Termai, SSTPhone, Young Men, and Tariffs
I don't know. I think podcasts are dumb.
Speaker 2:That what'd you say?
Speaker 1:I said I think podcasts are dumb.
Speaker 2:Hello.
Speaker 1:How you doing?
Speaker 2:I'm good. Sorry, I was actually late because I was on the phone with Kuti Tim, who actually is an early SST user. I don't know if you remember interacting with Kuti Tim in Discord.
Speaker 1:He actually helped fix some very serious performance issues in v two worked together a And I saw that he was working on like a two e that's like a cloud
Speaker 2:code That's what I was talking
Speaker 1:to him about. Nice.
Speaker 2:It's awesome. It's written in Go and Bubble Tea and I'm gonna work on it.
Speaker 1:Yeah. I had a feeling it was written in Go and Bubble Tea. I was Wait. Okay. Tell me more.
Speaker 1:Tell me more about because I I didn't look into it besides the screenshot, so tell me more about what it is.
Speaker 2:Well, it's a Cloud Code replacement. It works with other models. And well, what what like what do you want to know? Like, it's early.
Speaker 1:Is he taking a different approach to it? Like, is there anything that's
Speaker 2:Well, yeah. So he's he's a big Cloud Code user and it's clear talking through just we just had a twenty minute call where it's like, we have the same frustrations. I mean, one, just like knowing that you're stuck. Honestly, Cloud Code just kind of feels like a dead end to me because I just can't imagine I'm never going to want to play with other models in this life cycle of like frontier models leapfrogging each other. Like, I'm always going to feel like I'm behind.
Speaker 2:Yeah. Or I could be experiencing something better. I don't know, maybe that's just FOMO. I know
Speaker 1:I would say there's a good argument in favor of that because of the infrastructure behind these things. Right. In terms of like deploying these models, yeah, the big companies move slow, but we're seeing how when they do finally move, this stuff is really cheap. Yeah. You know, like like like the all the new Google stuff is like, just so much cheaper.
Speaker 1:So purely from a cost perspective, like, I doubt that Cloud Code is always gonna is ever gonna be like the most economical.
Speaker 2:Right. Yeah. So I guess like that's my starting assumption is Cloud Code is not my long term solution. Gonna want something open source that I can use with whatever model. Some of the annoying things like he he fixed I don't know if you've noticed in Cloud Code how much you're using it, but how often the tool fails to rewrite files because it says like the text is the same or something.
Speaker 1:Yeah. Yeah. Yeah. It's like we found multiple matches or like the Yeah. Text is the same.
Speaker 1:Yeah, exactly.
Speaker 2:The text is the same. Yeah. He fixed that bug, which is like super annoying to me. I'm doing such a terrible job because I'm feeling on the spot. I just talked to him and I'm so hyped because well, there's other things.
Speaker 2:There's other things.
Speaker 1:Okay.
Speaker 2:Wait for it. He's implemented a a very lightweight Vim experience, but I wanna I wanna like embed Neil Vim in there.
Speaker 1:Okay. So that I had the so that when I saw that, he said Vim like and I was like, man, I wish it was just Vim. And then I realized, oh, but then that require that would require the multiplexing stuff, which I guess is Yeah. A pain in the ass.
Speaker 2:Yeah, it is. And I'm gonna I don't know. I'm gonna dig around and see what's possible. I know I
Speaker 1:I think something simple would be to flip into full screen
Speaker 2:Yeah. When you're editing text.
Speaker 1:And like just just have that. Yeah. Like the multiplexing stuff shows up when you need to like embed a process inside
Speaker 2:Inside of a Like that, yeah.
Speaker 1:But, you know, it has like that alt screen concept. I don't know if you've come across that or
Speaker 2:What does?
Speaker 1:Yeah. So terminals technically have two, like, screens. They have, like, your main screen and an alt screen. Really? Yeah.
Speaker 1:And and you've seen, like, kind of when I'm trying think of an app that we all use that uses this feature. I can't think of anything off the bat. But basically, it's like it's like having two monitors effectively and you can flip to the alt screen and have something else running there. So you could have the input running there and then you just flip back and forth. So I don't know if that I don't know if the DX of that or the UX of that will will feel good, but it would be simple to implement.
Speaker 2:Yeah. Okay. Well, something like that sounds nice. I just it there's impossible to use a Vim like experience because I immediately try and do like any normal motion that I'm used to and it's like, no, that's not supported so
Speaker 1:Yeah. I've been using the other thing, an encode, an encode.
Speaker 2:It works? Does it work for
Speaker 1:you? Yeah, it does.
Speaker 2:I could I couldn't get it to like actually ever make a I'm
Speaker 1:using with Gemini.
Speaker 2:Really? Well, I wanna hear about Gemini. Let's we wanna we gotta
Speaker 1:get I haven't used it enough to have an opinion. I need to use it more but I think I'll just try so this thing that he's working on it, I'm assuming it's like model agnostic and you can plug whatever in. Yeah. You can use I wanna try it. As fast as these companies can move, these like like Clawd for example or Anthropic for example, my gut just tells me a random approach from some random person is gonna end up being better.
Speaker 1:I don't know why, I just feel like there's just always headwinds at at these larger organizations. Plus the open source part, think, will will help lot help a lot.
Speaker 2:Well, open source part for me, it's like, I know Kuti Tim, I like Kuti Tim, I can work on this too and it can be what I want it to be. Maybe that's not that's not best for, like, broad ambition and Kuti Tim can say, f off. No, thank you. But like, it's the promise, like the idea of like, oh, there's so many things I wish as a Cloud Code user that Cloud Code could do or didn't do. Yeah.
Speaker 2:He he said, so in terms of things he's doing that Cloud Code doesn't do, and maybe they'll add this, but he added LSP support. So somehow
Speaker 1:Oh, man. That's huge.
Speaker 2:It drops the LSP information into context. He said it makes a huge difference. And he's a big Cloud Code user.
Speaker 1:That's
Speaker 2:I worked with Cootie Tim. I don't know if you did I tell you I worked with Cootie Tim? I was like under Cootie Tim.
Speaker 1:I didn't know that.
Speaker 2:Yeah. So the the company after I left start or Statmuse in like 02/2019 whatever.
Speaker 3:Okay. After I was fired, whatever. I was fired.
Speaker 2:I don't know why.
Speaker 3:It's I just always say it that way.
Speaker 2:After I left literally Anyway, after I was fired at Satmuse, my startup, I joined a company called Brightcore, which is where Kuti Tim worked at the time. And it was like an insurance software company. I was there for like a year. I joined there randomly because one of my former Statme's employees was like a director of engineering at this startup. Much bigger company than Statenius.
Speaker 2:It was like, I don't know, 200 people? Yeah. Like a hundred engineers, something like that. It was a good experience for me just to like see how actual software companies run because I'd only ever worked for myself prior and since. So I have a little one year of experience, like, working in a it was like a safe organization.
Speaker 2:Yeah. Like, the ultimate in scrum memes. Yeah. Yeah. It was fun.
Speaker 2:It was Python too. It was just like all the things I hate,
Speaker 1:but That's funny.
Speaker 2:But I like the people a lot, including Kuti Tim. Anyway.
Speaker 1:Nice. I I
Speaker 2:feel like I had caffeine or something. I didn't. I'm just amped.
Speaker 1:It's just like, what a what a small world.
Speaker 2:Yeah. Right? Like, he's a NeoVim user. He uses bubble tea and writes Go and uses SST and knows you. Like, what Kudithim is basically me.
Speaker 2:I could just have him come on the show and fill
Speaker 1:in for You're like stuck writing Go and Bubble Tea even more now.
Speaker 2:Oh, yeah. I guess I do kind of like sometimes shit on it, but
Speaker 1:You just have to learn to like it. Well, it's
Speaker 2:it's not bubble tea. Like, the people at Charm are amazing. It's the elm thing I'm
Speaker 1:not super Yeah. I know. I just It just doesn't click for me either.
Speaker 2:It's not even like elm's fault. It's just my brain works some other way and it's like No, it's
Speaker 1:it's Elm's fault. Okay. I I would consider Elm being a failed experiment because a lot of people love it, but the component model is just is just better. The component model is just better and it's the reason React got so big, as big as it did. And you could preserve a lot of what's good about Elm without giving up the component model.
Speaker 1:It's an arbitrary like a random trade off we have to make, so
Speaker 2:Yeah. I guess you're right. It's Elm's fault because like who uses Elm today
Speaker 1:and React is like everywhere. Yeah. Anyway That's cool. I'm gonna I'm gonna I'm gonna try it. I I saw yeah.
Speaker 1:I saw a quick video the other day or whenever you posted it and I meant to check back but I I forgot. But, it's exciting. Yeah, these tools are are pretty cool. I feel stupid for not thinking about making them. Like forever, I was like, man, I wish I could have the cursor experience but not give up NeoVim and then my brain just stopped there.
