Loading summary
Transcript100 lines
- [00:03]
A
Hello and welcome to the vergecast flagship podcast of the Vercel Rest API. I'm your friend David Pearce, and today on the show we have a challenge. The way this is going to work, we're going to, we're going to do a bunch of these in a row. But the way this is going to work is we've gathered a bunch of people, we're going to give them all a challenge, and then we are going to go off for several weeks and report back on what we found. Today we are starting with two of my colleagues, Jake Castronakis and Hayden Field, and we are going to do a Vibe coding challenge. The first thing we're going to do is kind of lay the land, talk about where we all are in our Vibe coding experiences, and then we're gonna all go see who can do the best in Vibe coding. It's a lot of fun. I'm very excited about this project in part because it's gonna get me to like, actually get some work done for a change for the podcast. It's gonna be awesome. But first, here's everything else happening on the Verge today. This is 90 seconds on the Verge for Monday, June 29, 2026. Mythos is back. Kind of the anthropic model that was once deemed too dangerous to release publicly and actually threw even more chaos into the already chaotic relationship between anthropic and the is now being released to a Trump administration approved list of organizations. Who those organizations are, who they will be. Whether you'll ever get to use Fable 5 again, hard to say. Nobody seems to quite understand the AI regulatory environment right now. And given that the Supreme Court also ruled today that President Trump has the authority to fire two FCC commissioners, which changes the way that agencies work. Whatever happens here, it will start with the White House. Meanwhile, Comcast is splitting into two companies in the latest of a seemingly endless supply of media mergers and unmergers. Now, there will be Comcast, the Internet, cable and mobile carrier, and there will be NBCUniversal, which is the studios, theme parks and streaming networks, including Peacock. None of this, by the way, applies to Versant, which Comcast spun off in 2024. The media business just continues to be chaos. And a fun side note of this, by the way, is that I have a new disclosure to do here. I guess it's going to be something like Comcast is an investor in pmx, the parent company of the Verge. I'm going to have to get used to that one. And finally, quick PSA for you. WhatsApp is rolling out username support later this year so that you can find people and be found without having to share phone numbers. The feature's not live yet, but you might be able to go to Settings, account username and reserve yours. There are like 3 billion people on WhatsApp, so you should probably move fast. You can read more about all of this@theverge.com that is 90 seconds on the verge for Monday, June 29th.
- [02:26]
B
Support for the show comes from ServiceNow. AI is moving fast across the enterprise, but without visibility, it's just chaos. Different tools, different models, different teams using AI in completely different ways. ServiceNow turns that chaos into control. With the AI control tower, you see all your AI across the business in one place, what it's doing, what it's done and what it's about to do. So you stay in control. To put AI to work for people, visit servicenow.com
- [03:01]
C
we all do it. You have a night for yourself, but don't like the sound of the silence. So you turn on the TV just for the ambiance. It's a little trick that helps you feel like you've got company and aren't alone. And other insurers, well, they may make you feel alone, but when you switch to Geico, you've got claims reps available around the clock, so whenever you need, you'll have people around to help. And let's turn on the washing machine just for good measure. Isn't that soothing?
- [03:27]
D
It feels good to have support. It feels good to Geico.
- [03:30]
A
All right, now it is time for some Vibe Coding Shenanigans. Joining me now, the Verge's senior AI reporter, Hayden Field. Hi, Hayden.
- [03:38]
E
Hey.
- [03:39]
A
And the Verge's executive editor, Jake Castronakis.
- [03:42]
D
It's actually chief Shenanigans reporter.
