easy as buying generic off the shelf sameware. But it's flexible by the way in Outsystems is really good about this. Security and scalability are built in because they know you're building enterprise apps here. With AI powered low code teams can build custom future proof applications at the speed of buying. With fully automated architecture. You get security built in, you get integrations, data flows per permissions, the whole stack. It's all there for you. Frankly, Outsystems is the last platform you need to buy because you can use it to build anything and customize and extend your core systems. Frankly, we should have gone with Outsystems. You build your future with Outsystems right now. Visit outsystems.com TWIT to learn more. That's outsystems.com TWIT build custom apps and agents faster. This is brilliant. Finally a third answer to that build or buy question. Outsystems.com TWIT we thank him so much for supporting intelligent machines. It's an exciting time. I think we're living an exciting time. How's your Flappy Jeff?

[Paris Martineau]

I've really devolved. I had like three different runs where I got zero points.

[Leo Laporte]

Yeah, flappy was always really hard.

[Paris Martineau]

But remember it really, it depends on how the level ends up being generated because sometimes it's very hard and sometimes.

[Leo Laporte]

It'S ah, it's a randomly generated.

[Jeff Jarvis]

I'm having a senior moment. What was the Angry Birds, right?

[Leo Laporte]

Yeah, this is not that Flappy Bird was created by a Vietnamese kid some years ago and it was a massive hit on iOS. So much so. And he made so much and he didn't charge for it, it was free, but he had ads in it. He was making so much money he freaked out and he pulled it like weeks in he said, I can't take it.

[Jeff Jarvis]

I'm a communist. I can't do this.

[Leo Laporte]

It just, it bugged him that people were, you know, so addicted to it. It's really great. It's actually a really great story. And now of course, AI can easily build it because it didn't take long before a lot of people created Flappy Bird clones and so forth. Once he pulled it off the market, there was, you know, people were free to, to do that. There's a good article from Wired. The rise and fall. Get it? Of Flappy Bird.

[Paris Martineau]

And rise again.

[Leo Laporte]

Yeah, and rise again. Dong Nguyen uploaded a game in May. This was back in 2014, uploaded a new game to the iOS app store. One of hundreds new apps added to this app store every day. Win had created a simple game which the player controls a funny looking bird by tapping the screen. It needed simple name. He called it Flap Flap. Then he realized another app had the name, so it's so easy to rename it. He did it on the app Store and called it Flappy Bird. Perhaps you've heard of it. Wins game. He. Millions of people were downloading Flappy Bird. At its peak, Win was taking in $50,000 a day from the ads. But he was also under, and this is I think maybe what really happened. A constant barrage, according to Wired, of messages, insults, requests for interviews, even death threats. When decided last weekend, after just a few weeks, it wasn't for him. He stopped talking to anybody, pulled Flappy Bird from the App Store and that was that.

[Jeff Jarvis]

Wonder where he is today?

[Leo Laporte]

That's a great question. That's a really interesting question.

[Harper Reed]

Last we heard from him, he was railing against the crypto company that bought.

[Jeff Jarvis]

That Flappy Bird IP or something, Right?

[Paris Martineau]

Yeah.

[Leo Laporte]

Poor kid.

[Paris Martineau]

The current champ of Flappy Jeff is P. Hall on Discord.

[Leo Laporte]

Of course.

[Paris Martineau]

Paul Points.

[Leo Laporte]

I'm not surprised. Paul Holder, he's a. He's a genius.

[Jeff Jarvis]

How far?

[Paris Martineau]

33 points.

[Jeff Jarvis]

Paul, get a life.

[Leo Laporte]

We have, we have really recreated Flappy Bird. There are of course many places you can Flappy Bird online like Flappy Bird. Gg. This is a pretty close clone of the original. Just tap the screen. Oops.

[Paris Martineau]

Yeah, but it doesn't have a.

[Leo Laporte]

It doesn't have a jet.

[Paris Martineau]

A disgruntled professor.

[Leo Laporte]

Yeah, I love that. I think what a. That's that. There you go. In a nutshell, that's what's changing in the world. It's not twins pretending to talk.

[Jeff Jarvis]

How long did it take you to make it?

[Leo Laporte]

Well, it couldn't have been more than 15 minutes. He came up with that long.

[Jeff Jarvis]

Was it that time?