Speaker 1:It was
Speaker 2:like Yeah.
Speaker 1:I guess that's impossible.
Speaker 2:Yeah. It was like a couple months ago on the podcast, you said something about like the file system as just the source of truth. Oh, that's something so Kuti Tim's working on. This is very exciting to me. Because it actually has like a SQLite database and it actually stores I
Speaker 1:saw the SQL thing there.
Speaker 2:Yeah. So his idea is to like and it feels like a thing Git should do, but I get where maybe it's better just to do it in the app. He's gonna like store file states so that you could always go back to any point in the message history and then branch off or just revert back to that state. You don't have to constantly like git add stuff just to stage it and keep it safe. That's what I do.
Speaker 2:Like, in Cloud Code, I'm constantly staging files just to make sure they don't
Speaker 1:get Yeah.
Speaker 2:Borked. So he's kind of building all that into the Term AI app where it keeps track of file states and you can just bounce back.
Speaker 1:Damn. This is gonna be I feel this is gonna be really successful.
Speaker 2:It's gonna be good, right? Yeah. And I'm gonna work on it, so it's gonna be awesome. I'm just kidding.
Speaker 1:Yeah. No, mean, that too.
Speaker 2:That's where it's gonna go down. I'm not making a good sales pitch
Speaker 1:I mean, you're gonna work on it and it's something you use, I feel like that's just gonna be a good outcome.
Speaker 2:I do feel like I've probably burned more Cloud Anthropic tokens than many single devs have. Yeah. Maybe more than some teams.
Speaker 1:We we got this grant, this credit grant 15 k, which we got in 10 k Who's we? SST?
Speaker 2:SST got a credit grant for who? Anthropic.
Speaker 1:It's a 15 k in Anthropic credits.
Speaker 2:Oh, nice. And
Speaker 1:we were like, okay, we're gonna use this because when we launch our hosted version of OpenControl, that's what, you know, like users will consume this. But we've just been consuming it. Like, I've just been consuming it to to build this stuff and I'm
Speaker 2:like Yeah.
Speaker 1:How much of this is gonna be used by the time we actually ship this? But like, whenever I look back at my bill, it's always lower than I expect. Yeah. Always like, I'm spending a thousand dollars on this but but I'm not.
Speaker 2:It's like in the context of what it unlocks and I think just developing these skills right now is worth a lot. Like, I think flexing this muscle even if the models are I feel like we're like one generation, one is that the name of the models? Is it a generation? If we go from 3.7 to four
Speaker 1:I guess that like okay.
Speaker 2:Yeah. We're like one step away from this being incredible. I mean, it's already good and I think in a good code base, it it really can like boost things quite a bit. We've talked about that. In not so good code bases, maybe not.
Speaker 2:Maybe the next generation is just so smart, still it helps you kinda like get code bases to the point that they're good. That would be great. Yeah. It's it's expensive, but it's I don't know.
Speaker 1:There's the
Speaker 2:caching The Bedrock prompt caching stuff. You gotta like get accepted into it, I guess. But it
Speaker 1:Prompt caching stuff is so confusing to me because it's like it's so convoluted. Not because it's bad design, it's just like the solution is a complicated thing. Because like with prompt caching, you pay more but future prompts that share the same prefix will be cheaper. So it's like very hard to intuit like You know way
Speaker 2:more than I do. I just thought prompt caching make cheaper. That's all
Speaker 1:I know. Yeah. Like, turn
Speaker 2:it on, get cheaper.
Speaker 1:It does. But it's one of those things where like well, for OpenAI, it's turned on by default because I don't really see a good reason not to turn it on. Unless like unless you're very sure that every single time you hit the API, it's a brand new prompt. Like, it's it's like you you ask one question, you get the response, you never like follow-up. Yeah.
Speaker 1:Then prompt caching doesn't make sense. But yeah, like it makes sense to be turned on by default. The thing that's funny to me is when I look at the API, like in my head, when I, you know, saw these apps being built, I'm like, oh, you probably like There's an API call to, like, start a session and then you, like, append messages to the session. Like, every time you send API request, it's just a new message. No.
Speaker 1:Like, you send the whole context back The whole thing. Yeah. This is really inefficient, but I guess it's fine. But, like, yeah, it feels weird that you just send the whole thing back and forth.
Speaker 2:I I have a theory on that and it ties into Gemini and all kinds of stuff. But but just before I get to that, it kind of tie put a bow on Kuti Tim and Term AI. Kuti Tim, I know you're gonna listen to this because I'm gonna send it to you, and I'm sure you're gonna wanna hear what I had to say. I'm sorry for all of it. That was just that was pretty awful.
Speaker 2:You probably don't want me involved at this point, like, don't want me pushing PRs. I'm still going to. You can just shoot them down kindly as I know
Speaker 1:you will.
Speaker 2:Anyway, okay. Moving to the so this idea that Claw three seven is kind of the only model right now that can handle these giant contexts. Not that it can It's the only model that can handle giant context. I know Gemini has a bigger context window. I mean, like, it seems to be the only one that keeps trying Yeah.
Speaker 2:With all the tool calling It's persistent.
Speaker 1:And all
Speaker 2:the turn based. Yeah. The persistence. Is that also the reason people complain about three seven? Is that the same property that makes it do too much?
Speaker 2:People talk about it like it goes crazy and does too much in its actual, like, changes to your code base?
Speaker 1:It's possible, but to be honest, I my experience of it has been very good. So I don't when people say they're, like, not happy with it, I don't really know what they're talking about.
Speaker 2:I honestly I don't actually know either. I've not experienced it doing too much, but you hear people say, like, it created whole new some things and implemented third party libraries and did all this stuff like
Speaker 1:Yeah. But there was definitely a period of me refining it with like some of the rules I wrote up and cleaning up some stuff in my code base that it was getting confused by. But Yeah. I do stuff that's not mainstream. Like, I don't use Tailwind and like the way we use CSS is a little weird.
Speaker 1:And despite that, it like picked it up and it's I actually did something the other day which is pretty interesting where, you know, I mentioned to you like, in a beginning of ZR, I switched to using normal CSS and like no tooling. And it took me a few iterations of that to get my patterns right. So I had some older stuff that was in like my initial experimentation.
Speaker 2:Mhmm.
Speaker 1:So I was like, I need to clean this up because Cloud Code is getting confused by this older stuff. But then I just did that by telling Cloud Code, hey, like, look at this file. This is the pattern that I like. And then go like go back to this other file and and fix it. Yeah.
Speaker 1:And it did it It's so good at
Speaker 2:that stuff. It's so good at that kind of thing. Or like removing stuff. If you're like, we're done with this feature or we we're we're done with this concept and just removing it from the code base is so it's so good at that. Like Yeah.
Speaker 2:It can do its whole grip searching and then start just cutting stuff out like a pro. It's stuff that's really annoying to do, you know? Like, that's like the kind of programming task that's not a fun not a fun task.
Speaker 1:It's been a funny adjustment because sometimes when I'm working, I will actually and there's like a tedious task that I need to go rename something in like 40 files or like clean up type errors in 40 files.
Speaker 2:Mhmm.
Speaker 1:I have this thing where I kinda like doing that because like it would be like my break where I would just kinda like You're just checking and just like really focus on whatever I'm watching and kind of like, you know, do do this. Mhmm. But now, like, yeah, just like I never really should should be doing that. And technically, don't mind. Like, now that I have now I have the option of removing it, actually, it's much better.
Speaker 1:But I wouldn't say like I a % hated it. I like I kinda liked it to some degree.
Speaker 2:Yeah. Yeah. I I can see that. I mean, my brain's mostly off when I program all the time, as
Speaker 1:you can tell. Well, now I just get up and I go do something else, so Yeah. Well, it's doing
Speaker 2:it Or or
Speaker 1:I work on something else, yeah.
Speaker 2:Man, the future is bright. I am very excited about programming as an occupation in the twilight of my career. Do you feel like you're into the twilight of your career or you feel like you're in the prime? Are you in the you're probably in the prime. You're like way younger.
Speaker 1:I'm not way younger. I'm like, what's four than me.
Speaker 2:Is it only
Speaker 1:four years?
Speaker 2:I'm 38.
Speaker 1:I'm 33.
Speaker 2:Five years? Five years. Okay.
Speaker 1:I don't know. I don't think I'm ever gonna feel like I'm in the toilet of my career. I just feel like Well,
Speaker 2:I mean, none of us are retiring. Who knows where the economy is headed? The world's going to shit. I'm sorry.