- [03:44]
A
Thank you, Chief Shenanigans Reporter. What I want to do here is something that you guys actually did a little bit of while I was on parental leave last year, which is just all simultaneously attempt to go on a technology adventure together. You guys all did Vibe coding projects last year and I listened to that episode of the Vergecast with like sheer rage because none of you tried hard enough and you all gave up too fast and we're going to do some new stuff. So I have assembled the three of us and the way that this is going to work is at the end of this episode, I'm going to give all three of us a task. And it is, it is largely a Vibe coding based task and, and we are going to regroup in a month and just share what we have done on that task. Um, you will be relentlessly shamed for not doing a good job or trying hard enough. Um, you will, but the world is your oyster. So we're going to get to the task at the end. But what we were going to do is a task ahead of this. The first assignment I gave both of you was basically like, go vibe code. Something that you have made that is useful and interesting in your daily life. And you both were kind of like, we've done this already, so we're going to start our experience with some show and tell. And, Jake, I want you to go first and tell us a little bit about your vibe coding setup and some of the stuff that you've built.
- [05:00]
D
Yeah, I mean, I feel like what I should actually start with is the failures, which is, it turns out. So when we did this last year, vibe coding, as somebody who has no coding knowledge, it wasn't there. Like, it didn't work. We tried stuff and it was constantly flopping.
- [05:20]
A
Yeah.
- [05:20]
D
And then when I checked back in again, you know, this winter, oh, it had changed a lot, you know. And so the thing you will find if you start experimenting with vibe coding is like, you're gonna make a lot of stuff that in the moment, you're like, this. I'm cooking like this. This is amazing. And then you use it and you're like, I have no use for this. Like, this is total garbage. So I. I made a bunch of tools that I thought would be useful for work. Different kinds of news digests, different kinds of things that would, you know, pull together Twitter feeds, things that would. I just. One time I told Claude, I was like, hey, find every single, you know, open U.S. government database that is relevant to the Verge and, you know, pull all of the files that are relevant to companies we cover and then just, like, do some analysis on them. And it was just like, cool, let me just get the API for, like, the treasury. And. And it just, like, did it and it built it. And then I had all this data and I was like, there's nothing. Like, I don't know what to do with this. And so it's gotten there. Like, the technology has advanced a lot. But I think the thing that I found is that finding an idea that is useful to you specifically is harder than building it now, which is. Is nice.
- [06:41]
A
This is. I mean, this is, like, increasingly supported by the data, by the way, right? That, like, you look out into the vibe coding world and everybody is building tons of stuff that amounts to nothing. And this is becoming like an Enterprise crisis because everybody's like, we're using all these tokens for nothing. What is the point of any of this money that we're spending? So you've, you've encountered the same problem that every Fortune 500 company is happening in spades right now.
- [07:03]
D
Except I only paid like $20 a month, so it was not, it was not so bad. Eventually I did strike something though, right? I do. So I found a small focused problem I have at work, which is that I have a lot of meetings with people on our team every single week. Right. And by the, by the time the next week rolls around, I have forgotten everything we talked about. And that's because, like, I, I've tried a million different systems for taking notes for these, these meetings, and every single time, it's just, I, I, like, I don't read my old notes back. They're kind of a mess. They're in different places. I, like, had just like a long running text edit doc. Like, it's not. And so I kind of thought, like, oh, wait, what if I have a notes app that is specifically for meetings with our team and I can list everybody's names and I will have it automatically roll forward every single new entry with the action items from our last meeting and I can have it automatically summarize, you know, the past few weeks of notes. And so now when I go to have one of these meetings each week, I'm like, super prepared because it's giving me, like a brief on what we've been talking about. I have the action items that I'm supposed to be coming in with. And this has been great for me. And so, like, this, this was the first thing I built that was really, really successful.
- [08:23]
A
Okay, wait, can you just pause for one second?
- [08:25]
D
Yeah.
- [08:25]
A
Hayden, I want to know how that description makes you feel. I once described a vibe coding project to Nilay, and he, he told me that it sounded like I was describing a dream to him. And that has hurt my feelings ever since he said that. Uh, I want to know, like, does Jake's description of what he vibe coded do anything for you?
