Miami is Cold, Adam's All in on AI, Dax Cracks YouTube, and Kids Are Expensive
I knew it. I knew it. I love them so much. Oh my I'm
Speaker 2:not even
Speaker 3:would know to ask that question, but I knew to ask that question because I know you.
Speaker 2:It's been a while.
Speaker 3:When was your last recording? Was it more than two weeks ago?
Speaker 2:Two weeks? Yeah. We we didn't have two episodes. Does that mean it's been three weeks or two weeks? I don't know.
Speaker 2:Haven't talked to you since last year. That joke never gets old. It's always good.
Speaker 3:It's not true. We talked yesterday.
Speaker 2:Damn it.
Speaker 3:We had a recurring meeting yesterday.
Speaker 2:We did. We we had a recurring meeting. I'm actually super excited about this recurring meeting. I feel like it makes terminal more real. It gives us all a little more like, I feel more motivated after the call yesterday.
Speaker 2:Yeah.
Speaker 3:It's true.
Speaker 2:We're all on the same page. We got stuff we wanna do. Feels good. I think the recurring meeting is good as much as I hate meetings.
Speaker 3:Yeah. I'll I'll allow it in this case. There's mitigating circumstances that mitigate my hate of meetings, so
Speaker 2:What are the mitigating circumstances? I'm not smart on Just
Speaker 3:that's it's like not anyone's primary thing, so it's easy for it to flip.
Speaker 2:Yeah. Do you think we'll actually have the meeting every week or is it gonna be like seldom?
Speaker 3:No. Think I we'll have every week. I don't think everyone has to make it every week,
Speaker 2:but Sure.
Speaker 3:Yeah. Yeah. For example, I didn't make it yesterday.
Speaker 2:Yeah. For example. That wasn't frustrating at all. I don't know if you realize how important you are to those meetings because there are lots of ideas, Dax. Lots of ideas at our company.
Speaker 2:Everyone has ideas. We need someone to say, here's what we're gonna do.
Speaker 3:Yeah. I know. I I I, like, saw you guys in the call and I was like, I'm gonna come I'm gonna join here and they're just gonna it's just gonna be like in some place that I don't even I can't even imagine.
Speaker 2:What is the TV show? Is it Community when the the guy walks in with the pizza box and everything's on fire and like
Speaker 3:Yeah, exactly.
Speaker 2:Yeah, it was that situation. I wish we weren't on the call yet until you were on. I'm not throwing shade at any of our team members. It's just without you, we're kind of the Wild West. We would do a lot of things.
Speaker 2:I don't know if they'd be in any direction at all, but we would blow with the wind. We're gonna talk a lot of terminal stuff today just for what it's worth. Sorry podcast listeners, this is as much tactical as it is content at this point today after a few weeks. We didn't even like plan on taking a break. I don't even know if we talked about like, hey, let's just take a break.
Speaker 2:I think it just kinda happened.
Speaker 3:Yeah. My productivity wasn't great. It wasn't at zero because I do like the two weeks where everything is kind of quiet because I can do a bunch of stuff without like every day I wake up and the first thing I do is go through our support channel and I don't like doing that. So it's just this like every day starts off on a negative. In a lot of ways, it's just like hard to get started.
Speaker 2:Mhmm.
Speaker 3:And those two weeks, you know, it's not like, I know there's not much going on and no one's gonna like interrupt me in the middle of the day and stuff. So it's a special two weeks every year but I wasn't that productive either.
Speaker 2:Yeah. I could tell you were kind of like just chilling, taking a little time. Yeah. But between us not doing the podcast and not talking for that reason and just like I mean, everyone's pretty quiet during those two weeks. But that just kind of reminds me of why I hate those two weeks and vacations and summer and any time when I'm working and other people aren't.
Speaker 2:Yeah. I love when I'm not working and other people are. Fine. It's just the way the other way around is very frustrating.
Speaker 3:Yeah. I could sense it.
Speaker 2:Yeah. So well, maybe before we get into a bunch of terminal stuff. Hey, Ben. How how are things? We really haven't talked much.
Speaker 3:I'm gonna talk about the weather. That's what I'm gonna do.
Speaker 2:Oh my god, Dax. No one cares about your 54 degree days in Miami. Like, that's the warmest day I've had here in a month if if I had a 54 degree day.
Speaker 3:Okay. I wake up and I walk outside in my underwear and it's degrees. And that is not a good feeling, okay? So then I walk back inside and I'm like, okay, let me like put on clothes I guess.
Speaker 2:I guess.
Speaker 3:Ridiculous. And then I'm like, what do I even wear? It's like, oh, I guess I have to wear I'm wearing this like I don't even know what to call it. I got it I don't even remember what these clothes are called anymore. They're called
Speaker 2:pants. Backs. Those are pants.
Speaker 3:I'm talking I'm talking about the well, yes, pants. Yeah. Put on pants and then this is like is this a this isn't like a sweater.
Speaker 2:Looks like a sweater, yeah.
Speaker 3:It's a sweatshirt maybe? It's a sweatshirt. Sweat sweatshirt,
Speaker 2:yes. Okay. I'm not so good at this either.
Speaker 3:Put on a sweatshirt and go outside because you don't have to blow bubbles for Zuko.
Speaker 2:Mhmm.
Speaker 3:And my feet are freezing because I don't wanna put on shoes and then like my my plastic grass is very cold.
Speaker 2:Plastic grass. So
Speaker 3:that's very uncomfortable. Okay, whatever. Then I come inside, start my day. I'm sitting in my room. I don't have a heater in my house, So
Speaker 2:Oh, yeah. In Miami, guess you wouldn't.
Speaker 3:It's a Actually, I think I might have a heater. I just don't I've never tried.
Speaker 2:I don't know where it is or where the buttons I
Speaker 3:think if I turned up my temperature, I think the heater will kick on actually, but I don't know. It's weird. It's weird to use a heater. So now I'm sitting here and I've always had this problem where my hands get really cold, like my hands are just this blood just does not make it to my extremities. There's something called like Raynaud's or something.
Speaker 2:Oh, I don't have a condition, but when I work, like typing at my computer and then I go like upstairs and I like pick up my boy or something. He's like, your hands are so cold. Like my hands do get very cold.
Speaker 3:Yes. So right now my hands are ice cold. Yeah. So that's what I'm dealing with and it's it's really sad for me. It's probably the most tragic thing going on in the world right now, weather related.
Speaker 2:Oh, my word. Dax, it's 26 degrees here right now. I just looked at my watch. Low of 12. Would you die?
Speaker 2:Would you just fall over and die? I mean, you're from New York.
Speaker 3:Yeah. I I would just die. It's just the the resistance wears quickly, like, whatever. Yeah.
Speaker 2:The acclimation thing is so interesting. Like if I went outside and it was 50 something degrees right now, I could be in my underwear and it'd be Yeah. Perfectly
Speaker 3:You'd be wearing shorts.
Speaker 2:Yeah. I remember Yeah. It would it would sound amazing.
Speaker 3:But this is gonna last two weeks. I have deal with this for two whole weeks. And then it'll be warm again.
Speaker 2:Oh, is that that's the period of winter you have there in Miami? Two weeks?
Speaker 3:Mhmm. This is first couple weeks in January is always like this.
Speaker 2:Is the best time to be in Southern Florida in my opinion. I don't wanna be there in May or June, any of the summer.
Speaker 3:That No, that's better.
Speaker 2:But this time of year when you can like you can be outside on Christmas and it's not that bad, that's that's nice. I guess I'm always coming from like a visiting standpoint, so it's like it feels nice. But if you live there, I guess this does feel cold.
Speaker 3:I mean, you're okay risking death from freezing, yeah. I guess I guess that's fine.
Speaker 2:Oh, that's funny. Okay. Well, no one can relate with it.
Speaker 3:So that's how I've been.
Speaker 2:So I'm glad you're surviving your difficult weather there in Miami. Not hurricanes. Wait, wait. That's not difficult. 54 degrees.
Speaker 3:No. Hurricanes are easy. Literally no problem. Hurricane is not a problem at all. We have fortresses here that are just impenetrable.
Speaker 2:We're good.
Speaker 3:Yep. Okay. Here but here's another detail. So I got kicked out of the kitchen because you know I've been making two steaks every single day. Yeah.
Speaker 3:We don't have a hood like a
Speaker 2:rainbow. Yeah.
Speaker 3:So it's just been blasting like smoke into my kitchen for like three months and it's just it's just been covering everything with like a thin layer of something. I don't know I don't know what. But I kicked out of the kitchen, so now I have an outdoor setup. I got a propane tank and a burner outside.