Speaker 1:I did say something to I think it was Jay the other day where I was like, man, vibe coding feels is like too fun. Like I this is like way too pleasant. Like this just can't get there's no way this can continue to have like economic value because it's like too enjoyable. Because I literally just because you know, I always watch TV. I always have TV on the background while I work anyway, so it's the thing I've been doing for But now, I'm just literally just like, hey, computer, do this.
Speaker 1:And just sit and I watch TV and then and then it just doesn't work. And then I'm like, okay, computer, do this. I'm like, I'm making how much money to do to do this? Like, this this doesn't make any sense at all. This can't like the world can't work this way.
Speaker 2:When you use the term vibe coding, do you do you mean that in like some specific definition of vibe coding? I feel like some people, there's like a very specific I
Speaker 1:think when I use that phrase, it's when I'm like doing the most extreme version of this where I am building something new or experimental or like Mhmm. Building out like a new screen where I'm just like not at all trying to actively do any work. I'm just trying to get it to do everything. So that to me is like the epitome of vibe coding. Most of the time, that's not what I'm doing, it's more like intense where I like have a task, I find a way to split it up into two ways and I give it half and I do the other half.
Speaker 2:Yeah. And then
Speaker 1:and then like the moment I'm done, check to see if it's done and it's kind of like very intense thing.
Speaker 2:Mhmm.
Speaker 1:That does not feel good. It's very productive but like it's not like a pleasant feeling. It's Yeah.
Speaker 2:Yeah. Agreed. I like I still haven't gotten I I wonder if there's a point where the models are good enough that I turn off my OCD and I don't read every line of generated code like
Speaker 1:I don't read it.
Speaker 2:You don't read it. Oh, really? Well, that's how we ended up with a somehow an incorrect build amount on a terminal shop bag. It was like $10 extra for no reason.
Speaker 1:On Raycast? The Raycast extension?
Speaker 2:Yeah. It was like the Raycast thing. Right? But not sure it difference? Oh, that was that that was you?
Speaker 1:That was not me. It was because our API was unclear for a little bit because we had some we had no other field in there.
Speaker 2:Gotcha. Well, that's probably my fault then. I'm kinda on the API side. My bad. It reminded me of the terminal feud.
Speaker 2:Yeah. Yeah.
Speaker 1:Scoring thing. Exactly. I really don't wanna fix that because it it was so perfect. Like, that amount of error was so perfect.
Speaker 2:I know. Yeah. It wasn't off by one, it was like off by two.
Speaker 1:I like stumbled on the perfect algorithm for comedy. The perfect way to add numbers where it was hilarious.
Speaker 2:That's funny. Well, I still read every line. I generally go through and like make changes even and that's probably slowing me down a lot, but I don't know, it makes me feel more comfortable about what I'm putting out there. I don't know. It's just an OCD thing probably.
Speaker 1:Yeah. Well, so I'll say two things. The first thing is when I say unpleasant, I actually don't mean that I feel the tool is not good enough. I just mean the process is not enjoyable. And I don't know if we talked about this already, but it really feels like this idea of sitting down and doing focused work where you're in the flow is just history.
Speaker 1:So for me, for most of my career, that's kind of what I had. Then I started doing CloudFormation stuff and I was like,
Speaker 2:oh, okay, gotta wait for CloudFormation,
Speaker 1:then let me go work on something else. Oh shit, CloudFormation is probably done, let me come back. Okay, now So that whole thing, it just feels like that I'm working on five things at once and juggling them all.
Speaker 2:Mhmm.
Speaker 1:It's just getting Work is just looking more and more and more like that.
Speaker 2:Yeah.
Speaker 1:So that that's why to me it feels a little unpleasant. I think I can get used to it but
Speaker 2:I do prefer the single track, Like, I'm in the zone. I got music going. I'm like It's
Speaker 1:been a very long time since my work felt anything like that. Like, I'll get that like once a month maybe.
Speaker 2:I guess I still get it on occasion. I mean, I think there's there's still I do a lot of front end stuff and I guess you get kind of insulated from a lot of the nonsense.
Speaker 1:So the second thing I was gonna say is the thing you're talking about. So when I say vibe coding is really pleasant, a surprising thing is Liz finds it very unpleasant. Don't you think that
Speaker 2:I saw her tweet. I didn't know how much truth there was there.
Speaker 1:Yeah. So she so you would think, okay, she's now new to this, so you would expect this thing gives her way more capability. Even just stuff straight up like she's just not as good at using her editor as I am. So you would think this thing is like so much more useful
Speaker 2:for Oh, yeah.
Speaker 1:And it is, but it triggers her OCD a lot too where it's almost magnified because free so, like, it does something wrong and weird and you wanna go clean it up, right? Mhmm. Mhmm. And she feels the same way but it's magnified for her because when it does something weird, she has less experience to be to be even to be able to understand what it even did. So, to her, if it does something that she hasn't seen before, it feels like a hundred times worse than what you feel which you feel you're like, oh, I get what it did but it's like kind of messy so I'm gonna clean it up.
Speaker 1:For her, she's just like, what the fuck is this? Like, now now what do I do, right? So, yeah, for her she like I It's been like a bunch of back and forth for me to like get her to use it more. Mhmm. And I think she understands that it's practically useful for, you know, for all the practical reasons.
Speaker 1:But she's She finds it a much more frustrating experience than than I
Speaker 2:do. Interesting. Yeah. Yeah. Because that's like what the whole vibe coded thing is like targeted at, I feel like, is people heard Yeah.
Speaker 1:So you if you don't know yeah. Yeah. But I mean, guess some people just don't even look at this stuff. And I'm kinda trying to encourage her like, just don't look at it because I'm gonna have to clean it. I'm gonna clean up.
Speaker 1:Once you're in a good spot, I'm gonna clean it up anyway. Whether you hand wrote that or whether you
Speaker 2:Mhmm.
Speaker 1:You know, did it to this, I'm still gonna go through that pass and don't worry about it too much. But, yeah, like it it it's just tough because I think for a lot of people it goes against their instincts. And there's also this dynamic of when you're new, you also like don't know exactly what's good. Like you might have a sense of like what's good and what's bad but we've all been through that phase where we're like overly try to clean up stuff and it's actually just making it worse.
Speaker 2:Yeah, For sure. Yeah. I don't know I don't know where where it's headed in terms of like my confidence in just the not looking thing. It's it's hard for me to to just like not look at the code. I've just seen it do too many weird things like in new code bases and make those silly mistakes where you're like, actually this.
Speaker 2:And it's like, oh, I'm so sorry. And it just goes through that loop and you do
Speaker 1:like six iterations of that and still won't fix it.
Speaker 2:Yeah. That kind of stuff just makes it where like the terminal code base is a great example where it's like there's enough consistent patterns that it just seems to knock stuff out when you like give it something, it can follow the pattern and do the thing.
Speaker 1:And that's kind of why it took us so long to even like get to this point where we're really enjoying it because most of our work is on SST, which is like a weird Go CLI plus like really weird It doesn't look anything like a normal app.
Speaker 2:I've heard framework authors in general have a hard time this this Yeah.
Speaker 1:It just can't help you. Like, it's not Yeah. It's like so much of your work is like decision making and not like execution. So it feels very good to switch to like very normal looking work and Mhmm. That's where it it really shines.
Speaker 1:Oh, I was gonna tell you. So I I use it in a brand new way yesterday which is really exciting for me. I've had this PR because we need to support those for SST, finally getting around to doing that. And there was mean need to
Speaker 2:support what?
Speaker 1:Windows?
Speaker 2:Those? Oh, Windows. Yeah. Like what? Oh, like the operating system.
Speaker 1:The operating system, yeah.
Speaker 2:Not like a TMux concept or something.
Speaker 1:No, no. Windows Microsoft Windows. Yeah. And someone put in a lot of work to do a PR that they they just went through and just kept hitting every issue and then like just figuring out a fix for it.
Speaker 2:Wait, I'm sorry. I'm so dumb. SSD doesn't work on Windows today?
Speaker 1:No. It has Oh, I
Speaker 2:didn't know this. But I guess like what about WSL or that kind of stuff?
Speaker 1:It does work on WSL. I think there's like it's some people have issues with it. Don't know. It just it just it definitely works on WSL. It definitely does not work on Windows at all.
Speaker 2:And SSE, it's a Go app, right? Yeah. Does it you don't just like compile it to Windows or something?
Speaker 1:Yes. But they're just little details that are different that can't be abstracted away. Someone like went through and just kept hitting hitting those. Yeah. Okay.
Speaker 1:Continue. And they they kept patching things here and there. I mean, it's really stupid things where like when I spawn Pulumi, if I'm Windows, I need to add a dot exe, you know. It's just like it's just like things like that you need to be aware of.
Speaker 2:Interesting.
Speaker 1:And this is a big PR, so I had it. So I checked out the PR and I told Cloud Code, hey, compare this branch to the main branch.