- [08:43]
E
Yeah, I mean, I don't know. With 5 coding, I feel like the simpler the better. Like, it's usually solving a problem that, like, is really easy, but you just can't get it together. That's how mine was. At least what I, like, built. And so this seems similar. It's like, yeah, he could, you know, take the notes and look back at them, but he's not. So he's like, figuring out a small Tiny fix that's going to make him look. So I feel like, yeah, I mean, it seems like simple enough to work, in my opinion. Like, you gave us the prompt of the toothbrush test, like something you're gonna use twice a day. And that's what I kept thinking about. So it's like this seems to pass that. I feel like people try to go too crazy with it and then they just, like we were talking about, it just flops and like, they never use it again. So I like it.
- [09:26]
A
The feature creep problem is so real in Vibe coding because you're like, I can do anything. Maybe I'll just do everything. And it just all falls completely on its face.
- [09:34]
D
Yeah, I know exactly what Nilay is saying though, because it, it, it, you know, as a piece of software, like, it's. I, I think it's pretty polished, right? I could share it around, other people could use it. But it, this, doing this has made me re, like appreciate people who actually make software so much more. Because this, this thing is precisely designed to solve my problems and it doesn't make sense for anybody else. It is specifically tuned to my workflow. And I don't. Suddenly I'm like, how do you. If I, if I'm passionate and I'm gonna, and I'm gonna make a notes app. Like, how do I know it's a thing other people want? Like, how do I know that it's an interesting idea? Because we see, we. Right, we cover this stuff all the time. David, like, you've written about 9 million to do apps and each of them has like a, A, an interesting marketable idea. And like, I don't, I don't know what that is for my app, but it doesn't matter because it's only for me. But that's why it's so. It's like I can't pitch this to somebody.
- [10:34]
A
Jake, one more question for you before we get to Hayden's project. What is broken about the app that you made yourself?
- [10:40]
D
I mean, this app is like really simple. It's, it's basically just a Safari web view with some notes that I have. I, I hate to tell you, but, like, I haven't found it yet, though. I have like, out of paranoia, like, once a week I ask Claude to review it and be like, hey, make sure the data isn't going to fall apart. Make sure it's not going to delete all my notes. Like, make sure I'm not going to lose everything.
- [11:01]
A
Can I, can I just make a quick suggestion? Just go back to Claude and say, hey, can we back all of this up as text files?
- [11:06]
D
Just.
- [11:07]
A
Just do that for me and it'll make me feel better.
- [11:09]
D
All right, I'll look into it.
- [11:10]
A
You're. You're going to end up with, like, a web cache problem that is just. Everything's going to fall apart. If it starts with localhost3000, you've made mistakes.
- [11:19]
D
That's my other vibe coding app.
- [11:22]
A
Do you wanna tell us about that real fast?
- [11:25]
D
Well, I made an entire email client.
- [11:29]
E
Oh, my God, I can't wait to hear about this. This is my dream.
- [11:34]
D
This is. Dude. Okay, so, like, I've had this same email address for, like, over 10 years.
- [11:42]
A
Yeah.
- [11:43]
D
And it's been public the entire time. And so I get so much garbage. And so I was like, okay, how can I figure this out? How can I create an experience where I don't have to deal with a million emails at once? And the answer is just labels. Like, it's just Gmail labels. But Gmail labels are, like, on this, they're, like, hidden in the corner. There's, like, not like a fast way to access them. And I was like, oh, you know what would be perfect? You know how Gmail has, like, the big taps up top for, like, promotions and whatever.
- [12:13]
B
Yep.
- [12:14]
D
I was like, oh, what if I just put my labels up there? That would, like, solve my exact problem. Gmail doesn't let you do that. You can't choose what goes up there. So I was like, okay, I'll just build an entire app that puts my labels up there. And it's great. And it makes it super easy for me to, like, triage my emails because it, like, shows me all the stuff that's from, like, you guys, my coworkers. It has stuff from, like, known trusted contacts who I want to hear from. It has a tab for, like, literally like 5,000 plus unread newsletters. It's great. I love it.
- [12:46]
A
You basically built yourself, like, sucky Superhuman, which people pay 30amonth for. Like, that's. It's kind of great.
- [12:54]
D
Some.
- [12:54]
A
Some PM at Gmail is listening to this, is going to. And is going to get a promotion by going and launching what you just described as a new version of Gmail.