Speaker 2:Oh, no. You have to go outside in the cold.
Speaker 3:So now I have to cook outside and like I have to put my hands over I put my hand over the fire while it cooks to stay warm. So, guys, I'm really struggling out here.
Speaker 2:Sounds This
Speaker 3:is the hardest anyone's ever struggled
Speaker 2:ever. Mhmm. Probably.
Speaker 3:Yeah. I have to do that to eat my rib eyes, so
Speaker 2:To eat your rib eyes. Yeah. You have so many tweets about meat, it's crazy. Just how many of your tweets are like, here's a picture of wagyu or whatever. It's just it's like your whole timeline.
Speaker 2:I mean, you tweet a lot, so it's a small percentage of your tweets but
Speaker 3:Yeah. It's like I'm literally I like I like ration myself. I wanna tweet even more about me. I wanna tweet about it like twice a day
Speaker 2:You mean like all
Speaker 3:twice a week.
Speaker 2:Just an alt account. Yeah.
Speaker 3:It's just me. Yeah.
Speaker 2:There just needs to be circles. Remember, what was the when like Google came out with a social media thing and I remember do you remember that? Like, this is just I
Speaker 3:called again.
Speaker 2:Google one or what was it?
Speaker 3:I don't know.
Speaker 2:There was like a whole like media frenzy and we all felt it in the air. It was like y two k. It felt like Google's gonna take over. Facebook isn't gonna be the dominant social media anymore.
Speaker 3:Google Plus. Was Google Plus.
Speaker 2:Google Plus, which now is probably a different product. That sounds like an actual product. I don't know. You know what I mean? Yeah.
Speaker 2:Do you do you remember that feeling when that was coming out that it was like
Speaker 3:I I never felt like it was gonna be I I never felt any
Speaker 2:help It around
Speaker 3:just felt like this random thing.
Speaker 2:I remember feeling like this is gonna take over and Facebook's gonna be left in the dust. And they had this concept of circles and like you Yeah. Set up And all your different it kind of makes sense like you have different friends for different reasons and maybe they care about this and the other group cares about that and you don't want to share everything with everybody. I don't know. It kind of makes sense but in practice, I don't think anybody does it.
Speaker 2:Nobody wants to manage
Speaker 3:It's too much organization. How have you been?
Speaker 2:I've been good. Okay. I got a little obsessed with the thing I'm working on and it's Mhmm. That's that's dominated some of my thoughts for well, was like the Christmas break. It was it was kind like your period of not doing much.
Speaker 2:I took a week off the Christmas week and by taking it off, mean, I took it off from my normal work and I got obsessed with this little project and it's stupid and it's fun and It's literally
Speaker 3:unbelievable. Okay. We can't talk about it yet, but I do not even understand how Adam is doing this thing. It's like, it's so weird and crazy and I I like looking at it and using it. I'm just like, I don't understand how this is even possible or how
Speaker 2:it It's made possible by two facts that again, can't really say, man, we have too many secrets around here. We just I just It's okay. And this isn't even gonna come out for months probably because we're gonna wait and got we other stuff to But yeah, there's two little known facts. Two facts I didn't know about that make this possible. And and then yeah, this it just opens up all kinds of possibilities once you know those two facts.
Speaker 2:So stay tuned to terminal dumbness. There will be something dumb at some point. Well, there'll be a lot of dumb things between now and then.
Speaker 3:We I I think this might top coffee over SSH. This might top it.
Speaker 2:Spoiler. It may be another way to buy coffee.
Speaker 3:Yeah. This might this might top it. Even dumber.
Speaker 2:Yeah. Anyway, other than that, been good. Ace is back in school, so my oldest is back to school. We hired a nanny and then she was like, never mind. After a week, Archie fell in love with her.
Speaker 3:Oh, that's sad.
Speaker 2:Like literally told her every day she left like, I love you. Bye. And like every day, when is she coming? When is she coming? And then turns out she doesn't want to be a nanny.
Speaker 2:Was her first try at being a nanny, and I think she just got bored playing all day. Literally, we just wanted her to play with our youngest because he just needs constant, like, companionship, and it means nobody gets anything done around here. And she was basically like, it's such it's such an easy job. I even feel bad taking money for it. He's the sweetest.
Speaker 2:It's too easy. Sorry. I don't wanna do it. Like, what? Okay.
Speaker 2:So we're trying to find another nanny. We have one starting on Monday. We'll see how that one goes.
Speaker 3:Okay. So I I have a bunch of questions about this because Mhmm.
Speaker 2:I don't
Speaker 3:know if you saw AJ posted a tweet because he's looking for day care for his new baby that's like probably six months older or so now. And Boston, which is like probably I think he was saying it's maybe the most expensive place in the world for day care. It's like $3,800 a month.
Speaker 2:Wow. That's
Speaker 3:for that's for day care where you like leave them with a bunch of, you know, it's not like an individual person is taking care.
Speaker 2:Yeah. I got to hire a full time nanny for that here in
Speaker 3:Yeah, I was gonna ask like so I'm like, what I was like, should probably look into this and understand these costs. Not that
Speaker 2:Yeah.
Speaker 3:Like I think you're just you're just stuck paying for it. There's like no, there's not really a way to wiggle out of it. But yeah, I had the same thought. I was like that seems like you just have a full time person. Yeah.
Speaker 3:The person that you hired, are they considered full time or is it like a part time thing?
Speaker 2:Like, I mean, these nannies or most of them are open to both. They prefer full time. So Yeah.
Speaker 3:But in your situation specifically, is that is it a full time thing?
Speaker 2:Yeah. We're trying to get somebody for just like thirty hours. We weren't trying to do forty and the two people we've talked to have been fine with that like thirty versus forty. Like
Speaker 3:six hours a day? Is that what that Something comes out
Speaker 2:like that. Yeah. I think that's right. Yeah. Yeah.
Speaker 2:Five days, six hours. Yeah. That's right. It would just make a huge difference in our life because then Casey could actually have any time where she wanted to do something she wanted to do or just even like we just get so behind on house stuff, chore stuff that needs to be done. Archie's kind of a special needs ish child.
Speaker 3:Mhmm.
Speaker 2:He needs a lot of something. And we know we're not just crazy and he's not just spoiled or whatever because we've had another child and he wasn't like this at all. Yeah. We don't know. It's it's something we're gonna figure it out at some point, but we'll we'll see if nanny to take two works.
Speaker 3:Is this like a minimum wage job? Is it like like I'm trying to understand
Speaker 2:what Yeah. Don't know what minimum wage is, but it's around here anyway in Missouri. It's like 15 to $20 an hour Yeah. For a nanny. That's what we're seeing.
Speaker 2:First That's one we
Speaker 3:not too crazy.
Speaker 2:It's not nuts. Yeah. The first one we had was like in her mid twenties. This next one is 17. She just graduated high school and she doesn't want to go to college and she wants to be a nanny.
Speaker 2:So and she's actually like spent a lot of high school babysitting, like a lot. So I don't know, we'll see. It'd be cool if we could have somebody for a few years that we knew wasn't gonna get married and have babies and then leave us. Because we, you know, we had an assistant, Hailey, who helps with this podcast. Hi, Hailey.
Speaker 2:She is amazing. Hailey is the best and then she had a baby. Congratulations, Hailey. I hate you. Now she's having her second baby.
Speaker 2:So that's the thing, you find somebody you really love and they're great and they fit in your home and everybody loves them and then life events, you know. Just kind of the nature of
Speaker 3:What about going to the opposite extreme? Get someone that's like 65.
Speaker 2:So that's another thing we've talked about, like can we just hire and have like another grandma Yeah. We pay to come here. That could be interesting too. We I don't know if we've had a lot of people apply that were older.
Speaker 3:Older? Mhmm.
Speaker 2:Yeah. Most of them have been younger. Maybe that's just a geographic thing, maybe in this area, I don't know. It does seem like there's got to be a bunch of women out there, men, get probably women. We would probably prefer older women.
Speaker 2:I don't know. Is that okay to say?
Speaker 3:No. It's fine.
Speaker 2:Okay. Yeah. It's fine. I prefer it'd be a woman.
Speaker 3:So there, whatever. No mannies.
Speaker 2:No mannies. Like there there has to be older women that don't have grandkids of their own or their grandkids live far away and they're like, man, I wish I could just grandma all day. That's probably a thing too. We'll see. Hey, if you live in The Ozarks, specifically close to Springfield, Missouri and you would like to nanny for our adorable little five year old boy, just hang
Speaker 3:And and and you're an older older woman?