Speaker 2:Mhmm.
Speaker 1:Go through each, like, hunk of changes one by one. Describe to me what the change is and why it's there. And let's go through it one by one. Mhmm. And it literally did that.
Speaker 1:It was like, in this file, you know, we changed this because on Windows when you close a file, it's not guaranteed to like, like figured out why the change was made. So, I was able to review this PR, in detail and feel confident that, okay, now I understand what it's doing. And I'm horrible at PR reviews otherwise, I just hate doing them.
Speaker 2:And it's like such a chore. Hate.
Speaker 1:Yeah. Yeah. So this was like a really pleasant way to to do a PR review where it just tell showing me one by one. Look at this, this is good. Look at this, this is look good.
Speaker 1:This is probably why they made the change. This is exactly what was changed. Mhmm. And having that side by side with the actual diff, very very good experience.
Speaker 2:So Clogco did a good job with that. Yeah. Because I've doesn't GitHub have like a a feature for like AI code review? Maybe like a experimental thing, but
Speaker 1:I Yeah.
Speaker 2:Think GitHub has that. And I wonder if anyone's tried it.
Speaker 1:I saw something with Copilot.
Speaker 2:What what is Copilot?
Speaker 1:Okay. I gotta I gotta ask this because I get I've
Speaker 2:gotten so confused. Copilot used to be like line by Just
Speaker 1:just the tab complete. Yeah.
Speaker 2:Yeah. And then it became like a broader thing at Microsoft. Like, maybe maybe all of Microsoft is Copilot like Microsoft now. Because I feel like
Speaker 1:Yeah.
Speaker 2:It became this giant thing and I don't know what it is anymore.
Speaker 1:It's not just for code even anymore. It's just It's kinda like what IBM Watson did. Do you remember when IBM Watson was like a Oh, yeah. Physical computer, the Jeopardy computer? Uh-huh.
Speaker 1:And then it was so popular that they just put They just made out the name of their like AI branch and then everything learned
Speaker 2:like AI initiative. So Copilot is just like go it's GitHub or Microsoft's entire AI initiative is just Copilot?
Speaker 1:Yeah. Think so. I think that's how works. I saw it I think they even they even called just like the non coding one just that's built into Windows and they just called that Copilot.
Speaker 2:Yeah. Okay. Little tip for branding people. Copilot just means nothing to me now. Like, I just hear the word and I check out.
Speaker 2:Like, I don't Yeah. It doesn't have any I can't tie it to a thing. And that Yeah. That's a problem for me, like, as someone who's not immersed into products. It makes me not wanna dive into that world because it just seems like this big amorphous unknown.
Speaker 2:I don't know.
Speaker 1:Yeah. I mean
Speaker 2:I have like an aversion to it.
Speaker 1:This is a whole big question of Microsoft obviously has all the properties that are relevant and all the reach that's relevant and like the best sales team on the planet by far. Yeah. And this is like a massive, massive opportunity. And it's also in a space where they have historic like Microsoft and developer tools.
Speaker 2:Oh, yeah. Developers developers. Developers.
Speaker 1:Yeah, exactly. Been their thing forever. And I guess where a lot of money they've made here. How is it plausible at all that they're gonna miss out on this?
Speaker 2:Are they are they completely out of it if they're not like are they I don't know what the relationship with OpenAI is anymore. They they're kind of separated. There's a spat. Yeah.
Speaker 1:I mean, they they did say they're like working on their own stuff and
Speaker 2:So they don't have a thing outside of the OpenAI investment? They don't have like their own models?
Speaker 1:They they they are working on their own models. And I also think they technically don't they have some hold into Anthropic? I forgot what it was. Microsoft? They're like a little diversified
Speaker 2:I know AWS does.
Speaker 1:Even there. Or maybe it was Google that has something in Anthropic. I don't know. Anyway, point being, I don't think that it feels so easy to look at the landscape now and be like, yeah, Microsoft is like not anywhere because there's Cursor. Yeah, there's a bunch of stuff in GitHub, but GitHub just increasingly becomes like a product that feels like bad to use and like Mhmm.
Speaker 1:Why are they so slow with all this? What I really think it's I still think they're just gonna win. Like Well, that's how yeah. That's how I feel about Google, I guess. Like Yeah.
Speaker 1:Same.
Speaker 2:It's the same kind of situation where it doesn't matter how behind they are. It feels like they have the infrastructure, the bones to win in the end. But yeah. Microsoft, I guess Copilot, I guess GitHub and Copilot, that is their current like beachhead in the AI game. But like they don't have a big frontier model.
Speaker 2:That's kind of crazy actually.
Speaker 1:But I think they do. Like I
Speaker 2:Oh, they do? I just never heard of I think
Speaker 1:it's not like Is
Speaker 2:it called Azure
Speaker 1:or I've seen talk about it internally and or like talk about having it internally and stuff. And to me, this stuff isn't like difficult to acquire at this point. Like, if random labs in China are like doing
Speaker 2:it,
Speaker 1:Microsoft can do it given it's like Yeah. Almost purely a capital problem from one direction.
Speaker 2:Apple can't do it, that's for sure. Jeez.
Speaker 1:They they can do it too. They just I don't know. It's just Well,
Speaker 2:they're trying and they're not. Right? I mean, they're just bad at it.
Speaker 1:They're to integrate They're trying to build a good product, which they are failing on, which nobody There's no guarantee if we can make a good product, for sure, that if that's a problem.
Speaker 2:I want Apple to do it so bad because I just I want the like I want it on my phone. I want it just embedded into everything. Yeah. Is that bad to want
Speaker 1:Did you see my post about saying ST is gonna make a phone?
Speaker 2:Oh, I did. What was that about? I wanted to ask you about that. Like, I usually can understand your jokes. I did not get that one.
Speaker 1:It wasn't a joke. No. A hundred and serious.
Speaker 2:Okay. Why not?
Speaker 1:Well, I followed up saying this probably still won't happen but every day it's getting like two times as more likely.
Speaker 2:Okay. Yeah. You gotta explain this one.
Speaker 1:Well, we're like looking at all this stuff and it's not ridiculous to say there's maybe maybe we're gonna look back on this in ten years and be like, oh, there was like a new shift in computing that that happened during this time. Mhmm. What could that So then we try to imagine like what what could that be?
Speaker 2:Like from the ground up rethinking the phone.
Speaker 1:Exactly. Ground up rethink rethink what I mean, phone even even that word is like up up for question.
Speaker 2:Just the word Yeah, throw that away. Just small form factor thing you have in your pocket all
Speaker 1:the time. Yeah. Or like just like And so Jay's observation was So we had a few different observations. Jay's observation was these things are so much more powerful than we're actually using them for because we're bandwidth constrained. Like, all everything that you just talked about that you and like working out What what's what's the thing called?
Speaker 1:Term UI? Term AI?
Speaker 2:Term AI.
Speaker 1:Mhmm. That's just improving how much stuff you can shove into the LLM and like give it the right stuff, right? It's not about the LLM itself becoming better, it's like, it's already really good, it's just not being given the right And that's caught at
Speaker 2:right time. Mhmm.
Speaker 1:So, so many of the so many of the bottlenecks are just like Yeah. The AI If the AI could literally see everything you're seeing, it can do a lot of crazy stuff. It just can't do that. So, it's like bandwidth constraints.
Speaker 2:You're gonna make a pendant. SSC pendant. You're gonna take on friend. Is it friend? Is that the name of the one?
Speaker 1:Yeah. Well, there's there's there's like there's like five of them. Think they're they're all stupid.
Speaker 2:They're all stupid. But it could work for us. No. I'm just
Speaker 1:I don't think we're literally gonna try to solve that like, you can see whatever you can see problem but it demonstrates this like bandwidth constraint thing. Yeah. That that's the issue. Mhmm. So if you accept that, then you have to assume that some new form factor might make sense to enable to enable all this.
Speaker 1:Mhmm. So yeah, the question is like, you know, what could that be? What could that look like? That's one observation. The second one is what you're talking about which is forever not forever, but for a while, both iOS and Android have encouraged app users to like model their app as a set of actions so that they can compose nicer together.
Speaker 1:Like Mhmm. One app can trigger an action in another app.
Speaker 2:Mhmm.
Speaker 1:And it's so clear that that model is perfect for AI because AI can just compose these actions together like perfectly. Mhmm. And I think Apple kind of demoed something like that where you can ask it to do something, some like complex set of things and it can just like trigger the actions in different apps. And this got and so obviously, like, that's what you're saying, like, you want you want that, you want that on your phone, you want that you want that everywhere, you want to be like, hey, like, you know, I have a meeting later, like, give me directions to it. Like Yeah.