- [13:01]
D
I would really appreciate it, but. Yeah, but it's also, like, I had to, like, enable a bunch of weird Gmail API things every single time I did something. I was just. I would be like, claude, please do not accidentally send emails to my entire contact list. Like, like, it's a little scary. Like, I would rather Kind of not do that. Like a local notes app. Like the worst thing that happens is that I go into a meeting with you, David, and I forget what we talked about last week. Like, it's, it's fine. This one. The worst thing that happens is that I accidentally like spam every single person who.
- [13:33]
A
You hack your own email.
- [13:34]
E
Yeah.
- [13:35]
D
But to your local host thing, it's like for some reason that I haven't quite felt like dealing with, I have to log in once a week and then the login involves me going back into my browser and there's this local host thing and I just. Yeah, it's fine.
- [13:48]
A
But this is the kind of stuff that I really get a kick out of because my own experience, which we'll get to in a minute, is I have built a bunch of software that has a bunch of bugs that I would find completely reprehensible in anybody else's software.
- [14:01]
D
Yeah.
- [14:01]
A
But in mine I'm like, that's fine, I can deal with this. I don't know how to fix it anyway. Hayden, let's get to you. Have you, have you been vibe coding what have you built for yourself?
- [14:11]
E
Yes. Okay, so I have been vibe coding. What I made was a habit tracker, which I know everyone makes, but for me I have add and so I really desperately need one. Like I've learned about myself that I can't like start a new habit without accountability. Basically, you know, it takes like six weeks, three to six weeks of me having accountability with someone else or with like a inanimate object. Whenever I have a tracker, I need to have some sort of accountability for like three to six weeks and then I'm set, I'm in it. I don't need the accountability anymore. But it is very tough for me to start a new habit. So every time I join a new gym, like I make five gym friends, I'm like, I gotta really lock in here, you know, because I'm not gonna do something just for myself. I have to just not disappoint someone else or not get a star on my habit tracker or whatever that is gonna make me do the thing. Not just the long term knowledge that I'll be healthy or something.
- [15:03]
A
Totally.
- [15:04]
E
So at least in the beginning. So basically I used to download all these different habit tracker apps like you guys were talking about with to do list apps and you know, note taking apps. And so I've tried like every habit tracker app, there's always something that bothers me about it or it's really expensive. I even downloaded this one that I loved for a while called habit, but the problems with it were that you couldn't create a habit that wasn't every day. Like, for example, I need to take my iron supplement every other day and it's something that I'm always going to forget because you have to do it on an empty stomach at night and you have to take it with your vitamin C. So it's just like for an ADD person, that's just too many details to recall on the fly. And so, yeah, or I don't want to work out every single day, but I also don't want to be like shamed for not working out every day. And if I see it on the thing and it's like workout and then I don't check it off, I will feel shame even if it's not a big deal. I just, you know, I have to feel like I'm succeeding in order to have motivation. Basically, that's how I work. And so what I made was a habit tracker app that has like a section for daily habits. It's extremely simple. It's like rectangles that each have a color. And then when you swipe across the habit to mark you completed it, it becomes a deeper color and stars rain down. So you get, you get a reward, you get a gold star.
- [16:22]
A
Is this like a native mobile app you made yourself?
- [16:25]
E
Yeah, I made a web app and then I made a native mobile app. So I will say, just so you
- [16:32]
A
know, Jake Hayden's already winning. Just by making a native mobile app, Hayden wins this round of the Vibe Coding Olympics.
- [16:38]
E
I will say though, that it was really hard to build the native mobile app. Like, it took a lot of time and it kept breaking. And I tried, I did both of these things, both the web app and the mobile app. I tried both Codex and Claude code and I, I compared and contrasted and I had an interesting experience with both. But anyway, yeah, so it also has a section for three to four times a week, which is like working out and creative project or some sort, because I'm starting to try to like, learn how to embroider and stuff like that. And then only every other day does it pop up to say like, oh, take your iron. And that's another habit that appears. So basically I made it perfectly for myself and I think this is actually something that other people would want to use. But I will say, you guys, it said that I had like 15 high security vulnerabilities and 15 medium security vulnerabilities. So I'm not a developer, so I don't know what they are. But it seems like, you know, it could be dangerous for people who don't know what they're doing to, like, be using this to, like, deploy apps onto the open Internet.