[Leo Laporte]

Darren's very good at this, by the way. Darren does partner coding, pair coding with AI he does a lot of vibe coding. That's what he does. That's his. He's. He's been selling that, you know, he's one of the guys who does the advent of code with us. He's very. He's a really proficient coder. He and P. Hall Paul are our two, I think, our two top coders in the club. Wow. But Darren has gone all in on AI he does it professionally. Does it for work. Where do you work, Darren? Can you say? Is that a secret? He's mentioned it before. He does a very complicated kind of coding, as I remember.

[Paris Martineau]

No, I had assumed it was, you know, kind of flappy, bird or bust.

[Leo Laporte]

But I don't think he writes games.

[Paris Martineau]

He says a man of leisure.

[Leo Laporte]

He's out of work. Well, I think he works contract. He's, by the way, I'm pretty convinced a 10x. One of those 10x coders that they talk about. You know, these guys who are so good, they bring him in to solve intractable problems and get paid a lot of money. I'm pretty sure you're a 10x coder, Darren. Pretty sure. He's down under. He's from Australia.

[Jeff Jarvis]

Oh, what time is it down there?

[Leo Laporte]

Oh, it's a normal time. I know that because Windows Weekly was hosted by Richard Campbell and Sydney.

[Jeff Jarvis]

Wow.

[Leo Laporte]

And he joined us at 4am, which was 11am our time. So it's a. It's a normal midday now.

[Paris Martineau]

BY now it's 9:50am Yeah.

[Jeff Jarvis]

I did twig from Sydney once.

[Leo Laporte]

Yes, I remember.

[Jeff Jarvis]

I had to. The WI fi. The. The Internet was horrible. I had to buy a new phone to try to do it. I walked along the waterfront. I tortured you, as usual.

[Paris Martineau]

You did the show while walking along the waterfront because the hotel WI fi.

[Jeff Jarvis]

Was terrible and the cell connection was bad. I bought a separate cell thing.

[Leo Laporte]

And he.

[Paris Martineau]

I say, famously, isn't recording audio near bodies of water like the worst place to record it?

[Leo Laporte]

Pretty much. Hey, good news. According to the BBC, films that are made with AI can win Oscars. It's. It's official now. Boo. The Academy of Motion Picture Arts and Sciences issued new rules of on Monday which said the use of AI and other digital tools would, quote, neither help nor harm the chances of achieving a nomination. They can't speak for the voters, though. Remember Adrien Brody's fantastic film the Brutalist? I think lost the best Picture Oscar because it came out that they had used AI to help them with the Hungarian accents. I mean, the most minor use of AI you can imagine. But that. But see, the industry doesn't like it. They're terrified.

[Jeff Jarvis]

I thought it was a weird movie.

[Leo Laporte]

Oh, I loved it. Did you not like it?

[Jeff Jarvis]

No, I watched.

[Leo Laporte]

Did you think Anora was the best picture of the year?

[Jeff Jarvis]

Didn't watch that either. I don't.

[Leo Laporte]

Terrible anymore. Not terrible. It was. It was a fine movie, but not the best picture of the year. It wasn't terrible. It was. All right. All right. I think we should break. And when we come back, guess who has a pick of the week for. For us? You all do. Picks of the week. Coming up with Paris Martineau. Jeff Jarvis. I'm gonna take a little moment, just a quick moment to say I would love you to join Club Twit. I didn't realize this. We just crossed four years. Patrick Delahany told me the anniversary of Club Twit was. I know. I thought it was much younger than that. Four years ago, Lisa, my brilliant wife, had the brilliant notion of, you know, the audience is the most important part of this whole enterprise. We should let them join the club and support what we're doing so we are less dependent on advertisers. It was a lucky thing she did. This was during the height of COVID of course. So advertising was scant and it's gone up and down. And honestly, I think it's. I don't want to jinx us, but I have a feeling that there is a little bit of economic trouble ahead and it might in fact be a little troubling for us. But that's where you. You come in. Even today, advertising doesn't pay all of our costs. With the help of our Club Twit members, though, we can do the shows we do and bring in new shows all the time. When you pay $7 a month, that's all you get. Ad free versions of all the shows that we do. You get access to our Club Twit Discord and the events that go on in there. Like, as I mentioned, the AI User group this Friday, we do recordings of shows in there. Photo time with Chris Marquardt is coming up May 2, Micah's Crafting Corner May 16, and Stacy's book Club that afternoon. But I have to tell you, we have moved also our keynote coverage into the club. We got tired of the takedown notices from Apple, so we're just going to stream it in the club. If you're a member of the club. You can watch us as we cover the Microsoft Build Keynote. The Google I O Keynote Jeff, I hope you'll join me for that. Going to do a in between a special Gizwiz show. Paris, join me too if you would like that Google I O keynote is. Let's see, Tuesday, May 20, 10am we're going to also cover the Worldwide Developers Conference. In fact, now that we're doing it in the club, Mike and I are going to stick around for the whole day. We'll get not only the Apple keynote in the morning, we'll get the State of the Union keynote in the afternoon. So all of this is is, you know, by way of thanking you for supporting us. We want to make your $7 go a long way and I think that's a pretty good deal. $7 a month, $84 a year ad free versions of all our shows Access to our Club Twit Discord, where many of the club members hang. It's a great hangout. Special events that go on only in the club. I don't think any podcast network offers more for less. Please join the Club Twit TV Club Twit and thanks in advance. Imagine you're a business owner relying on.