Speaker 1:Yeah, Google Calendar can directly integrate that feature or it can just look up your meeting and then call Google Maps and then like do a bunch of stuff. Bad example, but like, you know, you can imagine some interesting things. So that got me thinking around something else, which is there's no angle to do this. No one's gonna switch off their iPhone. Apple has to do this, Google has to do this.
Speaker 1:This feature on its own is not enough, as compelling and amazing as it'll be. They just have so many hooks in everywhere that this is not gonna be the thing that makes people switch. So how do you even like try to disrupt the space? We don't have a lot of good ideas or any good ideas but the one idea that I had was there's also this growing desire for people and it's small and it's niche and this might not be big enough for people to say
Speaker 2:Oh, I think I know where you're going.
Speaker 1:I don't want this phone. Like Yes. This thing is just fucking like killing my
Speaker 2:want a dumb phone. I don't
Speaker 1:want a dumb phone.
Speaker 2:Yeah. That's a movement.
Speaker 1:Mhmm. And it's increasingly growing. There's been a few products that have shipped, but the products always look like this. They're like, they they look at the phone and they're like, okay, these five features are good and not distracting and we're gonna put that on the phone and we're gonna like not include everything else, right? Mhmm.
Speaker 1:So they're gonna be like, okay, you can get text, you can call, but that's it. But then you're like, well, I kind of want to listen to music, but they don't. They decided that, you know, music is not something you should listen to. Mhmm. And it's so it ultimately kind of ends up falling short because I don't think a general approach like that of saying these are the only five apps everyone needs is that compelling at the end of the day.
Speaker 1:But I think AI offers something kind of interesting which is there's two ways to use your phone. You're like, I need to do something and I'm gonna use my phone to execute it. Or I'm sitting on my toilet and I'm bored and I'm gonna like just my phone's gonna help me like discover stuff, right? Yeah. That latter thing is a thing that we're trying to give up.
Speaker 1:Yeah. And the every single app is optimized for that. If so we say like, oh, we wanna get social media off our phone, but there's technically nothing wrong for me at least, for me being like, I have a thought, I wanna post a tweet. Yeah. The problem is for me to do that, I need to open this app that is hyper optimized to show me a ton of stuff, prevent me from leaving just so I can send this.
Speaker 1:It's gonna show me my little notification thing, it's gonna show me my feed, it's gonna do all this stuff and all that is gonna grab my attention distract
Speaker 2:it. Mhmm.
Speaker 1:So that first category, I just wanna do something, I'm trying to execute on it, AI can do this, right? I can tell it what I wanna do, I can just trigger that specific action and not show me anything like super compelling. I can control the OS can control that like end experience. So I feel like that's maybe a good hook if you're like, is there a pool of people that are like, I am just sick of my iPhone. Like I just don't want it anymore?
Speaker 2:Mhmm.
Speaker 1:They're kind of already willing to move to something like, can AI help you reimagine the device that what that looks like and it's just this boring thing not meant to simulate? It's not gonna be useful on the toilet Yep. But when you're on the go and you need to do stuff, it is it is a perfect thing.
Speaker 2:So when you yeah, when you first when when I saw the tweet, when you brought this back up, I'm thinking you're trying to like make a phone like a phone phone and it's like, that's ridiculous. You're not gonna compete with Apple. But the I could imagine you guys actually making one of these in the class of these kind of like dumbed down phones because they're they're not like crazy hardware engineering. Right? It's not the Apple like, it's a very different device.
Speaker 1:We'd probably want to do something interesting there too, but
Speaker 2:Well, you do it you do it well, but like
Speaker 1:Yeah. I'm just saying You you can white label a phone today, like very easily. There's a bunch of Chinese manufacturers that just let you totally white label a phone and make it feel like it's yours. Yeah.
Speaker 2:If it's coming from China, good
Speaker 1:luck. Yeah, 50% more expensive but
Speaker 2:Yeah, how about tariffs at some point? I I just don't understand them and I'm afraid that our government doesn't either and that's scary.
Speaker 1:Yeah. Before we talk about that though, one last thing. So one of the piece of inspiration for this was the Daylight computer. I don't know if you saw that.
Speaker 2:No.
Speaker 1:So it's a tablet that's an e ink tablet. You must have come across it because it's going viral for a bit. Oh,
Speaker 2:I do remember. Was like very exclusive and hard to get Cause I always
Speaker 1:want things like that. I think it's kind of limited. So it's an e ink tablet. It's it's it's like kind of bulky, a little bulky, but crazy battery life obviously, but really responsive. So unlike like a Kindle, it's like you can do all kinds of interesting stuff in it.
Speaker 1:And I think that's a great example of how you need to approach Mhmm. If you do wanna disrupt this crazy like space Mhmm. You need to come up with like a completely different angle that people would get interested in. So a bunch of people are like, oh, this is a new form factor I haven't seen, I like it. If they get enough people to do that, they could do what I'm talking about, right?
Speaker 1:They can now explore like these like what does an OS in an AI world look like. But like that is not enough. You can't just say I'm gonna build OS with AI. You need actually Yeah. Invent something else that is compelling to be able to do to do the thing you kind of want to do.
Speaker 1:So very hard.
Speaker 2:I'm on board now. I want you guys to do it. I think you'll you'll do it well. I don't know how you're gonna do it. Under the SST branding, I mean, I just don't know like is it a
Speaker 1:different No, we'll start a new company and raise like a shit ton of money or something. Yeah.
Speaker 2:Yeah. Yeah. Okay. Cool. Well, you guys should do it.
Speaker 2:You're good at executing and this doesn't need to exist. You've convinced me. I'm sold.
Speaker 1:Yeah. It's it's funny I'm easily sold. If we take a bunch of steps back, where we're starting is we're gonna just try to make the UX of OpenControl really weird.
Speaker 2:Oh, okay.
Speaker 1:So when we do this hosted thing, we're gonna try to like experiment with all kinds of ideas to see if we can build something in like a very limited sandbox that kind of steps in this in this direction. We have some kind of conviction and it's like maybe everything should work this way.
Speaker 2:I saw your tweet about trying things in the open control UI. I didn't understand what you tried. It wasn't clear to me what was different about that Like, what was
Speaker 1:the That was just that was just an aesthetic thing. Like, I I that Just like
Speaker 2:the terminal UI kind of look?
Speaker 1:Yeah. Like, that that video probably wasn't a good example, but I'm gonna I'll post another one with like, our dialogues like look really crazy and weird. Really? It's just one of those things where you typically wouldn't try a crazier UI because you have to hand make it and that's a lot of work. Yeah.
Speaker 1:So if it ends up not working, you can't get to like start over which is a lot of work. But now, we can take all that and just kind of like vibe coded. Yeah. So, yeah, Jay built like He's an experiment. Very retro like dot MS DOS feeling like dialogue thing.
Speaker 1:Okay. I think I don't know if I like it or hate it. It's confusing because like I like I get to talk about it a lot. I'm very sick of this black and white UI stuff. I think it's very cool.
Speaker 1:Can't But it like kind of hurts my eyes at the same time, so yeah, I don't know. I'm very confused about it.
Speaker 2:I just I opened Twitter, which is a mistake. Do we just we literally just talked
Speaker 1:about But I What is
Speaker 2:it what could
Speaker 1:have been like? You could if you could have said precisely what you wanted and only pulled that That
Speaker 2:would have been nice. Yeah. Get on it. Give me a new phone. But I happened to hit one of your tweets, which is often because you tweet constantly.
Speaker 1:Like, you
Speaker 2:tweeted two tweets an hour ago. So at the same time, what are you what are you, an animal? 08:46AM and 08:29. You tweeted twice in an hour? Okay.
Speaker 2:Whatever. Your top tweet is about eight or could you tell me is eight or good? It's Python, right?
Speaker 1:No. Was gonna it because my top tweet is about how people need to stop making stuff in Python because I tried to install it and said 329 dependencies.
Speaker 2:I'm so tired of this. I feel like Python is growing in a weird, like, ass backwards way where why is it it's growing because ML researchers, that's all they know? Like, they don't know anything about Wait.
Speaker 1:Saw your post about the software engineering bench is only Python. Is that true?
Speaker 2:It's true. SWE bench. The one bench that everybody talks about for like software programming capabilities. Now, I know these frontier companies are like training it on data from all the languages. Clearly, it can write other languages.
Speaker 2:But the one metric they always point to that's like, this is the best coding model, it's SWE bench and it's all Python. SWE bench is all Python. That's crazy to me.
Speaker 1:Yeah. That is really stupid. I did not realize that.
Speaker 2:And then you just keep seeing people come out with new tools written in Python. Where did this come from? We need to kill Python. So tired of installing Python or different Python versions or pyenv or VirtualPyIns or It's like there's so many stupid little layers to that. I just hate
Speaker 1:It's funny because because UV actually works really well and it fixed everything but I can never tell if I can use it with a project. So then I have to like then figure out like, can UV work with it? Like if I don't see it if the default instruction isn't like use UV, I just assume it's a clusterfuck. So even the fact that UV exists didn't really help.