- [17:43]
A
So my favorite thing is you go through these things and you say, like, oh, Xcode is telling me I have all these problems, and cloud code is just like that. Don't worry about it.
- [17:50]
E
Yeah, exactly.
- [17:51]
A
Probably nothing, but we sure it's nothing.
- [17:54]
E
I literally asked Codex like, is this a problem? And it was like, oh, don't worry about those. That's what it said. I was like, okay. So, yeah, I mean, it works on my phone now as long as I'm connected to the same wifi. So, you know, I mean, I would like it to work all. It seems to. It hasn't broken yet, but I just, you know, it works. So far it's going well, and, you know, it has made me more likely to do my. My new habits, which is great. So that's really.
- [18:24]
A
That's really impressive. I do think that's a fun one because that is the sort of thing that I. I can absolutely see how you're like, oh, other people would. Would be like this too. Whereas, like, there are some things like, Jake, the first thing you described in particular is like, this is very specifically for me and my thing. But then it's very easy to be like, well, lots of people have running one on ones at their work. Maybe other people will want this too. And then it's like, I think an easy trap I've seen a lot of people fall into is you're like, okay, well, what would other people want? And then you've accidentally built an app you don't want anymore. Whereas I feel like, Hayden, everything you just said to me was like, oh, this is very cool. That's something I would use too. And then you're like, oh, and stars come down. And I'm like, I would turn off the stars. I don't want the stars. Get the stars out of my face.
- [19:07]
E
You can adjust those so you have
- [19:10]
A
an adjustment for the stars, even though you like the stars. Stars.
- [19:13]
E
Well, I did originally, and then I was like, no, I'm just gonna keep them. So I. I left them. At first it was like, it was trying to be too elegant. It was like a quick sparkle. And I was like, no, this needs to be like a reward function. You need to make them more intense. And then it did.
- [19:26]
A
I want this to rain stars. That's really good. Mine I've talked about a bunch already on. On this show and on the website. But basically I built myself. I've. I've built two things now that I use every day. The first one is just a, basically a way to see my calendar and my to do list in one place, which is shockingly hard to do in a way that is not like hideously ugly. Um, yes, offense to every app that does this already. I know there are lots of you and I've tried them all and you're all bad in your own special way. Um, but I can. I just now have like a running list of the events that I have for today and the tasks that I have for today. And all the events are synced to Google Calendar and all the tasks are synced to Todoist, which is what I use for my to do list. But I can just see them and add to them all from one place. And that has been like life changingly great. It exists on my phone. It's just the thing I open to see what's going on today. The other one, which I just built yesterday and is working very well for me, is I constantly have 700,000 tabs open. And I discovered the problem is that I don't. I need to put my tabs somewhere. Some of them are like, oh, this is going to be a story. Some of them are like this, this is going to be a thing that I'm going to put in the installer newsletter. Some of them are like this I just need to come back to later. And I have. I built a web clipper that is just a filing system. So I click the thing and then I click where it goes and it just sends the link to the right place and I don't have to think about it anymore. And I am using the least RAM on my computer that I have used ever in history. Uh, it's this tiny little piece of software. It like barely works for anybody, starting with Vibe coding. By the way, I really recommend starting with browser extensions because they're very simple to make, they're really easy to load. Um, you can obviously do lots of like security disaster things, but like the, the there are fewer horrible ways for it to go wrong when you're just trying to like do a simple thing with a web page on your computer. Highly recommend messing around with Chrome extensions to solve some of your browser problems.
- [21:26]
E
When I got a new car, I thought my insurance premium would increase and empty my bank account. Like if Fatween won the lottery.
- [21:33]
D
I've invested most of my winnings in chicken tenders because they're bomb but bro, I bought a house and it's sick bro.