[Harper Reed]

A dozen different software programs, each one.

[Leo Laporte]

Disconnected, more expensive and more complicated than the last.

[Harper Reed]

It can be incredibly stressful right now.

[Leo Laporte]

Picture Odoo Odoo brings all the tools.

[Harper Reed]

Your business needs into one simple CRM.

[Leo Laporte]

Accounting, inventory, manufacturing, marketing, HR and more. All seamlessly connected. Everything works together, giving you the peace.

[Harper Reed]

Of mind that your business is running smoothly from every angle. Odoo's open source applications are user friendly and designed to scale with your business.

[Leo Laporte]

Saving you time and money. Say goodbye to juggling multiple platforms and hello to efficient integrated management.

[Harper Reed]

Stop wasting resources on computer complicated systems.

[Leo Laporte]

And make the switch to odoo today.

[Harper Reed]

Visit odoo.com o d o o.com and discover how Odoo can simplify and streamline your business operations.

[Leo Laporte]

Odoo Modern Management Made simple like your.

[Harper Reed]

Favorite startups growth curve, T Mobile's coverage keeps scaling because T Mobile helps keep you connected from the heart of Portland to right where you are on America's largest 5G network switch. Now keep your phone and T Mobile will pay it off up to $800 per line via prepayment prepaid card. Visit your local T Mobile location or learn more@t mobile.com backslash keepandswitch up to 4 lines of your virtual prepaid card. Allow 15 days qualifying unlock device credit service report in 90 plus days. Device, knowledgeable carrier and timely redemption. Required card is no cash access and expires in six months.

[Leo Laporte]

Wow. Out of sync is now up to 98 on Flappy Jeff.

[Paris Martineau]

Whoa.

[Leo Laporte]

98.

[Paris Martineau]

98. That's the number to beat, people.

[Jeff Jarvis]

And by the way, Paris's week is going to be lost.

[Leo Laporte]

Another benefit of Club Twit, access to great games like Flappy Jeff. That's true, that's true, that's true. Paris pick of the week.

[Paris Martineau]

All right, my pick of the week is something I came across called the. I'm probably going to mispronounce this the well project, which aims to illustrate the rich ecological history that underwrites the development of New York City. And how you do this is you enter in an address. So, for instance, let's do five. Five, five. I'm not going to dox myself. I'm just gonna choose a random address, 555 Fulton street in New York, New York, and it takes you there and tells you what it looked like. Long ago, this used to be an oak tulip tree forest community, home to plants like starved panic grass, prairie fleabane and whitewood aster, and animals like the meadow vole and broad winged hawk. This block was the suitable habitat for over 199 species of plants.

[Leo Laporte]

And then the asters move in and that's all was left. Wow. And so you zoom in, so you find out this is what it looked like back in the priest. Now can I say later, like a hundred years later or.

[Paris Martineau]

I don't think so.

[Leo Laporte]

It's just ancient time. Here's what Grand Central looks like.

[Paris Martineau]

If you then go click in over on the left hand panel, you can figure out what the plants were. Animals were Princetons on that part of Fulton we were just talking about, it was likely used by the Lenape people in part because the block was 135ft from the nearest Lenape trail and a thousand feet from the nearest Lenape campsite, Yankee Stadium.

[Jeff Jarvis]

Wow.

[Paris Martineau]

The state go around all of it. Yeah.

[Leo Laporte]

It's kind of. Kind of depressing, actually. Battery park, they got some parts of.

[Paris Martineau]

New Jersey in there too. Jeff, if you're interested.

[Leo Laporte]

More woods. Oh, it was kind of pretty. Yeah. I like watching old movies about New York.

[Jeff Jarvis]

Right.

[Leo Laporte]

It's because of this. I mean, in New York, the history of New York is really quite fascinating.