Speaker 2:Yeah. Older Yeah. Woman or Really anybody. If you listen to this podcast, that's a pretty quick filter that I'm gonna hire you. So because I'm a narcissist and I want anyone in the home to know that I have a great podcast.
Speaker 3:So they're basically they're in the house all day or they're like taking him out somewhere or how is it?
Speaker 2:Mostly in the house all day. Yeah. I guess we haven't really sent we haven't put either of our boys in the car with other people very much like grandparents, but that's about it. Oh, that's scary. Oh.
Speaker 2:Yeah. Oh, man. My oldest is gonna be driving in like five years. What is that gonna feel like? I get it.
Speaker 2:I get all the tropes about like like crying as they pull out of the driveway because that's terrifying to think of my kid. I mean, he's 10 now, so I guess it would be terrifying if he was dressed right now. But we'll see.
Speaker 3:Yeah. I get it. Yeah. So and we yeah. We were talking about this too and it's yeah.
Speaker 3:We have some options because we have a lot of family.
Speaker 2:Must be nice.
Speaker 3:And but like, you know, we have this whole thing where Liz is very unsure how confident she is in that and taking that option. Yeah. Yeah. And then and then there's like and I don't know if this is I don't know the where you are is populated enough for this to be a thing, but there's a type of nanny you're talking about, then there's like more professional ones that are like trained in certain ways. So people Mhmm.
Speaker 3:And obviously, the the price there's a good price difference there.
Speaker 2:Yeah, sure.
Speaker 3:And then there's the local daycare which we can walk to, which to me sounds fun, especially when the kid is like not an infant because there's like other kids and stuff. But yeah, again, how expensive is it? Is it full? Is there like a wait list? Like all that stuff.
Speaker 3:People will talk about how impossible it is. In New York, this was like so crazy. All this stuff was just like so crazy, so impossible, like so difficult.
Speaker 2:Yeah. For I guess for our situation, he's just a year away from he probably goes to school in August. We think he's gonna enjoy school.
Speaker 3:Term thing.
Speaker 2:Yeah. It's really not long that we would need it unless he waits two year because he could technically he wouldn't have to be in school until the following August. So it's possible we need it for a couple years, but we'll see.
Speaker 3:Yeah. The the other option that exists here, which again, I don't know how maybe this is an everywhere thing, but if you look up daycares on Google Maps in Miami, you'll just come across just like normal looking houses because it's pretty normal for like one person in the neighborhood just to like kind of start a business where they just take a bunch of kids. And it'll just be like, they'll have like five or six kids at a time and it's just at their house and it's like, you know, on a street. And obviously, it's probably more affordable than Yeah. Some of other options.
Speaker 2:Yeah. I guess I guess we haven't looked at daycares. I mean, we're just both home, so it'd feel weird to like take our kids
Speaker 3:It's nice that it's nice. That that's what I was telling Liz. I was like, I feel better if they're just again, they're not old enough where like, they would even interact with other kids, especially when they're like infants or whatever.
Speaker 2:Right.
Speaker 3:I was like, feels better if someone's at our house versus us dropping them off somewhere orange.
Speaker 2:Yeah. And it's crazy that the the cost isn't that different. Well, I guess I'm comparing Missouri and Boston, so maybe very different. I don't know. Okay.
Speaker 2:That's enough about nannying and babies and kids and all those things. Let's talk terminal because I have so many thoughts and questions. And Yeah. By the way, this show is sponsored by Terminal. Buy our coffee.
Speaker 2:No. Subscribe to Kron. No. Become a Kron member. We're gonna make Kron.
Speaker 2:It's a membership. It's exclusive club. You're gonna have so many benefits if you become a member of Terminal, Kron. Yax, tell them what they're gonna get.
Speaker 3:You know, it's so crazy that we don't even wanna talk about it Mhmm. To anyone besides our Kron members. So Mhmm. We're we have a lot of really weird crazy ideas we need to pull people to do it with. So, yeah, like if you're a Kron member as Boire, we're gonna that's the first place we're gonna go for that stuff.
Speaker 3:So if you want to support us, if you want to help us do some of the weird stuff we do, please subscribe. And we're gonna keep saying this every episode.
Speaker 2:Yeah. And the way you do that, we're gonna say it every episode. You can just sshterminal.shop. So go into our s s h store and you can become a member from in there. Just buy cron.
Speaker 2:There you go. So the the thoughts I have about terminal. I I have in that call yesterday and Began Began bot has been joining just like in our Slack. He's in calls. He's just he's a wealth of ideas.
Speaker 2:He had an he had the idea that nerd sniped me into this obsession I currently have. Yeah. So he's got all kinds of wacky ideas. Love VeganBot. Just being on that call, I realized like there's kind of this like we got the left brain, right brain thing going at Terminal.
Speaker 2:I feel like me, you, and David could execute and build anything and build it really well. Like, we're very like precise and we make really high quality things that we make. And then Prime, Teej, and Vegan are like our crazy cousins and they have the most wild ideas and they could just like run with some stuff. I feel like they should be able to just run with like a lot of Vegan's ideas around like what digital membership could look like for Cron. Like, I want to like unlock them and let us keep building the very normal looking coffee company that's high quality, a great experience buying our coffee.
Speaker 2:Not that we don't overlap and that there's not like cross pollination happening, but I wanna feel like they can like do stuff without us having to be a bottleneck to their stuff. Does that make sense?
Speaker 3:Yeah. Yeah. I gotcha. And I think the way I always look at this stuff is I feel like companies always fall into these like two distinct phases where there's an initial phase where you have all these crazy ideas that you wanna do Mhmm. But you just can't do them in this initial phase.
Speaker 3:Yeah. Then you hit some kind of milestone where then you can now start doing a lot of things that was kind of the point of even starting the company in the first But that like sometimes takes like years and sometimes companies like never even get there. So to me, I feel like we're like in that initial phase. I don't think we're far from being out of it. But it's gonna feel uncomfortable whenever we have these like really exciting things because we don't really have a mechanism to there's so much foundation to lay before we can do that.
Speaker 3:Yeah.
Speaker 2:That explains the way I feel a lot, think. Because we we do have so many ideas and it just feels like because it's everybody's kind of side thing, it feels like we're always gonna have more good ideas than we have bandwidth to execute on them. But you're saying maybe there's just this phase we're in and eventually we get into a phase where more of these kind of things that we dream up and wanna do actually happen?
Speaker 3:Yeah. Yeah. I don't think we're that far from it. There's just some stuff that we need to set up. And it's funny to think like it's it's always easy to forget because we're always focused on the next thing and we like forget of all the things that are added up already.
Speaker 3:Yeah. Like we don't think at all about our coffee purchases at all Yeah.
Speaker 2:Right now. Just happens.
Speaker 3:It's like totally Yeah. That was like that used to be something that was like such a focus and
Speaker 2:Good point.
Speaker 3:Every single week we were like trying to make that happen and now we like don't even think about it at all. Yeah. I think eventually some of stuff we're working on now, we'll get there and then, you know, we'll have more leverage to do some of these bigger things. Yeah. But it's gonna be uncomfortable till we're But
Speaker 2:did what I said does it make any sense to you like
Speaker 3:That that makes total sense. Yeah. I think the way you described just like the leanings of the people makes sense and people yeah. I think not being able to make the stuff they wanna make happen is is not good.
Speaker 2:Yeah. It's like like I would never do like, I'm thinking from like a marketing standpoint and just like creating content and putting stuff out there. I would never do anything like what those three would do. It's just not in me. I'm like very perfectionist y and like everything's very polished and I wanna do things like a certain way.
Speaker 2:But like they're like fire from the hip, crazy ideas and like some of them not even a lot of effort, but they would really resonate with the people who like Terminal. I feel like that has to come from them and how do we allow them to go direct how do they not like I don't know, have to like run stuff through the company or something like how does that make sense? I feel like there's like a disconnect where they can't just do those things and I don't know what it is.
Speaker 3:So like, I think what you're describing is the idea is born and then nothing ever happens after. Mhmm. And part of that is because maybe
Speaker 2:they
Speaker 3:like literally can't make it happen. Mhmm. That's half of it. But I think the other half is actually just a little bit of a I don't what the right word here is, but it's the same dynamic where everybody in the planet, not everybody, every engineer in the planet has ideas for like, side projects or whatever and like, such a small fraction of people get it to happen because
Speaker 2:Mhmm.