Speaker 2:Yeah. That's like Neovim plugins now for me. If it doesn't say LazyVim at the top, then I'm like, probably not a good plugin. That works.
Speaker 1:And it what kills me is on Arch, it install it tries to install all dependencies system wide for these things. I I don't know why this is a pattern. Like, when I install Cloud Code through the Arch Package Manager, it's just like NPM installing anything and like has a scoped node modules for it. Yes, that results in duplication, but at least random other shit is like OBS stops opening up because I installed Ader chat or something, right? This is my experience with Python where random shit on my system will break because they're like, oh, totally, we can just have one version of dependency for the whole system, that's not gonna be a problem.
Speaker 1:Yeah. I don't get why that's like a pattern in this world. So just makes me so angry.
Speaker 2:It's so funny that that you ran the same thing because I everyone keeps saying Ader, and I went to the center. Was like, nope, Python. I didn't even try. I was just like, it's Python. I'm sorry.
Speaker 2:I'm not gonna
Speaker 1:do it. It's just not gonna go through
Speaker 2:the pain.
Speaker 1:I always give it a shot. I'm always like, maybe someone in the that someone packaged it as like a standalone nice thing and it's always the the the So if you watch that video, it's actually extra funny because it's like 300 something dependencies. A third of them are Haskell Haskell dependencies and like a fourth of them are like Ruby dependencies. What? So I'm just like, this is just This not right.
Speaker 1:This is not right. Whatever's going on here is not right.
Speaker 2:Well, it's not right. I don't understand how Python and Haskell interop but that's not right.
Speaker 1:Yeah.
Speaker 2:Oh, two things I wanna still talk about. We gotta talk about the tariffs just because we can do that at the end. There's like a I listened to the Bill Simmons podcast. Mhmm. During the NFL season, he has Cousin Sal on, and they talk about NFL stuff.
Speaker 2:And they always do this segment at the end of the show called oh, I'm blanking on it. I always skip it. That's the whole point of this little segue. Parent corner. They do this parent corner thing.
Speaker 2:Like, they talk about football for, like, forty five minutes, and they spend fifteen minutes talking about things that happen in their family, and I just always skip it. That's how I think of like putting world events at the end of our podcast. It's like if people just wanna skip that part, just skip it. Like, it's fine. Care about tariffs?
Speaker 1:But while you're talking about it, just felt like checking my portfolio, and it is looking awesome.
Speaker 2:Yeah. We gotta we gotta go on the tariff thing. Before we deal on the tariff thing, I I listened to this podcast that Casey sent me and I'm blanking. It's Diary of a CEO. You've seen him maybe?
Speaker 1:I don't know. I think podcasts are dumb.
Speaker 2:Di what'd you say?
Speaker 1:I said, I think podcasts are dumb.
Speaker 2:Well, I don't know anything about Diary of the CEO, but it was this episode he had these people on talking about, have you heard some of the statistics on men in the younger generation? Like, there's a young man problem
Speaker 1:Yeah. In the world.
Speaker 2:For sure. You heard
Speaker 1:I mean, I've heard the little statistics, but to me it's like the statistics must be crazy because it's like visible without even looking at the statistics. Like, I can see it in the real world.
Speaker 2:And I just had brushed off so much of the like people just on Twitter, it's so flippantly call people incels. Yeah. And I didn't even know what that word meant, but it was just kind of like, yeah, I guess like just annoying guys. But it's like, it's a serious, serious problem. And the reason I'm bringing it up is like, feel like as millennial men, I think we do have like a modern take on men, manhood, masculinity.
Speaker 2:I hate the word masculinity. It just makes me think of podcasts that are like, masculine man. Like, how to be a man and I just hate that stuff. But I do think like our flavor of masculinity of being a man seems better to me than like our older generations. We're much more like hardened and cruel.
Speaker 2:I mean, see Trump. Mhmm. Like, think we we're different as a as a generation of men. But like, are we not passing it down or something? What's going on that I mean, I guess a lot of it is like single family homes, no male role models, all this that they talk about in the podcast episode.
Speaker 2:How do we like, feel like we as men, you and I, we're the men. Remember the elevator thing where
Speaker 1:Yeah. I'm the man. Yeah. Yes.
Speaker 2:We we gotta be role models. And I don't know how to be a role model. I don't feel like I could ever be a man role model, but I feel like we're men and adult men and I kinda stop saying the word man. I don't know.
Speaker 1:I have my take on this is gonna make you laugh because it's so you're never gonna expect what I'm gonna say.
Speaker 2:Okay. Perfect.
Speaker 1:But me and Liz talk about this all the time. I think the problem is people aren't horny anymore. I think that's a root that's the root issue that we I think we
Speaker 2:talk about in the podcast.
Speaker 1:We live in a sexless society and I and I'll kind of explain what the main observation I make that is so different from how I was in my twenties and as like a teenager. Yeah. There's so much debate and like criticism of like from men and women about each other these days where men are like women are like, they're like always doing stuff like this and that's like so annoying and like And then women are like, well, like men are like, there's all just bitching at each Yeah.
Speaker 2:There's a lot
Speaker 1:of content at age. I would just agree with anything any girl said because like all I cared about, like there was one thing I cared about. Yeah. And I just I just can't imagine like being like, oh, I wish you would be different, you know? To me, it's just like Yeah.
Speaker 1:Because like that and to me that's so unnatural to like let something overpower that. So I think that's just like the root thing at the bottom of all this where what worked is what what created, I guess, like peace was that dynamic of being like, well, yeah, like the like, we want to sleep with each other so we're going to like not be so annoying about technically superficial things. Yeah. But now, I feel like that's gone. So everyone is just like constantly criticizing each other about like what women culture is and what man culture is and it just does not lead to anything sexy.
Speaker 2:Like that's just Well, why do think that Yeah. Why do you think that's gone? Like, is it the internet that we're just in person less where there's like interacting online, you're It's easier to be a jerk to somebody online?
Speaker 1:That I think it's just everything you get. Think that's part of it. I think people say like just the prevalence of like pornography, they say that's like another another part of it. Because like all these people, like when I like the vibe I get from these people, like they're so proud that they're like not letting their sexual desire over power like whatever Mhmm. Quote unquote standards they have for the opposite sex.
Speaker 2:Mhmm.
Speaker 1:But the underlying feeling is like, I feel like you're getting that elsewhere in probably like a really fucked up way. That's kind of the vibe I get. Yeah. Like they pretend like they're so like, you know, above it. So I think that's a part of it.
Speaker 1:I think the work from home thing is also like Mhmm. Yeah. It's hard for me to like not see any other way because I met my wife at work. Everyone I know from my life is from work. Mhmm.
Speaker 1:So I just can't imagine that like helping, like it's hard enough to meet people adding that on top of it. Also, the way that the online dating thing, think at scale just is not like it just puts you in like a weird mind spit like, headspace when you're just like, you know, flipping through those apps as well. So I I don't know. Something is just not working, like, these things together.
Speaker 2:It's not working And I think you're just kind of like hitting all the points in the podcast. So we've established there is this problem. You don't feel like we should be I just feel like increasingly compelled that we should be role models somehow, and I don't know how to do that. I feel like Prime does a good job of this. Like, he's kind of like a role model to young men.
Speaker 2:It just feels like every millennial man that has an online presence should probably be doing something about this problem so we have a generation coming below this one because birth rates, all that.
Speaker 1:I I for me personally, I am very reluctant to over I'm very like cautious of over correcting, right? Because Yeah. What we're witnessing is something was working and it stopped working. So the natural thing what's gonna happen is gonna be a crazy over correction to, like, to being very High match guillotine. Yeah.
Speaker 1:Yeah. Or that also, right? That that that's also kind of happening. So I don't even know what being a role model looks like. I mean, me, I just like making fun of people.
Speaker 1:So I just want to make fun of like I'm like, you are so like, to me that like, me growing up, if there was like, if a bunch of guys in our group and one of the guys was like, oh, isn't it so annoying when girls like are like this? Like, that person would just get demolished. Completely demolished. Yeah. And we all know exactly how they'd be demolished.
Speaker 1:I'm not gonna say it, but we all know what everyone would call that person.
Speaker 2:East coasters. Not here in the Midwest, okay? We had manners.
Speaker 1:But we're like, what are you are you not trying to get girls? Like what? Like that I just can't. It's just so it's just so different. So I don't know.