[Paris Martineau]

It's just a very interesting tool to look into, like what used to be here. The other thing I wanted to show.

[Leo Laporte]

You, that's before Robert. Robert Moses, though, right? I mean, yeah.

[Jeff Jarvis]

Let me just Mention while we're on here one more second. Russell Shorthose new book. Have I. Have I plugged it? It's one.

[Leo Laporte]

Yes, you have because I, I mentioned that I, I loved the book he wrote called the island at the center of the World and now he has an updated Taking Manhattan. It's one Taking Manhattan. I will read it because I loved the Island.

[Paris Martineau]

What is it about?

[Jeff Jarvis]

It's about the Dutch time in New York and how the English took it over without any shots being fired. But how the Dutch culture prevailed in our diversity.

[Harper Reed]

This a larger seven nations of America about all the cultures that actually make up America. And there's like technically about seven cultures that make America.

[Leo Laporte]

Native American cultures.

[Harper Reed]

No, no, I'm talking about. No, no, no. Post, post Native American.

[Leo Laporte]

Oh, our cultures.

[Jeff Jarvis]

Yeah, yeah, yeah.

[Leo Laporte]

Ah, so like only seven. Yeah.

[Jeff Jarvis]

You know how like the Dutch are.

[Harper Reed]

Like in the, in New York, stuff like that. It's like how in California it's the settlers, you know, and things like that.

[Leo Laporte]

I, I was a while ago after finishing the Power Broker. How are you doing on that, Jeff, by the way?

[Jeff Jarvis]

I'm about 2/3 of the way through.

[Leo Laporte]

Very good.

[Jeff Jarvis]

It's riveting. It's riveting.

[Leo Laporte]

It is really well written. You would think a book that long and that detailed, you'd get bored.

[Jeff Jarvis]

And it's fascinating how obviously he knew what he thought of the guy by the end, but you just see velocity to it.

[Leo Laporte]

He develops it so beautifully.

[Jeff Jarvis]

He just becomes an such a. The section about his brother is just heartbreaking.

[Leo Laporte]

Yeah, there's a lot of heartbreak in that story. I've been looking for a good history of New York. There is one two volume set by Mike Wallace and Edwin Burroughs called Gotham and I think the follow up to it. What's the name?

[Paris Martineau]

Oh, I've heard Gotham is supposed to be pretty good.

[Leo Laporte]

Yeah, I kind of want to read it.

[Paris Martineau]

I had some friends that did a book club where the theme was obnoxiously long books and Gotham was in it.

[Leo Laporte]

It's definitely. So it's the history up to 1898 and then the sequel is Greater Gotham 1898-1919.

[Paris Martineau]

I do think that's a good naming custom for sequels is you just add greater to whatever the.

[Leo Laporte]

And then the greater Greater Gotham.

[Paris Martineau]

Greatest Gotham.

[Leo Laporte]

The greatest Great, Greater, greatest. I like it.

[Paris Martineau]

No, no books can have. Can be anything more than a trilogy.

[Leo Laporte]

But I, but I thought the Power Broker, the Robert Moses story is just so gripping and really does cover kind of most of the 19th or the 20th century. In New York because he really was so important from the 20s right up to the 70s. Taking a hat and I will read that. Thank you.

[Paris Martineau]

My last brief shout out for a pick because I don't think we can play the song is 18 years ago today. Tay's on day. Uploaded Chocolate Rain.

[Leo Laporte]

Chocolate Rain.

[Paris Martineau]

Some say dry and dry.

[Leo Laporte]

Another spiel. The pain. You were like 8 years old when that came out. You remember it?

[Paris Martineau]

Yeah, I was.

[Leo Laporte]

I could totally play it, but I remember it.

[Paris Martineau]

Oh, we can. Okay.

[Leo Laporte]

Oh, sure.

[Jeff Jarvis]

I don't know.

[Paris Martineau]

I guess we can Download the free MP3.

[Leo Laporte]

Creative comments. So there are a few reasons this caught on. One, he looks like he's eight and he has this big voice. The other one is he leans away.

[Paris Martineau]

From the mic to take a breath and then calls it out by doing on screen. I lean away from the mic to take a breath and that phrase has been stuck in my head for 18 years.

[Leo Laporte]

The prisons make you wonder where. He actually got a lot of mileage is still kind of around. Camera out and see the light. Chocolate Rain. He. He knew what he was doing. He knew what he was doing when he created that classic Chocolate Rain. I love the thumbnail that no no youo Meme uses of him leaning away from the microp. That's what he does. I wonder. So let's see what's going on.