Speaker 3:It's very easy to under underestimate like, all this other invisible stuff around these things that you need to do. And then when you start doing it, you're like suddenly become aware of it and then your motivation just or you get distracted or it goes it goes away. So, yeah, like that muscle I think is just a little weaker with them because that they just this is like not something that they've done a lot.
Speaker 2:So there there's this perfect example where one of them had an idea for if we're we're late, like we can't ship our first cron membership box on February 1 and we have to do it March 1, we could just send a blank shipment that has like a note in it that's like project delayed. And it was kind of funny like programming humor. And that idea just sounds like, yeah, let's just do that. But it's like when you start thinking about, okay, how would we do that? The actual logistics of that are kind of a pain in the ass.
Speaker 2:And like Yeah. There's actual steps that have to happen to make that happen. It's not just like, oh, okay. Yeah. That's just like buys us time thing.
Speaker 2:Like, no, that's actually a thing now we'd have to do. What does that note look like? Is it printed? Is it written? How are we getting all that set up for Maher?
Speaker 2:How big of a box? What are we are we putting in an envelope? What are we doing? Like, there's so many questions still that like an idea like that can sound so easy and like this is a thing we can do because we're not gonna have the hard thing ready. But even that is like a whole bunch of work and it's kind of a hard thing.
Speaker 3:Yeah. Yeah. And and this just comes down to like like for a high performing team, everyone has to understand everything. There's like just no way around that. So if you don't like really know how the details of what it looks like from having an idea to like making it happen Mhmm.
Speaker 3:You just don't have this you just like don't build up the sense around that.
Speaker 2:Yeah. So I think those three have the desire and the bandwidth to like execute on these things. Do we just need to help coach and guide them towards taking those steps? Maybe that's it?
Speaker 3:Yeah. Would say so. I I haven't thought about this at all. So I haven't like Yeah. I haven't like been able I haven't like done anything to to help that.
Speaker 2:Yeah. Use your big brain. Do that thing where you just stare at the wall and think and maybe come up with some answers on this. I wanna us to do more of the crazy things.
Speaker 3:I'm of two minds. So one is what I just described, like, the best team, every single person does everything. Like, they they just know how to do everything. Mhmm. And that's just always gonna be the best.
Speaker 3:But it's not realistic to take any group of people and have that situation come out because we have some advantages. It's not like like there's some upsides to the different people on the team that like are very specific. Yeah. And the thing I'm describing is not very specific, it's a very general way of doing things. Mhmm.
Speaker 3:So the other other part of me thinks, okay, like maybe we need to build a way to make it so they don't explicitly don't need to like know anything, you know, anything else about the So, yeah, I'm not sure, like I just need to think about and understand. Yeah. Like, I guess I I I haven't haven't thought about it too much but
Speaker 2:That's fine.
Speaker 3:Yeah. If we just get better at some of the common logistical stuff that always happens whenever we do something and it's just kind of like something that's not it doesn't feel like going from zero every single time Yeah. Then it's possible that like they don't even need to know much and it's kind of all enabled. Yeah. The other thing is like, literally if we just have if we just make enough money to hire one person that is like basically full time
Speaker 2:Mhmm.
Speaker 3:That kind of just serves as like the solution to a lot of this stuff.
Speaker 2:Okay. Like the it's not just stuff we send to people who buy stuff. It's like just our social and Mhmm. Content properties like that could all just be blown out so much more than it is. And I feel like they have that skill.
Speaker 2:So that would be great.
Speaker 3:Yeah. Yeah. Yeah. It's it's a bit there's there's a bunch of stuff but think ultimately like you're you're identifying a problem correctly. I I just don't know exactly what to do about
Speaker 2:it What solutions are? Yeah. Okay. That's fair. Totally fair.
Speaker 2:Yeah. There's just a lot of stuff on the terminal front now with the the coffee business that I care about. I feel like toward the end of last year, it was like Terminal is an events company. And Mhmm. After our last event, it's like very excited about doing these events and that was just a lot of fun.
Speaker 2:And now, I kind of think like in January, I'm just feeling like we got a lot of stuff we could do to make the whole like we're an ecommerce business thing more real and yeah, I guess I'm excited about that again. I don't know about events.
Speaker 3:Yeah. I mean, that's that that that has to be the foundational thing because that's the only thing that is scalable that we can like like like like I said, we don't even think about our week to week coffee sales anymore.
Speaker 2:Yeah. It just happens.
Speaker 3:And that's
Speaker 2:It's kinda nice.
Speaker 3:Yeah. Whereas an events thing is never gonna really get there. Yeah. Yeah. Yeah.
Speaker 3:We we just haven't like tapped. There's like so much untapped stuff here.
Speaker 2:Oh, man. We should another little teaser for those who haven't become a Kron member. Please, if you're listening to this podcast and you're not a Kron member, how dare you? What are you doing? What are you even doing with your life?
Speaker 2:If you're not a Kron member, just a little teaser that the first bag has been designed. It's hopefully gonna go out February 1 and it's amazing. It's hilarious and amazing. And David is incredible. Like, art Yeah.
Speaker 2:He's the best designer ever. He David is just the best.
Speaker 3:It's so funny because when we were describing the idea during the meeting Yeah. I had, like, an image of it in my head. And then like an hour later, David designed it and posted it and I I was just laughing because I was like, this looks nothing like what I imagined. And I was
Speaker 2:like Good.
Speaker 3:A thousand times better than anything I
Speaker 2:could even comprehend. So it was an hour after the meeting, he posted the artwork. And during the meeting, when we're describing the idea, he's having like Google and look up what we're talking about because he's never heard of this. So he went from not knowing this was a thing in programmer land to like designing this amazing bag that represents the idea so well in an hour. It was impressive.
Speaker 3:That's so funny.
Speaker 2:Yeah. Okay. So that's terminal. Anything else about terminal? I feel like there was something else, but I can't remember.
Speaker 3:No? I think that's
Speaker 2:No, no? That's it. That's it. Okay. What else are gonna talk about?
Speaker 2:What's been going on in tech? You wanna talk about technology?
Speaker 3:Tell you some I've been doing a lot more YouTube. I've been Oh,
Speaker 2:you have? I've seen some of your YouTube.
Speaker 3:Yeah. So we've been that's like it's a bunch of stuff with like the actual, like, SC itself has gone to a good place. So now, like, we're reallocating effort to
Speaker 2:Marketing.
Speaker 3:The YouTube side. Yeah. And it's it's kind of funny. I'm like, alright, this is gonna sound really arrogant. So prepare prepare yourself.
Speaker 2:Dex?
Speaker 3:No. No one is surprised. Okay. I'm like, okay, we're gonna do YouTube seriously. Let's try to like max out what we can get out of it.
Speaker 3:Mhmm. I looked at a bunch of other company YouTubes trying to like figure out like, okay, what's the bar? Like, what's possible? Looking at other like tech individual YouTubers in our space. I was looking at their numbers and I was like, okay, let's set some goals for ourselves.
Speaker 3:Let's get to a place where for for like a educational type video we put out, get 10,000 views in a week. That's like from what I saw, was like, that seems like a decent goal. Everyone always talks about how hard YouTube is and even getting there would probably be pretty difficult. The first video we did got 25,000 in two days.
Speaker 2:Look at you. In two days.
Speaker 3:And I'm kind of like looking at this and I'm like so I started watching a lot more YouTube in the past month. Not like in our space, just like in general. I've been trying to like reduce I like reallocate like what I'm watching and doing more YouTube so I understand it better. And I like and I see a bunch of people that I like and I watch their stuff and it is so so much better from every like micro like the obviously production quality but also like every single word they say in the video is like perfect. Whereas for me, I feel like 30% of the words in my video were correct and like, you know, it's like I'm definitely right at the beginning, like it's already doing better than what everyone kind of makes it out to be.
Speaker 3:I feel like YouTubers are always like, this is so hard. Like, it's so hard to be a YouTuber and get views and like, it's just such a grunt. There's like talking about this stuff, I'm like, okay, maybe I just had one lucky video and like the rest are gonna suck but like, I've made 2¢ and they've all crossed that initial goal that I thought was gonna take us a year to get to. So I say it just feels a little bit like
Speaker 2:This is too easy.
Speaker 3:No. I'm like I'm more commenting about like the the attitude from the YouTubers of being like
Speaker 2:Oh, I see.
Speaker 3:I kind of wonder like, are you just making it so you have less competition? Because like, everyone talks about it. A lot of people are like, oh, I'm kind of interested in this, but it sounds really hard.
Speaker 2:Like gatekeeping, gatekeeping. Is that what that's called?