- [21:39]
E
I'm thinking the floor is gonna be
- [21:40]
D
all trampoline bro with a helipad on the roof. The contractor said it's structurally unsound. They're just being babies.
- [21:47]
E
But switching to GEICO saved me hundreds so my bank account is safe.
- [21:51]
D
It feels good to save some hard earned cash. It feels good to geico.
- [21:57]
F
Support for this show comes from Fetch Pet Insurance. Do you have a pet? Every six seconds a pet owner in the US gets hit with a vet bill of over $1,000 and it's almost always an unwelcome surprise. That's where Fetch Pet Insurance comes in. Fetch is the most complete pet insurance get paid back up to 90% of vet bills. You can use any vet in the US and Canada. All vets are in network. Go to fetchpet.com save right now for your free quote. That's fetchpet.com Save heat up your 4th
- [22:34]
G
of July at the Home Depot with our wide variety of grills under $300 and make every gathering one to remember. Give your outdoor space a glow up whatever your budget is with savings on seasonal plants starting at $5. With the grill fired up and your backyard set to perfection, you'll be able to invite friends and family over to kick off the party. Start celebrating with low prices guaranteed at the Home Depot. Prices may vary by store. Exclusions apply. Seehomedepot.com Pricematch for details study and play
- [23:05]
D
come together on a Windows 11 PC and for a limited time, college students get the best of both worlds. Get the unreal college deal, everything you need to study and play with select Windows 11 PCs. Eligible students get a year of Microsoft 365 Premium and a year of Xbox Game Pass ultimate with a custom color Xbox wireless controller. Learn more@windows.com studentoffer while supplies last ends June 30 terms@ aka mscollegepc the email
- [23:36]
A
client you described, Jake all of a sudden has my brain spinning where I'm like this is a thing that I want. I have gone back into a bunch of the supposedly useful AI email clients recently. My inbox is the most out of control it's ever been. I don't know how you two are as email people, but like I was inbox zero person for a long time. I have like 700 unread emails right now and it's stressing me out.
- [23:56]
E
You know how I am.
- [23:57]
A
You're a mania. Yeah, you're a. You're a. You're a psychopath when it Comes to email. We don't. We don't talk about that. Um, but I think, like, I feel like we've all done a pretty good job here of, like, solving a very specific problem for ourselves, which is good. Um, oh, wait. Can I tell you the thing that's broken about mine, which I think is very funny? Uh, I set up a bunch of keyboard shortcuts so that I can, like, navigate around my little daily planner app. Uh, there are three tabs, and I can type 1, 2, and 3 to switch between them. I can type command T to have a new tab or a new task. This app, for some reason, captures all of the keyboard shortcuts on my device. So I can't open a new browser tab inside of while I'm looking at this planner app, because command T, it takes to me in a new tab when I just want it to be T. I cannot make cloud code fix this. This is where I'm like, any person who is a remotely functional developer would know how to fix this. I do not. So I am like, every keyboard shortcut on my computer ceases to exist whenever I have this app open, and it drives me insane. But I don't know how to fix it.
- [25:00]
D
That seems unusable.
- [25:01]
A
It's not. I just have to go over and click the new tab. But everything comes back and it's fine. As long as I switch tabs with the mouse, we survive. But anything. Once I. When I'm in the app, it refuses to acknowledge that anything else exists, which strikes me as a very normal Vibe coding problem.
- [25:16]
D
Yeah, this is what's delightful and terrible about these things. It's just like, oh, but everything else is so perfect for me. Like, I will deal with this incredibly major inconvenience.
- [25:25]
A
Yeah. Have you guys used anybody else's Vibe coding stuff yet? And obviously, it. It's hard to know to some extent.
- [25:31]
G
Right.
- [25:31]
A
But I mean, like, have you seen anything else somebody just, like, made on GitHub and been like, oh, I'll take that for myself?
- [25:36]
E
No, I haven't.
- [25:37]
A
Would you. Does that feel risky?