[Paris Martineau]

What is he up to today?

[Leo Laporte]

His real name is Adam Nyerere Bonner, but he called himself Tay Zonday. He was a grad student when he recorded it in 2007 from Minneapolis.

[Jeff Jarvis]

What did he study?

[Leo Laporte]

Ah, that's a good question. Chocolate Ray made it to YouTube's front page. This is early on in YouTube, right? And 4chan showcased chocolate rain all over the page, so give 4chan a little credit as well. He starred in a number of commercials, including one for Comedy Central's Last Laugh and a cherry Chocolate rain ad for Dr. Pepper.

[Jeff Jarvis]

Oh, good.

[Leo Laporte]

Oh. In February 2010, he was in a Vizio super bowl commercial with Beyonce.

[Jeff Jarvis]

Geez.

[Leo Laporte]

He appeared in Weezer's Pork and Beans music video. Performed with Stephen Colbert at an autism charity event in 2010, he has received some work as a voice actor. Was hired by NASA to voice part of their presentation celebrating 50 years of Nat. So history. He was voice acted in Robot Chicken in a Chocolate Rain spoof episode.

[Paris Martineau]

Okay, cool.

[Leo Laporte]

So there's. There's many parodies, including this one by Chad Vader.

[Harper Reed]

Stay behind.

[Leo Laporte]

His colors celebrate Chocolate Brain. He's short the same crime has a higher price to pay. Chocolate.

[Paris Martineau]

All right.

[Leo Laporte]

This is when the Internet was not full of crap. Right? Right.

[Paris Martineau]

Yeah. This was the early YouTube days. Two years into YouTube, how much work.

[Jeff Jarvis]

They had to do to do that.

[Harper Reed]

That's a lot of work in 2007.

[Leo Laporte]

Yeah, you're right. It wasn't easy. Yeah, you're right. Here's Cherry Chocolate range. Oh, I got too many videos going. I gotta stop.

[Paris Martineau]

I was about to say it's crazy.

[Leo Laporte]

Chad Vader's still singing. Sorry, I don't know what's going on.

[Paris Martineau]

I think that's a good way to.

[Leo Laporte]

Thank you for bringing back tay on day Chocolate rain, some stay dry, weather this, feel the pain. Mr. Jeff Jarvis, a pick for you.

[Jeff Jarvis]

So last two weeks ago, as you know, I went to Salem, Oregon to testify against Senate Bill 686, which I think is crap because it just tries to hold up technology companies for newspaper companies, for lobbyists. I saw this, I submitted my testimony in writing, and I gotta say, somebody else did a better job than me. I'm Mr. Morris.

[Leo Laporte]

Dear Committee members, was this a comment that somebody posted?

[Jeff Jarvis]

Yes, it's actually posted to the comments.

[Leo Laporte]

On the legislation from Samuel Morse on behalf of telegraph operators of America. Dear committee members, as inventor of the telegraph, I am in strong support of SB686. Many, many years ago, my invention became obsolete when new and more efficient means of communication were developed. Suddenly, all my hard work was for nothing. People no longer had to slowly tap out messages and dashes and dots and then hand deliver those messages to people mil on horseback. Frankly, I lost my shirt. If the opportunity had risen for me to get a bundle of cash from those creeps who invented the telephone and the fax, mail machine and email, don't even get me started. I would have grabbed it in a second. Be sure to get a pile of money from those pesky Internet jerks to prop up the local newspapers. And while you're at it, consider taxing offset presses. Those things near it really screwed over Google. Gutenberg. Oh, wow.

[Paris Martineau]

A man after your own heart.

[Harper Reed]

Yes.

[Leo Laporte]

Isn't that a great comment?

[Jeff Jarvis]

Some thought it was me, but it wasn't. I wish it were.

[Leo Laporte]

It should. It's brilliant. It's almost as good as Larry David set up of oh, Bill Maher.

[Jeff Jarvis]

Oh, my God. Did you see what the New York Times then did?

[Leo Laporte]

No.

[Jeff Jarvis]

The deputy opinion editor then has to write a separate column explaining a joke, number one, which you'd never do, and number two, saying, oh, he wasn't comparing Trump to Hitler.

[Leo Laporte]

Oh, that's not.

[Jeff Jarvis]

Yeah, he was.