Speaker 3:I don't know if that's what it is, but it might be subconscious, but I'm like, I don't know.
Speaker 2:Well, yeah, I wonder like your your case. Do you chalk that up to like people are hungry for more SST information or are you positioning the videos just broadly to like developers and teaching like, what is your your sort of like positioning of the videos?
Speaker 3:Yeah. So we see two gaps. So one gap is within the AWS space, there's just Yeah. So for example, we made an Aurora versus RDS video. So I searched up on YouTube.
Speaker 3:Okay. Aurora versus RDS. Classic question that like everyone has like what the hell is it? Video that I thought was kinda okay, like okay. Everything else was like just nonsense.
Speaker 2:Yeah.
Speaker 3:So there was really just one result for that. So we're like, okay, That's there pretty crazy. Isn't There just isn't a lot of stuff in this space. Yeah. So when we make videos like that and we have a lot of we have infinite ideas there because there's just so much stuff we can do.
Speaker 3:We don't expect those to go that big because, you know, that we are Yeah. Focusing on some on some some audience. But to be fair, we have like it's not like we're starting from zero. We have a community that like Yeah. Will drive stuff there.
Speaker 3:But you know, most of the views are that from there are coming from like 80 plus percent of views on those videos are coming from people discovering it from within YouTube. Yeah. We're not driving them from Twitter or anywhere else.
Speaker 2:It's not they're coming from Twitter. Yeah, sure.
Speaker 3:Yeah. And then but then I also we make a bunch of those and I'm like rough the ratio that roughly have is for every four of those I make, then I'll make one of these other kind of videos which is trying to come up with stuff that's as broad as possible. So the first video I made was just on Go. I was just like talking about my things around Go.
Speaker 2:That's the one I was thinking of that yeah.
Speaker 3:And that one did like I would say that one did probably like two or three times better than our AWS stuff, which makes sense. It's it's a bigger audience. Yeah. And when I made that video, was watching it, I was excited for the idea, then I made it and I was like, man, like this isn't that good and I was kind of like not super excited about putting it out. But then, like I said, it did it did pretty well compared to what I'm seeing out there.
Speaker 2:Yeah.
Speaker 3:Like in my head, 50,000 views was like, if you're gonna get 50,000 views on a video, you're like super legit to the point where you can get sponsored. Like it's like
Speaker 2:Yeah.
Speaker 3:Like a big deal. But I feel like we're maybe a year away from that. It doesn't feel like unattainable.
Speaker 2:So it's weird to me. You've talked about this AWS content gap. It's weird to me that like AWS is that much more niche than Go. Because like everybody uses AWS. I don't I feel like for there to be so little in terms of content out there in the third party world, like AWS puts out a ton of stuff.
Speaker 2:But like YouTubers that just do AWS stuff, there's not that many and like tons and tons of people use AWS. AWS is the
Speaker 3:thing That's the most common.
Speaker 2:That more people use and don't talk about. It's so weird to me. Like Yeah. Why on Twitter is there so little discourse about AWS? When like you go to I remember being at TwitchCon and TwitchCon is like the most opposite even though Amazon owns Twitch.
Speaker 2:It's like the most opposite thing to AWS I can imagine. And we're sitting around there, all the developer streamers at TwitchCon and we're talking to AWS because they all work places where they use AWS and it's just wild to me.
Speaker 3:Yeah. I I think what we've what I found is I don't know how to position these videos so that they're they're stuff that a broad audience would actually try to watch. So the Aurora versus RDS one, so that was my that was me experimenting with, okay, let me just like hardcore do like a very explicit this is an AWS video, obviously. Like, you might even find it from searching about this topic. Mhmm.
Speaker 3:But I feel like not that many people even if the video is good, there's not that many people are even gonna like go watch that. I don't know why it just doesn't I can see how it like doesn't feel like appealing. Yeah. Whereas I can definitely explain some AWS stuff in a much broader way where a lot of people do watch and like the AWS part is almost like a Trojan horse to like explain this bigger topic. Mhmm.
Speaker 3:But yeah, I don't know. I just think there's not a lot of it's very common but I think there's not a lot of like interest in people
Speaker 2:Mhmm.
Speaker 3:Trying to spend their like educational hours on that.
Speaker 2:So interesting. Yeah. It's like, is it not it's not sexy? Like, if I wanna Yeah. Learn something
Speaker 3:like developer. Yeah.
Speaker 2:Like, I wanna learn some new language. I wanna learn about Zig because I got some time now. I got an hour and I wanna watch something like Yeah. Watching like how to choose between RDS and Aurora versus like, I'm gonna learn a new language is definitely less appealing, I guess.
Speaker 3:Yeah. But I also think you're right in that it is a giant population and I feel like once we crack something, like those videos might end up being much bigger. Yeah. Like there's no reason why they shouldn't be 10 times bigger.
Speaker 2:Yeah. I'm glad I'm glad you brought this up. This is actually a thing I I forgot that I wanted to talk about when we were talking about terminal, which is I got a message this morning from Bash who is working She works for Charm, right, on the Charm CLI. She's in our terminal Discord and she wanted to show how to fix a bug in our shop and she's like, well, I can't get I can't get it running because I don't have SST set up. And I've just generally when like we've open sourced the repo, I want to have the open source community feel like they can come in and do stuff.
Speaker 2:We have so much stuff we'd like to do in the shop if we could have other people doing stuff for free, that sounds amazing. But how do we make it so that it's like super simple? Maybe it is just like more making a more clear set of steps if you don't want to run the whole back end. Maybe just people who want work on the front end, that's like a happy path where they don't have to do anything with SSD. But like more broadly, do you think there's like still just some kind of gap?
Speaker 2:Because SSD is super easy to get running and it's super simple, but it's just is it lack of familiarity?
Speaker 3:No. The issue here is they don't have access to AWS account. That's actually
Speaker 2:the It's the AWS account dependency. Can we make that that's what I had this idea. I had that thought and I had this idea. Can we just have like a an AWS sandbox account that we make available to people who wanna do open source contribute contributions?
Speaker 3:If they're just working on the SSH app, that requires two variables. S c shell is is really nice because we don't have to, like, think about it. But they could also we could just say, like, make a dot end file with point it to like the dev API server and the dev auth server or whatever whatever it needs. I forgot what the two are. Mhmm.
Speaker 2:And they can just go run it. With just node or whatever?
Speaker 3:No. They can just call they can just do go run.
Speaker 2:Oh, go run. The the SSH source. Sorry. That's what we're talking about.
Speaker 3:Yeah. Yeah. Go the language, run it. Yeah.
Speaker 2:Yeah. Okay. So but what what if they do wanna work on the whole project, but having people set up a new AWS account or if they've never had an AWS account, that whole process just so that they could contribute to our thing, is it possible that we just give people access to our AWS account? Is that a terrible idea?
Speaker 3:Yeah. It is a terrible idea. Well, I mean that that that literal thing is
Speaker 2:Not like they can log in to the console, but just like some limited permissions that that they can like interact and use ISS.
Speaker 3:I mean, we we think about this at we think about this at ST all the time because it's like a new user general new user problem, like someone wants to try SST but then Mhmm. But they don't
Speaker 2:But
Speaker 3:have to get set up but they don't have the user account. There's like a whole set of things there. Yeah. I mean, it's against there's something to think about. Like, I think there's a way where we can make them because if you think about the components, like, we could make it run all locally.
Speaker 2:I wondered. That's exactly where I was headed in my mind was like, well, what if you just had a little bun served that was like your your router and what if you just had what if you had local versions of all this stuff?
Speaker 3:I mean, if we care about this, we can rework our architecture to like optimize for this a little bit. So for Radiant, I did do that. I like don't I'm not using I'm using like almost no AWS specific services, like it's just a container for the back end. So it can run like very simply locally, like even if it's SSD dev like, it's just bringing up local commands for the for the most part. Yeah.
Speaker 3:So yeah, we can like if we if there's something we care about, we can rework it. I even built, there's another thing I've been working on, this is what I worked on over over the break a little bit, is our next open library, open event. But this is like a durable workflow async events thing that just depends on your database. So if we if we start using that kind of stuff, then like, yeah, there's no dependency on some of these more AWS specific things.
Speaker 2:Yeah. I'm trying to think about our current terminal architecture. Is it it's a router with a bunch of lambda function
Speaker 3:URLs for the API? It's just called no. So we can just swap that into a container if we want.
Speaker 2:Yeah. Could just be a container. Well, it be advanced point. Locally with Docker and everybody's happy?