Speaker 1:Yeah, I think I think the problem is the role models that I'm saying I'm gonna say role models in quotes because Yeah. It's still just a minority of people that like kind of fall for this. But the people that are gaining popularity are the people kind of espousing this kind of thing of being like, oh, like, you know, you're cool if you like don't need like don't care about women at all, right? We gotta like make fun of these people because that's what we
Speaker 2:Could you name names? Like, is this the Andrew Tates
Speaker 1:the whole Andrew Tate thing like, he's just such like a freak in every single way.
Speaker 2:Mhmm.
Speaker 1:And I don't wanna get too specific but there's people that we've we know in our lives that kind of follow the kind of like are kind of trapped in in that whole thing.
Speaker 2:Mhmm.
Speaker 1:And all they can think about is how like every woman is like an OnlyFans girl and like they're all like this and they're all like this and like it's so hard to find like a good But what they don't realize and this is Liz's sister's observation is that their model in their head of what women are like are just the equivalent to Andrew Tate on the opposite side. Like it's just another extreme that is like not really real. So there's like caught in this weird fake perception of the world when like most people in their lives are like pretty normal. Mhmm. But yeah, it's like a bitter, sad road to head down.
Speaker 2:Yeah. And I feel like that's the role models that do exist online. Like if if young boys in the world today don't have a father in the home, so there's no male role model like in their home and then they don't they talk in the podcast about how most teachers are women. Like if they go
Speaker 1:to school and it's
Speaker 2:mostly women, if they don't have experiences around men and then they only experiences they have are online and it's the Andrew Tate's the world, then shouldn't normal people like b roll models too somehow? I don't know. Should we I I don't even know what I'm saying because like I don't wanna do something unnatural.
Speaker 1:I I don't know what what what that could be, like
Speaker 2:Just like saying things to young, I don't know. To men. Because we like say things that were specific that would help. Like we have families and jobs and we're normal and I feel like that that's worth aspiring to or is that boring to these
Speaker 1:people? No. The problem is I think these role models are as much a symptom as they are a cause. Like there is just this emergent thing where this is a dynamic between men and women these days and you see it among young people. Whenever I hear about what dating is like these days, it just it just seems really difficult and impossible and crazy.
Speaker 1:I'm like, yeah, if I was this age, I don't even know what I would do. So it's not gonna I don't even need advice for them. So yeah, think it's just like a it's like these role models are just kind of like symptoms as well.
Speaker 2:Okay. Okay. Well, you made me feel better because I just had this weird like, I don't know what we should do about it. I hate when someone presents a problem and it's like, could I be the part of the solution? And I don't know and my I just I think about it too
Speaker 1:have no clue. It's it's weird and I think it's it's really embarrassing. And I hope that these people get older and it's kind of one of those things where we're all kind of embarrassed what we're like at a younger age. I think it's just ultimately one of those things but
Speaker 2:It's a good outcome because like then we still have a civilization. We're not I kinda know if I'm actually scared of the birthrights. People bring up birthrights like this is a real problem. And like, I don't know, how is it? It's less people.
Speaker 2:Is that bad? I
Speaker 1:don't know. Yeah. I mean, yeah, birthright thing. I think that that also is like a huge the it's just like a multifaceted thing. Right?
Speaker 1:There's like so many economic challenges of having children, so that's that those need to get fixed too. Like
Speaker 2:Okay. That's a good segue. Economic economic economic problems, troubles. Okay. I'm gonna tell you everything I've heard about the tariffs in twenty four hours.
Speaker 2:Mhmm. I've been tossed in the winds every direction and I want you to tell me what's really going on, so no pressure, you better know.
Speaker 1:Okay. Okay.
Speaker 2:So I first saw the tweets where people were like, what an idiot. He literally just did this in chat GBT and reversed the trade deficit and set the tariffs on every country based on that. Yeah. He he dumb. We're going into the Great Depression.
Speaker 2:That was the first set of tweets. Then I saw tweets where people are like, actually, he's big brain and he's lowering the yields on ten year treasury bonds so they can convert a bunch of debt. It's like literally like this master plan. And I'm like, I don't know. Is there some truth in the middle that's neither of these crazy scenarios or are we just what's the is the first one the right one?
Speaker 2:We're just doomed?
Speaker 1:So I saw us I watched something yesterday which helped me a lot because I think the the biggest question here is has been the question from the beginning, right? It's ultimately is the not just a president, but is this new right
Speaker 2:Coalition.
Speaker 1:Like the new right coward
Speaker 2:and the right, yeah.
Speaker 1:Are they pro free trade or not, right? You can say if they're pro free trade, then the tariffs are simply a negotiating tactic with the ultimate goal to be But there
Speaker 2:are tariffs anyway. Right? Yeah. And a lot
Speaker 1:of people argue for that. They're like, oh, you guys are being stupid. It's just a really good negotiating negotiating tactic like. Mhmm. They're like, that's all he's doing and they've been saying that from the beginning.
Speaker 1:But if that's true, if that's your goal, you can't also have these other things, right? You can't also say, no, our goal is to bring manufacturing back here. No, our goal is to like make these jobs available. Yeah. That's not actually your goal.
Speaker 1:Your goal is to get back to better Right.
Speaker 2:It's one or the other.
Speaker 1:It's one
Speaker 2:or other. Gotta have to.
Speaker 1:Can't be both. Yeah. Right. Mhmm. So anyone arguing that the people that are like like kinda infinitely argue that everything they're doing is right.
Speaker 1:I mean, they've become they find themselves in conflict of of these things.
Speaker 2:Mhmm.
Speaker 1:So and I've been confused about that too. I'm like, what's the actual goal? Is it just negotiating Or is it that these are this is the way they view things and they think that it's like a rip off and America's in a bad situation globally, economically? And this is the new world order that they want. They want
Speaker 2:Mhmm.
Speaker 1:It to be that we're not gonna absorb the production of the whole world. You guys are on your own. We don't want armies across the whole world. You guys can develop your own armies. We're we're tired of of funding all this because Mhmm.
Speaker 1:Things aren't good back home and we we need to fix that and this is like our our silver bullet for fixing that. So it's confusing which it is. It doesn't help that I don't think Trump himself himself has particular clarity about this. So I think almost all the actions that are coming from him are pretty much due from the hip The
Speaker 2:team.
Speaker 1:Week to week, new justification for this, roll it back. Like it's it's like not not a lot of clarity. But the thing that I watched yesterday was an interview with an economist that is in charge of developing the policy for this new right. So this is him like talking about what he's been trying to push for for for for many years. And it is very much not pro free trade.
Speaker 1:They view things as the last forty years have been effectively a disaster. They think that the middle class and low and like middle class and blow have have suffered and they paid the price for globalization. They're very much eager to move to a system that does is not globalized. Where America is insular, focuses inside and we we're we're gonna have these high barriers to do trade. So that is their actual goal and they see that as this is the way to do it.
Speaker 1:What I'll give them credit for is I think they're genuine in saying they are really looking at the middle class and generally trying to help them. It's not some like trick to like benefit someone rich. I think this person, as a person that I was watching interviewed, that's generally his goal and he's been looking at to that lens. I ultimately think he's wrong, but if he's the one and his group of people are the ones shaping policy, then what that means is these are not just negotiation tactics.
Speaker 2:No, it's actually
Speaker 1:It's an effort to reshape how America fits in to the whole world going forward.
Speaker 2:Briefly, because it is a very complex topic. But could we briefly hit on like the the theory and your expected outcome? Because I guess the theory is it improves like manufacturing jobs and stuff like we start doing things in America instead and that gives the middle class better jobs, but all the prices go up. I don't understand.
Speaker 1:Yeah. Well, I guess the way they try the way they look at things is the world produces a lot of stuff and Americans buy it, right? So they're like, we're absorbing the We're paying people in other parts of the world with our wealth Mhmm. For the stuff that they produce. And we're not betting, we're not benefit They view us we're not benefiting, we're just like a consumer economy.
Speaker 1:Mhmm. And I think that's kind of a valid characterization, like most income from a lot of these big countries that have grown a lot and have become a lot more developed in the past ten years has come from from The US. But I think the part they're missing is that we pay them with money that we print. Like we print money and we give it to them and we're like, you have to do all trade in dollars, you have to buy all dollars because, you know, we use our military to make sure that all trade globally happens with dollars. It's a very sophisticated and crazy control scheme that the America has built in the past, you know, since post World War two.
Speaker 1:Someone made a really good observation that it actually might be too effective and that it's totally that the control mechanism is totally invisible
Speaker 2:Oh, saw this.
Speaker 1:To the point where the people running the system don't understand the system itself and so they accidentally start to dismantle it.
Speaker 2:And then we we lose so if you play it out, we start manufacturing in The US, we're not doing as much trade globally, maybe trade globally just generally declines. Yeah. And there's we lose those levers of control like presumably?
Speaker 1:We lose levers of control like more countries with our armies like that leads to like more variability like
Speaker 2:Conflict and Yeah.