[Leo Laporte]

It wasn't the point it was, you know, God, I want. I. I don't watch Bill Maher. My wife loves Bill Maher. So I heard it in the background, him saying, you know, there's too much hate in the world. So I went and had dinner with Donald Trump and Kid Rock at the White House, and he was a very courteous fellow. And so Larry David did a brilliant job, I think, of puncturing that.

[Jeff Jarvis]

The best piece and people, many people have admired it is this is the tan suit. Hitler came in wearing a tip where his tan suit.

[Leo Laporte]

It's. It's actually good. Good on you, Larry. Was it. Was it in the New York Times?

[Jeff Jarvis]

Yes. New York Times, yes. Yeah.

[Leo Laporte]

Amazing.

[Harper Reed]

Yeah, I.

[Leo Laporte]

It's a shame they felt like they had it. Explain what it all.

[Jeff Jarvis]

Oh, it's. It's just so pedantic, and we don't normally do satire. And if the satire cares about this.

[Leo Laporte]

Network, Larry, next time go to the New Yorker, please. They'll understand it better there.

[Jeff Jarvis]

Well, he got more attention, I think, from the New York Times, which is good.

[Leo Laporte]

Yeah, it was brilliant. And I was just glad to see it because I had the same exact visceral reaction.

[Jeff Jarvis]

I did, too. I couldn't watch it. I watched the beginning of that segment, and I just said, no, not going to do it. Can't do it.

[Leo Laporte]

It's my problem with Bill Maher in general.

[Jeff Jarvis]

But I will be curious this Friday whether Bill Maher responds.

[Leo Laporte]

Oh, he will. You know, he has to.

[Jeff Jarvis]

He has to. He loved it. He loved the attention.

[Paris Martineau]

I'm being silent because as someone under the age of 50, I have no opinions on Bill Maher.

[Leo Laporte]

Is he really for old folks?

[Jeff Jarvis]

But Paris, he would have opinions about you.

[Leo Laporte]

Oh, he does. He has.

[Paris Martineau]

A lot of people have opinions about me. Doesn't he have opinions about them?

[Leo Laporte]

In the words of my calculator.

[Paris Martineau]

Can'T see it. Oh, oh, don't panic. There we go.

[Leo Laporte]

Panic.

[Paris Martineau]

Let's burn another one, guys. Come on. Bring back the sticker.

[Leo Laporte]

Peel it off. All right, ladies and gentlemen, thank you so much, everybody. Wait a minute. I have pictures of my son going to dinner with Martha Stewart. Oh, well, wait a minute.

[Jeff Jarvis]

Did you ask him about the Tesla?

[Paris Martineau]

You should ask him to ask Martha about the Tesla idea. Would she want to take a crack to support.

[Leo Laporte]

Apparently he said, I decided to wear a car cardigan to. To meet Mar.

[Paris Martineau]

There's no Martha in there in that photo.

[Leo Laporte]

Well, there will be. This is. He's going to.

[Paris Martineau]

The cardigan is the right move.

[Leo Laporte]

Is it right? It's the Right thing to wear a cardigan on the way to Martha.

[Jeff Jarvis]

The mustache is quite dashing.

[Leo Laporte]

Oh, I know. He's got. He's stuck with it now. He cannot ever shave. It's his trademark. He's like Chef Boyardee. He's just. He's stuck with it. Ladies and gentlemen, thank you you so much for putting up with us for the last couple of hours. Thanks to Harper Reed, our special guest.

[Jeff Jarvis]

Harper was delightful. It was great.

[Leo Laporte]

Isn't he great? He's just a character.

[Jeff Jarvis]

Fun.

[Paris Martineau]

He was fantastic.

[Leo Laporte]

Great energy.

[Jeff Jarvis]

Some people can't pull off character. They try too hard. He just is.

[Leo Laporte]

Yeah, he's like a nice Shingi, which no one understands. What? I just. That one.

[Jeff Jarvis]

I got it. But I can just show a picture of Shingi. Come on, you got to show.

[Leo Laporte]

Jeff Jarvis is professor of journalism. Is Shingi still at aol? He was at aol.

[Jeff Jarvis]

Oh, no, no. He got booted a long time ago.

[Leo Laporte]

He was AOL's digital prophet.

[Paris Martineau]

Oh, he is reverse Shinki.

[Jeff Jarvis]

Yes. People couldn't stand shaky. I met Shiggy once. He was very nice.

[Leo Laporte]

I'm sure he's nice. It's just doing the hair like that.

[Jeff Jarvis]

It's that embarrassing moment where, you know, I have said funny things about him and I didn't know if he knew it. He didn't. What? That's good.