Speaker 3:Not even with Docker. Just run it just
Speaker 2:But our database oh, database is playing at scale.
Speaker 3:You'd to bring up MySQL locally, yeah.
Speaker 2:Okay. Bring up MySQL locally, yeah.
Speaker 3:So I that's why I wanted to understand better like what I don't expect anyone to work on the back end.
Speaker 2:I don't either, I guess. But like, it'd be yeah. I guess maybe not. Yeah. Maybe the use case is just I want people to be able to come in and work on our front ends.
Speaker 2:We have a few of them. Yeah. And add that's the
Speaker 3:case then. We just need to make that a little bit easier. Yeah. That shouldn't require too much rework.
Speaker 2:Yeah. Okay. Yeah. I don't wanna do a big thing, a big effort just to support a use case nobody has. I don't yeah.
Speaker 2:I can't imagine somebody wanting to change our API and us being like, yes, merge that PR. It sounds far fetched.
Speaker 3:And if they're if they have something that they do wanna do like that, they should be good enough to submit a PR without ever having running ever having run the app.
Speaker 2:Just sight unseen. Just first try. I don't need to debug.
Speaker 3:If you can't do that, like it's unlikely your PR would have ever been good enough for us
Speaker 2:to merge. Good point. Oh, wanna talk about AI coding assistance because something switched for me over the Christmas break where I decided I'm gonna use these things way way more. And maybe you've been using them this much, but like hit the cloud limits every day, twice a day kind of thing. And like just using I don't know.
Speaker 2:Do all the do does like o one pro have this idea of projects where you can just drop a bunch
Speaker 3:of files? That's like Yeah. Could be the projects. Yeah.
Speaker 2:Okay. I
Speaker 3:remember when switched to Claude, that was like right when they added the projects feature and then there was this huge deal.
Speaker 2:So that that thing is important for me because I'm my workflow is to drop like all my current files I'm working on, all the context it needs into the project And then I just ask questions in the project to like make improvements. Yep.
Speaker 3:I I do that with with all of our database schemas because then I can just ask it for like Mhmm. Anytime I need to like have an analytics question, I can be like, oh, how many users did x y z things and it spits out a SQL query. Nice.
Speaker 2:And it gives you SQL queries?
Speaker 3:Yeah. Yes. I actually just give it the drizzle definition so it like can understand all
Speaker 2:And it of figures it out from there. Yeah. Yeah. It's actually kind of incredible some of the things it can do. But then also, it's kind of incredible how bad it can be sometimes.
Speaker 2:Like, can get it stuck in a loop, especially if you have one long conversation and you keep going and keep going, where it tries like 15 times to fix a problem and it just never does it. Like, I just keep going because it's funny and I wanna see like, will it ever figure this out? You you definitely have to figure out like when to leverage it and when not to because there it's I haven't figured out the like what problems are just not good for it and what ones it can nail and do them so quickly. I don't know if you have any secret sauce there, but like the process of exploring leveraging it way more. The workflow is kind of clunky like Yes.
Speaker 2:Copy and paste the stuff around. Yes.
Speaker 3:That's what I wanna talk about. So there's so I have a few thoughts about this. So one, someone linked me to a CLI tool called Glimpse. It was made by someone on Twitter, I forgot. Works on Dingboard.
Speaker 3:Mhmm. It's a simple CLI where you just do glimpse and then a folder. Mhmm. And then it will scan all the files and organize and like understand the file tree everything and just put something in your clipboard that you can paste right to Mhmm. Any LLM.
Speaker 3:And then you can ask your question. So it'll like tell the LLM like, here's the file structure, like here's each file and like it makes sure that
Speaker 2:Oh, interesting.
Speaker 3:It'll like cap it if you're like exceeding the token amount and stuff. So Mhmm. That's like one thing to help smooth that out a little. But here's actually what I think should exist. There is a file system API in the browser.
Speaker 3:So instead of us like having to like upload files to this project's concept
Speaker 2:Mhmm.
Speaker 3:I should just be able to share a folder Mhmm.
Speaker 2:Yes.
Speaker 3:Like, give permissions for this app to access this folder so it can read and write to it. That way, when I'm asking the LM like, hey, fix x y z things, it can just write those changes. Yes. To the file and I can go back in my editor and it's right there. It's not like going through git or anything.
Speaker 2:No. Right.
Speaker 3:So I think that's like a really cool workflow that I actually
Speaker 2:What were you saying about the browser though? What how the browser come into play? Oh, you're saying like if you're using you're using Claude in the browser is what you mean, like Yeah. Like the I gotcha.
Speaker 3:There is an API in the browser for a user to share a folder.
Speaker 2:A folder.
Speaker 3:Yes. Okay. With an app. Yes. And the app could read read read the contents of the folder.
Speaker 3:So it can like use that and like load context in x y z, whatever.
Speaker 2:Yeah. Where you can just ask questions, it can write directly to buffers and then or to files and then you don't have to have this whole back and forth copy paste job.
Speaker 3:That would smooth out the
Speaker 2:clunkiness Oh, of it'd amazing.
Speaker 3:Yeah. I did see one app that had that. That was it says with all these things, it's just like they build this giant solution for like the company they wanna build and all of it is like roughly wrong. But then they'll have like one feature where they like really nailed it that nobody else Mhmm. And I saw I saw that in in one of one of these products.
Speaker 2:So I'm really excited about this idea of like this workflow of like, I'm an engineer, I'm in charge, but I'm leveraging LLMs to like think about my code base, make improvements, add new features, fix bugs. There's also the Devon thing which is exciting too, like I'm excited about
Speaker 3:I churn from Devon. Oh I try really hard So to make it
Speaker 2:I thought this could be like both. It could be like they both live on and like the the these agents are kind of doing these agentic workflows and submitting PRs. That's an interesting input output for me. Like, it's not stuff I'm actively working on. It's like out of band.
Speaker 2:I thought that could coexist, but I'm more excited about this idea that like I want to improve, yes, the workflow aspects of what seems like a very powerful mental model where the LLM has awareness of your code base and it's improving files or suggesting changes. So, yeah, I feel it's gotta get better and I don't wanna use cursor. Everyone is like, that's what they're doing with cursor. But like, I don't wanna change my IDE just to have this. It needs to be able to work with any LLM and it needs to be able to work within my text editing workflow.
Speaker 3:That's why I think so that's what's great about dev. So I think think about this two different approaches. There is Devon, which is I think if I'm gonna if I was to build a company in this space, like, think that's the right thing to try to go for Mhmm. Because Devon is basically saying, you don't need to change anything about anything that you're doing. Yeah.
Speaker 3:We're we're trying to build something that looks like a human. So if you already set up to collaborate with another human
Speaker 2:Mhmm.
Speaker 3:This can work for you. None of your workflow changes like they submit a PR. And second thing is it's unsupervised. So you give it a task and it's supposed to come back
Speaker 2:Yeah.
Speaker 3:Hours later. You're not going back and forth with it every every like thirty seconds Yeah. Like you are today. But I think that's great. The reason I churn from it is the results just aren't good Yeah.
Speaker 3:Enough where it it still feels like I have to watch it and get the feedback very very frequently. Mhmm. So I think the raw power just isn't there. But in terms of Yeah. Shape of the product, I think it it's it's really great.
Speaker 3:Then you
Speaker 2:Do you think that can coexist? Like, in the long run when that improves, do you think that is a a use case, but you also have this like supercharged local Yeah. Development workflow?
Speaker 3:Well, I I think if that exists, that throws into question so much because that that thing is scalable. Like you can spin up a thousand of them if you want.
Speaker 2:The other
Speaker 3:thing we're talking about today, is more like help me as an individual be more efficient, like, yeah, you can like improve that but you can't horizontally scale that because it's so depending on you the person.
Speaker 2:Yeah, sure.
Speaker 3:So I think if Devin crossed that bar, it's a high bar potentially impossible. So like they might not cross Yeah. Which is why there's this other class of bets which is let's just make let's augment like the human. Mhmm. And I feel like this category is like the product execution here is like horrible.
Speaker 3:Mhmm. This is what you're saying, it's like so clunky. They make you massively shift your workflow.
Speaker 2:The fact that I'm still doing it even though it's that clunky though speaks to like this needs to exist.
Speaker 3:The potential. Yeah. Yeah. Exactly. So like, yeah, just make it so That's why I'm excited about going to the file system because that is a true common thing that exists across the board.
Speaker 3:So I would like to see something like Bolt or there's another one we're looking at here, Lovable, I think it's called.