Speaker 1:Like when you have armies you tend to use them. The other thing is The other thing that I try to focus on is you have to really understand yourself and play to your strengths. Like, The US, we are like a globalist country and we can play that game better than anyone else. So much better than anyone else. Nobody else can even try to play that game.
Speaker 1:So for me, and there's downsides to that game, but I like to look at the solution as let's play the game even harder find a to address those downsides without stopping playing the game. I feel like now we're trying to play China's game where we're gonna like suddenly manufacture stuff. We're like
Speaker 2:But we're like decades We're
Speaker 1:dropping stuff that China literally cannot do, China would love to do, literally cannot do, to now try to do stuff that anyone can do, right?
Speaker 2:Interesting.
Speaker 1:So that's why I ultimately disagree with this because I feel like it's kind of the like golden age thinking like, oh, things are so much better in the past and kind of let's go back to the good old days. I think it's that mixed with I think Trump's just been and we've talked about before, he's been in a part of business that is very zero sum, not in tech Mhmm. Where things can really grow. Yeah. So I think that also colors colors a view of things.
Speaker 1:That's what's going on. So I think these tariffs, whether they end up sticking or not, because like I said, Trump is a politician and politicians are fickle. Like, they need to like, they will go with where the wind pushes them and if like there's a bunch of pushback, they'll say they'll roll them back and be and they'll say the tariffs were went perfectly according to plan and they helped us achieve our goal of rolling them back and they'll declare the victory and they'll roll it back. So I don't know if they're literally gonna be in place forever but behind the scenes, I see like the underlying push towards this direction of like dismantling the the current system. And so that's why your portfolio is down.
Speaker 2:And mine is You took the other bet. That was wise.
Speaker 1:Yeah. I pulled out all my money from like the market like a like a month ago or so and
Speaker 2:Good timing.
Speaker 1:It's just funny what psychological impact that makes in your head like and this is everything that I was saying like, I don't care who you are or how like moral you are. We're all short term thinkers who just care about our own self interest. So the moment I did that, I love the tariff news. Like I love that this happened. All turmoil.
Speaker 1:It's impo like everything I said even though even though I disagree with it, the emotion, it's hard to not feel happy to see the chart go down because like Yeah. My money is going up, right? So all the people that are saying stuff like there's always people saying the stock market doesn't matter, like that's not what's important. Like all these other things are important. I know for sure those people literally are either 14 years old because they don't have a retirement account they just straight up have no money at all, like they're unemployed.
Speaker 1:Because at the end of the day, if money goes down, people get mad. Yeah. So we'll see how this plays out. But anyway, that's that's my summary of it.
Speaker 2:Okay. Well, I know more than I did before we started talking about it. So that's
Speaker 1:good. Yeah. I bet. Interesting, fun, funny times.
Speaker 2:Funny times.
Speaker 1:The the memes have been really good. I've been loving the terror The
Speaker 2:side note on the memes, We had this giant explosion of Vance memes, and then we had this giant explosion of Ghibli images. I never saw one Vance Ghibli, not one. How is that possible?
Speaker 1:Oh, that's a good point.
Speaker 2:How's that possible?
Speaker 1:Can I tell you something?
Speaker 2:I feel like they're they should be all over the timeline. Yeah, what?
Speaker 1:I don't I hated the Vance meme thing. I don't know why, I just didn't like it. I didn't think it
Speaker 2:was I didn't really understand it.
Speaker 1:And I was just like
Speaker 2:I mean, I thought he sounded like a doofus in that meeting in the Oval Office, like just some of the things he said were just kind of like annoying to me. Yeah. But I don't understand where people just took off with his face and like started doing stuff to it. I don't understand
Speaker 1:And I Whoever made that like chubby curly hair version of him like Yeah. Very impressive because it it, you know, it like went crazy viral. I hate the way it looks. Like, I hate it. But I also can't stop looking at it.
Speaker 2:This is like it's like how you hate the way that boxer that boxed Mike Tyson looks. What's his Wait. Dammit.
Speaker 1:Jake Paul.
Speaker 2:He's not even a boxer. Jake Paul. Jake Paul's brother. The other Paul. Whatever he's
Speaker 1:I mean, I definitely hate the way he looks. Like, he looks like a douche, but
Speaker 2:Oh, I just remember you saying how much you hated the way he looks that night in the boxing matches. I'm sorry. I don't know. We were we were in person that night.
Speaker 1:Yeah. He just looks unlikable. And I and and I deeply this is probably like somehow racist or like very wrong, but in the past couple of years I just keep seeing more and more proof that people are what they look like. And I'm like, look Adam, you look like a nice guy and you are a nice guy. I cannot picture your face with my personality.
Speaker 1:It just like does not it's it seems impossible. How could someone that look like that looks like you have my personality? Doesn't that seem impossible? Yeah. Imagine someone looks like me that has your personality.
Speaker 1:Impossible.
Speaker 2:I just, yeah. I guess I just know I do have questions we can get into at some other time about what you mean by I have a nice face. I've heard that before, like that I look like a nice person. I don't know what that means.
Speaker 1:But that's what I'm saying. It's like this mysterious innate thing where Okay. I'll explain it another way. There are some actors that are always in evil roles. Why?
Speaker 2:Yeah. No, it's true.
Speaker 1:Because somewhere in our head we're like that's what an evil person looks like.
Speaker 2:Their face just looks evil.
Speaker 3:Yeah. There is something there. Okay.
Speaker 2:And I'm not
Speaker 1:saying they're definitely evil but you know, like, I know they're not gonna have like a bubbly personality in real life, you know? Yeah. They're gonna be, like, kinda brooding. But I'm I'm just convinced that this is a this is a real thing.
Speaker 2:And that's why the Vance thing bothered you?
Speaker 1:Wait, what were you even talking about? This is this this this was completely something this is completely something that I just went totally off topic. It's just
Speaker 2:Oh, okay. Got it.
Speaker 1:I just, yeah, I just think I mean, I think that's why with Logan Paul, I'm just like, he just looks like he looks like a dick.
Speaker 2:Yeah. No, I agree.
Speaker 1:It's just it's just his look.
Speaker 2:He does he does look like a bad person.
Speaker 1:Maybe he looked that way and the world treated him that way so he became that. Maybe that's But yeah, whenever you meet someone with a strong personality, like you can like clearly have keep their like shape, articulate what their personality is. Imagine literally what they look like with the opposite personality and like you just will have no frame of reference Yeah. For anyone like that. Yeah.
Speaker 1:So there's something here.
Speaker 2:There's something there. Something here. Okay. We gotta get off here but before we do, terminal. We're shipping cron.
Speaker 1:Oh, yes. We
Speaker 2:gotta do that. Days away from the first cron. Yeah. Let's do an ad. So the ad is there are hundreds of cron subscribers and you should be one of them.
Speaker 1:Yes. Go to s s h. Open your terminal and sshterminal.shop and subscribe to cron. You will get a really beautiful package. I'm really excited to get this package myself.
Speaker 1:Oh, it's so good. It looks so good. I'm a box. All of it. UPS destroys it I know.
Speaker 1:Yeah. We'll see.
Speaker 2:It's gonna be in a bag, which will keep it nice and safe. Well Sure.
Speaker 1:We'll see.
Speaker 2:Yeah. It's gonna look super cool. There's all kinds of little details that I'm very excited to see in person.
Speaker 1:Remember how funny our first set of boxes were where it looked like they just like shredded?
Speaker 2:Yeah. It looked like that meme on Twitter.
Speaker 1:They like just like ripped it apart as hard as I could.
Speaker 2:They were so bad. I've seen so many terrible shipping mishaps like that's very common, I guess. Mailmen are angry angry people.
Speaker 1:It just it can't be that. It just must be like it blows my mind that just incidental contact can cause a box to get shredded.
Speaker 2:Yeah. I don't really understand, like, tossing the box around how some of the states that the boxes were in, like, they had to stepped on, some of them. It it was bad.
Speaker 1:And not just stepped on, I think they like grabbed it and ripped it, like they were shredded. Like, when I say shred shredded one. It was like Oh. Yeah. Like a paper shredder or something.
Speaker 1:Oh. Ridiculous.
Speaker 2:E commerce, they said. It it'll be fun, they said. Anyway Okay. Terminal dot shop.
Speaker 1:You should you should you should sign up for this experience, the one that
Speaker 2:we just described.
Speaker 1:It's gonna
Speaker 2:be so good. First first batch going out very soon. Okay. I gotta get up here.
Speaker 1:Alright. See you.
Speaker 2:Thanks, Sax. See you. What do you want?
Speaker 3:HTS, sir. And how are you this afternoon? Alrighty then. I have a package for you. Sounds broken.
Speaker 3:Most likely, sir. I'll bet it was
Speaker 2:something nice, though.