[Leo Laporte]

Yeah.

[Jeff Jarvis]

But I was also event at Columbia on Friday, and the head of the Times editorial page was there and she was a little frosty.

[Leo Laporte]

Oh, she knew that. You thought it was the. What do you call it? The broken time.

[Paris Martineau]

Hashtag broken times.

[Leo Laporte]

Hashtag broken times. If I could. You know what? Shingy. I'm just jealous. If I could do that with my hair, I would.

[Paris Martineau]

Yeah, you should grow your hair out. Leo.

[Jeff Jarvis]

What is Shiggy doing now?

[Leo Laporte]

I don't know. I just got a haircut. You're telling me to grow it out now?

[Paris Martineau]

Just saying. Could be.

[Leo Laporte]

I don't know. I think it's now my trademark. Much like Saul Hank's mustache. I think I have to have perfect hair.

[Harper Reed]

Oh, God.

[Jeff Jarvis]

You got.

[Paris Martineau]

What? If you had perfect hair that was shoulder length, you could do the.

[Leo Laporte]

I would love that. I would love to. I. Look what happened to your dad. They thought he was a.

[Paris Martineau]

Listen, I will say it's changed his whole vibe. People think he's a celebrity now just because he's has long hair.

[Leo Laporte]

It does do something. I mean, Howard Stern is just a schlub without the long hair.

[Jeff Jarvis]

Yeah, you got to go to shingy.com.

[Leo Laporte]

Oh, he's still around.

[Jeff Jarvis]

Oh, yeah, he's still selling.

[Leo Laporte]

Gotta be selling.

[Jeff Jarvis]

The shingi artist, globe trotting speaker and market seeker. He is a storyteller who identifies emerging transformation, inspires clients to think differently.

[Paris Martineau]

Every scroll down made me laugh in a different way.

[Jeff Jarvis]

Keep going.

[Leo Laporte]

Yeah.

[Paris Martineau]

The. The prayer hands. Woof.

[Leo Laporte]

Oh, what the heck do you do?

[Jeff Jarvis]

He's a renowned speaker and if you want to.

[Paris Martineau]

This is like technology. This is like a satire of a guy.

[Jeff Jarvis]

Brilliant.

[Leo Laporte]

David Shane at the table.

[Harper Reed]

So let's get on with with it. We are the connected generation. Consumers are way smarter than we think they are.

[Leo Laporte]

The secret about. I wish I had had the presence of mind to have a trademark, you know, that's the key. Yeah, that's the key. That's why you got to have a mustache. Like now.

[Paris Martineau]

That's an ideas guy.

[Leo Laporte]

That's it. That's a guy with ideas. He's got ideas, you can tell, flowing.

[Paris Martineau]

Out of his head. They can't be contained.

[Leo Laporte]

Official.

[Jeff Jarvis]

The official title is he is a creator, creative omnivore and visionary futurist.

[Leo Laporte]

And he spits up whatever he eats so you can enjoy the regurgitation of his brilliant.

[Paris Martineau]

He's a master baby bird.

[Jeff Jarvis]

Shiggy's boundless creativity and design mastery establish him as a preeminent authority in crafting iconic identities. His razor sharp instinct, visual art, artistry and verbal eloquence make him an indispensable advisor.

[Leo Laporte]

Orchestrating judge. Anyway, I'm not going to judge.

[Paris Martineau]

I'm sure a lot of this is true, but if anybody. If I ever had a website describing my own verbal elegance as using the words verbal elegance, I would hope that someone puts me out to pasture.

[Leo Laporte]

But, you know, this is why we need to promote ourselves more. We need to be more out.

[Jeff Jarvis]

Shiggy's got it. We're. Are we doing.

[Leo Laporte]

Yeah, we're just doing podcast.

[Paris Martineau]

He's clearly got it and we don't.

[Leo Laporte]

We're in our houses doing, my hair is flat. Yeah, you know, we didn't knock it. Jeff Jarvis. I'll tell you who he is. Professor of journalistic innovation at the Craig Newmark Graduate School of Journalism.

[Jeff Jarvis]

Don't rush it.

[Leo Laporte]

City University, New York. I try to rush it every single time. He is now at Montclair State University in New Jersey and also a SUNY Stony Brook in New York State. Author of the Web We Weave the Gutenberg Parenthesis now in Paperback magazine and more. And he is a beloved figure on the international stage. Jeff Jarvis. Dot com. Thank you, Jeff.

[Jeff Jarvis]

Thank you.