Speaker 2:What are what are these? I've never heard of these.
Speaker 3:It's like v you've tried v zero, I'm assuming. Oh, Claude is like a general purpose LLM Yeah. Whereas these are like more focused around code. So they theoretically the product experience around the LLM should be like better than using a general purpose thing like Claude. Yeah.
Speaker 3:But where they're lacking is they just because it's not this unsupervised thing, I'm like needing go back and forth with it, which is the point. You need to make that back and forth process like very seamless and they don't, for some reason, do that.
Speaker 2:Yeah. So okay. That just reminded me there is a a third workflow thing that AI is doing for us and that's the like I mean, it's the copilot. What do we use?
Speaker 3:It's paid Super Maven.
Speaker 2:Super Maven. I use that as well because I do whatever you do. And then somebody has been recommending where is that tweet? I'm on Twitter as we speak. Hang on.
Speaker 2:Here we go. A bunch of local models people recommended for like, that are tailor made for programming. Things like I'm gonna find it. I swear. DeepSeek, and there's another one somebody said Quinn 2.5.
Speaker 2:So there's these models that are, like, specifically better Copilot models, I guess, the open source models. That that's like a whole another class. It's like line by line. I want AI to just predict my next characters. But then there's like this Claude projects, like broader context on the whole code base, rewriting whole files, refactoring across files.
Speaker 2:There's that thing and then there's the like agentic Devin maybe never exists thing. It's interesting how AI within programming is like many things.
Speaker 3:Yeah. Yeah. There's so many different ways you can you can go for it. Yeah. Devin's like the big bet and the rest are like incremental.
Speaker 2:Yeah. I mean, is Devin like the we don't have a job if it does work, like who cares? We're not gonna even be around?
Speaker 3:I don't I don't know. I don't know if it's that crazy but it definitely extremely disruptive. Yes. And we kind of rethink what our jobs are Yeah. In that case.
Speaker 2:Well, I really like the idea of this supercharged version of programming. It just It allows you to do things that I wouldn't try to
Speaker 3:do. Exactly.
Speaker 2:It allows me to take on tasks I wouldn't even try to do on my own because I don't have enough context. I've not dabbled in that thing enough. I don't know the high level like how you can go about solving that kind of problem. I just feel so brave to go out and do big crazy stupid stuff and it's because Claude can give me a first pass that's like amazing. It's it's incredible what it can do.
Speaker 3:And you can like you can like create a course for yourself effectively to get up to speed on Yeah. On any like obscure thing. As a I mean, I say I think it like just massively lowers a risk around attempting any feature. Mhmm. And it's so clear that the ST Multiplexer and ST Tunnel, both are like really useful good features that are like maybe core of ST now.
Speaker 3:Neither would have existed pre LM. Just we we wouldn't have even tried it. It would have felt like too weird or crazy of an idea to take a risk on that we we've because we would have felt like I probably wouldn't have been able to put it off.
Speaker 2:Yeah. It it does away with that question of like, am I able to do that? Like, is this something I should try and take on? I should just leverage something else. Like, no.
Speaker 2:I feel like I can take on anything because it'll help me fill in the gaps quickly.
Speaker 3:Yeah. OpenAuth also kind of falls in this category because previous iterations of it, I always hit a wall of like very obscure questions of it'll always be something like, why shouldn't I use a JWT for the refresh token or like, why can't I use an identity token for authorization? It was like very specific and nobody has like clearly spelled it out. But interrogating an LM, I was able to like really understand a bunch of things that were kind of eluding me for for years. Another thing is like, yeah, if I sat down and I'm like, the next two weeks of my life, I'm just gonna like master author or whatever, I'm gonna understand the spec, I'm sure I would have been able to do it.
Speaker 3:But this let me like kind of casually acquire that experience, that like knowledge while working on other stuff just like here and there. And eventually hit a tipping point where I was like, okay, now I like fully get it and I can go Yeah.
Speaker 2:Go do it. Yeah. There's a very exciting world where an individual individual developer is doing things not possible before. Like, I used to feel that way about some of the like AWS managed infrastructure stuff. I mean, really what SST is doing.
Speaker 2:I and I still feel like it allows smaller teams to do things that used to be really painful. This is just like a different version of that and it builds on top of that even. Like you can you can still do all those things, but now you can like take on really weird tasks that I can't talk about because it's a secret. I've like rebuilt so many things for this very stupid specific environment that I'm building this. I've rebuilt so many like open source projects in like thirty minutes that like work for this particular case.
Speaker 2:It's really it's interesting Things that I never would have tried to do and like Yeah. Yeah. I just can't talk about it yet. It's so stupid that I can't talk about it.
Speaker 3:This feels similar to I remember like earlier in my career career, my process for learning a new thing, I I can remember this very clearly. I was like, I need to use RabbitMQ. I need to learn RabbitMQ. So my approach was I would just keep googling until I acquired like 10 to 15 articles that like roughly covered everything I had a question on. It was like this process know,
Speaker 2:tabs open, that's what I do.
Speaker 3:Yeah. Process that took a bunch of whatever. And then I realized like I'm just building a shitty book. Whereas if I just if I just buy the RabbitMQ book, like the RabbitMQ book and just read it, that's just someone that pre compiled all of that stuff for me. And then since then, whenever I had like when I was definitely going into like a new thing, would just buy the book on it and just read the whole thing and I could skip this whole let me like kind of scrounge around for for the information.
Speaker 3:And I feel like LMs are kind of similar in that you can just ask it about a topic and it just like pre researches everything for you. You don't have to like go like look up all this stuff. When I was working at multiplexer like, you know, what are all the different ANSI codes that exist? And like what are the common ones and what are the ones that are special that term Like I would have to expect someone to have like documented that exact question, but you know, Lem could just pull that up for me on demand.
Speaker 2:It it just creates a bridge so much better than like traditional Google page indexing. It creates a bridge from like the actual world and everything that's been done so far to your brain and what your context is. Like it creates this amazing map that like is so hyper specific to what you're doing and where you're at. Whereas like googling, you'd have to go from like yeah. Like said, like 15 different sources of information and you're trying to cobble it all together in your head and make it all make sense for your context.
Speaker 2:The LLM just bridges that gap so nicely.
Speaker 3:And and even going back, like I would even say that multiplexer thing was the first thing I did that was so heavily LM assisted that I didn't even do it right. Like I was still using some of my old habits of being like, I gotta find a library that someone implemented that kinda has what I want and I'll go from there. Yeah. And that's what the current multiplexer is, it's something I found that's like kind of working and then I use LM to like add all the stuff that I needed. But looking back, I should have just gave up that first thing.
Speaker 3:I should have been like, I can just implement this from scratch because Yeah. The LM would have made it really easy. Yeah, my ordering was all wrong. Like I tried to find it, then I used LM to help me like understand what this kind of thing even does and I like to and I filled in the gaps. But I could have just had LM explain to me everything about how terminal stuff works upfront
Speaker 2:Mhmm.
Speaker 3:Then implemented it and then I would have been in a better place. And it's probably what I'll do for a v two of it.
Speaker 2:Yeah. How weird if you would have like told us a year ago that we'd be spending so much time building like terminal things. Like just how much like we've had to like build two e's and like environments that run two e's. I mean, you're doing like the next layer down. It's just wild how much our careers are weird right now.
Speaker 2:We're not building React apps.
Speaker 3:I struggle with this because I like I can't tell if I'm just heavily in some weird bias because I'm looking at a bunch of things and I'm like, wow, a Tui just like There's so many things that could be a Tui and an aren't and it would be such a good compelling experience. But I can't tell if it's a bias or not, but I'll give you an example of this. So someone tweeted yesterday, just try it as a tea, wow, it's so great, except it sucks that I have to like go set secrets one by one. And like, it's such an annoying process and I agree. It is like a very annoying thing when you add a new secret, you gotta like, do or you make a new stage and you have like five secrets to the thing.
Speaker 3:It's like a pain in And the obviously, I'm like, yeah, SST secrets should just open a TUI that just shows you all the secrets and then it helps you type it in and like makes it very easy. And that would be a fantastic experience. But like, am I just being biased to like doing everything from the terminal or like, is this truly a good experience? And what it it like composes so well, right? Like, we have that standalone ST secret UI that's just for setting secrets.
Speaker 2:Mhmm.
Speaker 3:And then ST dev, that just becomes a tab
Speaker 2:that's Oh, already in your SSD In your dev yeah. In your what's it called thing?
Speaker 3:Multiplexer, yeah.
Speaker 2:Multiplexer, yeah.