- [25:40]
E
I don't know. I mean, after building my own and seeing all my security vulnerabilities, I'm like, yeah, I don't know if I would. I mean, I feel. I feel like I would just rather make it myself, especially because, I don't know, like, when I was. When I was comparing and also, like, comparing with Claude Code versus Codex, there were, like, 30 security vulnerabilities on both. And I feel like anyone else Vibe coding would be like, me and just be like, oh, it's no big deal. So yeah, I don't know, but I think I would get ideas from it.
- [26:09]
D
The thing that I'm really curious about is if when I was doing this email client, I was like, oh, why don't I just like start from an open source project and then that will be like a really sturdy center and then I can just cut, like I can cut it down and build it back up to what I need to be. And like my dream is that like as this stuff becomes more and more common, people just build a core to a notes app, a core to an email app, core to a to do app. And that stuff. I don't if it does exist. Like, I'm not good at finding it and figuring out how to build off of it. And also some of these things are just so simple. But yeah, I think if it's something like that where I'm like, this is a known trusted entity, a community project. Yeah, absolutely open to using it. But some of this stuff, like, I'm more likely to like be inspired and then do my own thing than be like, I mean, this is the thing. Like they're, they're dreams, they're, they're very specific to you. And so it's like, I don't know if I really would use somebody else's thing.
- [27:11]
A
Yeah, I went. When I was building this little web clipper thing for myself, I fed there. There's this thing called Marky that is just like a neat little tiny text utility for Mac that somebody I think vibe coded and put on GitHub. Uh, and I, I put that into Claude code as like design inspo that I was like, this is what I want this to look like. Basically very simple and straightforward look like that. And cloud code was like, oh, I like Markey a lot, great job. And I was like, okay, sure, sure, Claude code. But yeah, I, I have not yet figured out my own tolerance for this stuff because I think even if you assume that most people are doing this stuff with good intentions and not malicious intentions, which I think is true, certainly people will have malicious intentions. But like you said, Hayden, even the people with the good intentions might screw this up. Which is one of the reasons I have been reluctant to share any of the stuff that I've been doing with people. I even got some feedback from people who were like, I like the way that looks. Like, will you share the thing with me? And I'm like, A, I don't really know how to do that and B, no, because I Don't know what's going to happen to either of us if I do that. And it feels like, Jake, to your point about last year, this question of, like, how do I even stand up any of this stuff? We've sort of solved that in a way, right? Like the. The ability to get a thing up and running for yourself is much, much, much easier now. Like, Hayden, you and I have figured out xcode. If we can figure out Xcode, people can do it, that's for sure. The next phase we have not figured out is, like, how to make these things safer and share and more shareable and more sort of durable out in the world. That doesn't feel like anybody has figured it out yet.
- [28:45]
D
And that it's also interesting. Like, you're talking about how you would customize Hayden's thing and as soon as you go to share something, it's like, oh, I have to clean up this, like, wacky settings thing that everything is just like, hard coded to what I need it to be. Right? Like, if it was an actual app, I would have to code. The ability to make custom tabs, the ability to like, rearrange. And it's like, it's funny because it's like literally a machine does it for you. It's as lazy as it gets. But, like, I don't want to babysit that. Like, I. There's like, you still have to do some, like, you have to do a lot of troubleshooting, particularly for bigger projects. Like the. The notes app that I made took like 30 seconds. The email app took like a weekend, which is. Is again, all things considered, not very long. But, like, it was, it was annoying. And it's like, it. I. I don't want to go deal with that and like, potentially ruin my app in the process.
- [29:44]
A
Yeah, I agree with that.