[Leo Laporte]

Paris Martineau has a shorter story. She is creative omnivore, ready to suck up all the information you've got. If you've got a tip, email. You said tips at Paris nyc. Is that what you said?

[Paris Martineau]

Yeah, that's true.

[Leo Laporte]

That is spitted us.

[Paris Martineau]

No, I sucked up.

[Leo Laporte]

I sucked it. You were sucking it up.

[Paris Martineau]

Information.

[Leo Laporte]

Okay, that's yes on the signal. Martino01 give her some good tips. You know what? This is the thing right now. This is your opportunity to write the book that blows the lid off at X. Whatever that X might be.

[Paris Martineau]

Not x dot com.

[Leo Laporte]

Not that X. No, no, no.

[Paris Martineau]

Different X.

[Leo Laporte]

The N. I'll say N. The. The thing, whatever. The thing is the blank. The blank. Fill in the blank.

[Paris Martineau]

We'll figure it out.

[Leo Laporte]

Thank you, Paris. Thank you, Jeff. Thank you to our very long suffering audience.

[Jeff Jarvis]

And crew.

[Leo Laporte]

And crew. We're glad you made it to the end. They had to. You did. So pat yourself on the back. We do this week in intelligent machines. No, we decided not to call it that because it would be twim. We do twim.

[Paris Martineau]

We love to go twimming every week with you.

[Leo Laporte]

We're going twimming. No, it's just intelligent machines. We just truncate it, make it easy. It's im. It's wacky week, so we're a little punchy today. We do it every Wednesday right after Windows Weekly, 2pm Pacific, 5pm Eastern, 2100 UTC. We always kick it off with a great guest and then stories about I am Intelligent Machines and I hope you will come back and enjoy the fun and festivities next week. You can watch us live if you're in the club on Discord, but everyone else can watch live on YouTube, YouTube, Twitch, TikTok, X.com, facebook, LinkedIn and Kik after the fact on demand versions of the show available at TWiT TV IM. There's a YouTube channel dedicated to intelligent machines and you can also subscribe in your favorite podcast client, get it automatically the minute it's available. If you do that though, please leave us a five star review on there because it helps spread the word about the hottest new show on the Internet. Thank, thank you everybody. We'll see you next time on Intelligent Machines. Now peel me off.

[Harper Reed]

Like your favorite startup's growth curve, T Mobile's coverage keeps scaling because T Mobile helps keep you connected from the heart of Portland to right where you are on America's largest 5G network switch. Now keep your phone and T Mobile will pay it off at the $800 per line via premium prepaid card. Visit your local T Mobile location or learn more@t mobile.com KeepAndSwitch up to four lines of a virtual prepaid card. Allow 15 days qualifying unlock device, credit service port in 90 plus days, device and eligible carrier and timely redemption. Required Card is no cash access and expires in six months.

[Leo Laporte]

Ever notice your dog slowing down and having health issues and wonder what can I do to make them better? Well my friend, add rough greens to your dog's food for 90 days and I guarantee you'll see changes that will emit amaze you. Greetings naturopathic doctor Dennis Black, inventor of Ruff greens here and I invite you to give your pup the Ruff Greens 90 day challenge. In the first 30 days you'll see shinier coats and increased energy. By day 60, your dog will have a stronger immune system, less shedding, improved joint function, all due to the live nutrients that you've added to their diet. And at 90 days, better digestion, reduced inflammation, improved heart health health and you may even have reduced their cancer risk. Fetch your dog a free Jumpstart trial bag today. Go to try roughgreens.com use promo code Try rough. That's T R Y R U F f. Go to tryruffgreens.com use promo code try rough. You just cover the shipping. You don't have to change your dog's food to improve your dog's health.

[Harper Reed]

Just add a scoop of rough greens.

[Leo Laporte]

Marketing is hard, but I'll tell you a little secret. It doesn't have to be. Let me point something out. You're listening to a podcast right now and it's great.

[Harper Reed]

You love the host.

[Leo Laporte]

You seek it out and download it. You listen to it while driving, working.

[Harper Reed]

Out, cooking, even going to the bathroom.

[Leo Laporte]

Podcasts are a pretty close companion. And this is a podcast ad. Did I get your attention? You can reach great listeners like yourself with podcast advertising from Libsyn Ads. Choose from hundreds of top podcasts offering host endorsements or run a pre produced ad like this one across thousands of shows. To reach your target audience in their favorite podcasts with Libsyn ads, go to libsynads. Com. That's L I B S Y N Ads. Com. Today.