- [29:45]
E
The. The phone app or turning it into a phone app was what took me the longest. And then I was like, testing both the web. I made the web app and the phone app each with Codex and cloud code. And it was interesting because OpenAI, like Codex had better taste, but it was extremely confusing. Like, I kept having to ask, what do I do now? Or what does that mean constantly. I never made it to the phone app at first either because it messed up on the app version setup and it like, didn't set up for a compatible Expo Go version. Interesting. Um, so. And it shows like a two new animation package for the stars, things like that. Whereas Claude code was better at building web previews and really easy to understand and just better at Solving problems I couldn't conceptualize on my own. Like, I honestly didn't know how to conceptualize the three to four times a week thing. I was like, yeah, I need this, but I don't really know how it will show up and not make me feel bad. And it was like, oh, why don't we do it like this? And it was right. So I was impressed with Claude code more. But I will say it was. I don't know, it was not. It didn't have great taste without hand holding. Like, the one it made at first was functional but extremely ugly, whereas Codex's was like, beautiful. So I had to do a lot of hand holding to change the design. But I mean, the functionality was better at first. So that was interesting.
- [31:05]
A
That is really interesting. Claude code has a very, let's say, unsophisticated design sense by default. Um, okay, we should get outta here. But before we go, let me give you your task. And, and as a reminder, you have four weeks, give or take, to pull this off. And then we're gonna, we're gonna regroup in four weeks and we're gonna see how everybody did. And there are going to be prizes at the end for who did the best. And if you're stars, dating stars will rain down upon you if you win. Here is the task. The task is to vibe code a website for something going on in your life. Hayden, you're getting married is a thing that's going to happen. Would you like to vibe code yourself a wedding website? Knock yourself out. My. I already know what mine is going to be, which is I will just spoil it for you. Now I'm going to go vibe code myself a way to share photos of my kids with my family without having to either, like, create a private Instagram page or just try to, like, teach everybody how Google Photos works. I don't know how any of that's going to work or what I'm going to do, but that. I have a. I have a month to attempt to figure that out. But the, the. The prompt is very broad. You just have to build a personal website for something going on in your life.
- [32:19]
D
Are we being judged on security flaws?
- [32:22]
A
Let's stay with you. You can safely assume this is just for you, and if it is designed for someone else, fine. Like, I'm going to have to make a version of my thing that is accessible and useful to my parents. But this will not be for public consumption. So to, to whatever extent that makes you feel better about leaving the security warnings untouched is up to you, but you will not be judged on the relative security of your website. At least not yet.
- [32:49]
D
Yeah, all right.
- [32:50]
A
I'm excited if you have tips for us, if you have thoughts about what we should do. We are all vastly in over our heads on all of this. We're all being very vulnerable on this podcast as we go through this stuff, so help us out. Vergecast of the verge.com, call the hotline 866 verge11 Jake and Hayden, thank you for doing this with me. And as always, the best thing you can do to support everything that we're up to here at the Verge is to subscribe to the Verge theverge.com subscribe it gets you access to all of our podcasts ad free, including this one. It gets you all of our exclusive newsletters and it gets you some big stories. I've been writing about Vibe coding all of the other stuff Gen twohey like Vibe Coded her smart house. That was awesome. We have lots of great coverage. You can get all of it. Theverge.com subscribe thank you in advance. Vergecast is a Verge production and part of the Vox Media Podcast Network. The show is produced by Josh Kahas, Eric Gomez, Brandon Kiefer, Travis Larchuk and Aaron Locasio. We'll see you tomorrow. Rock and roll.
- [33:46]
H
When I scraped my car in that parking garage, I was worried that it could be a long process to take care of it. Like a landscaper's first day trimming a hedge maze.
- [33:55]
A
I have definitely already been here.
- [33:58]
E
Now, was it left right or right left?
- [34:02]
A
Well, maybe I'll cut a path out and find my way back later.
- [34:06]
H
But it wasn't like that. I filed a claim in under two minutes on the Geico app and they handled it from there. It was taken care of almost as quickly as it happened.
- [34:13]
D
It feels good to get help quick. It feels good to Geico this episode
- [34:18]
I
is brought to you by Google Chrome. You think you know a browser, but Gemini and Chrome? That's new. It can help you with practically anything on the web, like restoring a vintage motorcycle from a 50 page restoration block. Or finally break down that long article you've had open for weeks. Gemini and Chrome is here for it, ready to make anything online make sense. There's no place like Chrome. Check responses Setup required. Compatibility and availability varies 18.