Speaker 3:So like, every feature can be as it's like, it can be a pure CLI command, sc secret set x y Mhmm. Or it can then be a two e or it can be embedded in as a C dev. I'm like, this feels so compelling to me And I could to me, I'm like, once I have that kind of thing, I'm like never gonna wanna use something else. At the same time, I tweeted the other day that pissed off everyone. I don't if you saw that.
Speaker 3:People got so mad at this. I was like
Speaker 2:Oh, I didn't see it.
Speaker 3:Makes me sad that people use a Versus Code terminal.
Speaker 2:Oh, I saw that tweet.
Speaker 3:I swear the reason I posted that is because when people use SSD dev in the Versus Code terminal, there's a whole set of things that are buggy and don't work. So for me, I'm just like, man, it sucks that they don't get like the full nice experience out of it. And people, I would say understandably, interpreted it as you suck for using the Versus Code terminal, which is fine to interpret it that way, but then the number of comments I got replied being super pissed is the part where I'm like, you're just an idiot. Like, you guys really went so far away to get to like really be bothered by this. Yeah.
Speaker 3:So on that hand, and a lot of people replied being like, don't get it. Like, that's just where you like start your your program. Like, why would I like Yeah. What do you mean? And so then I explained like, my issue is that it's not that like integrate terminal in your editor makes a lot of sense, I get it.
Speaker 3:I just wish it was better. And they're like, what do you mean better? Like, how could it be better? Like, all I do is run like basic commands there.
Speaker 2:I run npm start or whatever, yeah.
Speaker 3:Or like, you know, it's just where you like do git commit. So yeah, part of me wonders like, am I just making this up or is it that if I built good experiences then people would be like like more people would actually use use it? I can't tell which way. Yeah. But
Speaker 2:But that that reminds me that Ghosty came out and like now I'm I'm realizing like it's they became the thing of the day. Everybody's talking about like how awesome it is. And I realized like so many people don't really take advantage of it or they're not really like They're like, yeah, this is cool. But like you said, if they're just running git commit in it, then like Yeah. So it wasn't like
Speaker 3:And why it's so much Yeah. So hard to use
Speaker 2:Like they wanna be part of using the cool new thing but it's like they're not really benefiting that much if they don't use the terminal much. And I'm not saying developers have to use the terminal a lot. I feel like it made my job a lot more fun getting into that kind of stuff last few years. Whereas before I would just made changes in Versus Code and get committed. I think in the last few years, I've dove more into the like programming is fun, but like I need more hobbies.
Speaker 2:And that the terminal kind of became that, I guess.
Speaker 3:I mean, I I just I like don't really when people the thing I can feel confident about is at the end of the day, you have to bring up your back end and you have to bring up your front end. Like that Mhmm. I would say like at minimum you're dealing with that if you're like a typical web developer. Yeah. That means you need two two things.
Speaker 3:Like, even if you're using Versus Code, need like two tabs or two whatever. And to me, I feel like that is maybe the good sign that like this stuff is useful because I think you can like anyone can understand how like a multiplexer makes that nicer. Yeah. And once you have multiplexer, you can see how adding more things in there makes it nicer. You can see
Speaker 2:how Yeah.
Speaker 3:Bringing up everything at once makes it nicer. Mhmm.
Speaker 2:Oh, I love the SST dev experience. I mean, I've gotten so used to it now.
Speaker 3:And then Aaron's been doing his thing, is like the exact same thing and he's done a really good job. I don't know if you've seen the stuff he's worked on solo.
Speaker 2:What is it?
Speaker 3:It's like the same thing as SST dev Solo.
Speaker 2:Solo? Oh, Aaron Francis. I'm not gonna lie. I was like trying to figure out who you're talking about.
Speaker 3:There's a lot
Speaker 2:of Aaron I love you, Aaron.
Speaker 3:Out there.
Speaker 2:Yeah. No. There's just so many Aaron's.
Speaker 3:Yeah. It's not related to us. Like, he like built it totally independently running across the exact same problem and like landed on like very similar solutions. And his implementation is very good. Like, we're gonna we're probably gonna steal a bunch
Speaker 2:of ideas Tell me what it is really quick because I think I saw him mention it, but I think I missed, like, the announcement or something. So I don't know what it is actually.
Speaker 3:It's it's the same thing as ST dev just for the Laravel world. So it, like, understands your Laravel app, brings up your front end Laravel, it'll like put your logs in a separate tab. He has this really cool feature which I think I'm gonna steal. He has an all there's something called dump in PHP. It's like it's effectively it's like a special log command.
Speaker 3:It's like a user only when you're debugging. He pulls those out into its into its own tab. Nice. So I was like, oh, it would be cool if I like if you did console log something
Speaker 2:Mhmm.
Speaker 3:It wouldn't show up in the main functions tab in SCTEP because that's like a mess. It got pulled out into a separate tab. So if you're like debugging something, you can see like a less noisy tab with just like your your debug statements.
Speaker 2:That sounds nice.
Speaker 3:He also made it a horizontal thing at the top, which I considered at first, but I was like, well, you're gonna run out of space. But then he made it scrollable. I'm like, oh, that's actually perfect.
Speaker 2:Like tabs, you mean? Like horizontal like
Speaker 3:Yeah. Ah.
Speaker 2:Now, what what did he actually make? I'm so dumb when it comes to this stuff. It's like a multiplexer like you made?
Speaker 3:It's a multiplexer, yeah. It's team It's like a specialized team mux. That's all Like, that's what SC Dev is, that's what Zero is, that's what Sorry, that's what Solo is and that's what and Turbo Pack also has the same thing.
Speaker 2:Oh, yeah yeah yeah, okay. I was a contributor to Turbo Pack way back in the day.
Speaker 3:Oh, wow.
Speaker 2:Back in my Next. Js days, you know. That was Golang. I guess that was my first exposure to Golang. I had written Golang.
Speaker 3:It's not Golang anymore, it's not on Rust, but
Speaker 2:Oh, well, of course. But Yeah. That's a whole another conversation. I actually I'm not gonna lie, I don't know what's going on with Rust right now. I think Prime I heard you guys talking about it on call yesterday.
Speaker 2:Like, the Prime did something and now Rust is a main topic again. I haven't even caught it.
Speaker 3:I actually was gonna post about this today. So Prime made a it was funny. I remember there was like this week where he was like, oh, should I just like write this in Rust or should I like just kind go back to writing Rust for this thing? And then everyone in the chat was convincing to do it. So he was like, fine, I'm gonna do it.
Speaker 3:And then day later, he made a video of being like, I'm done with Rust. And I watched the video and you know, it was like a really good reasonable video. It wasn't like Rust sucks, you shouldn't use it. It was just like, he's used it a bunch, he's been in that phase where he loved it, he's been in the phase where he like was really enamored by some the ideas about it. And now he's in the phase where he feels like, you know, it's just really really really really heavy for what he's getting out of it.
Speaker 3:And he's just like not super happy with where it actually ends up practically.
Speaker 2:Mhmm.
Speaker 3:But I think I think it's very reasonable and like he's like, I just want something simpler that is still low level, that feels better. He's like, I'm just not having fun writing this anymore. It's like
Speaker 2:Yeah.
Speaker 3:The the parts that I enjoy it don't make up for the parts that, you know, are painful. But I was like, this is very like reasonable and I I kind of that's it. That was like also my same feelings around Rust which is why I don't I like don't really use it for anything anymore. But like, people got really bad about it. And like, people are like just saying all like, I saw the reason I was thinking about it is I saw a comment being like, he doesn't understand Rust.
Speaker 3:That's why he's saying this. Like, he doesn't understand it. He's not really a Rust user. I'm just like, no, the stuff he brought up is like you can't get to there unless you've like been through the whole journey. So clearly Yeah.
Speaker 3:Like he he does. But yeah, people are upset for whatever reason.
Speaker 2:Is it like Prime, I like Prime, he's a friend, he's a co founder. Is it biased that I have this memory? Did he not like, is he not like a huge reason Russ became the zeitgeist? Like, all the like trendiness?
Speaker 3:Yeah. He was he he was so obsessed with it to a point where I found it annoying. Like, literally found it annoying. And I
Speaker 2:was like Yeah. Like, when people talk about Rust being this like cool thing that people use and it's mostly like streamers, that was mostly Yeah. Prime. Yeah. Yeah.
Speaker 2:I didn't know if there's people before Prime that I don't even know about that were like shilling Rust and making it cool. But I feel like he brought he giveth and he taketh away. He he brought Rust into prominence and then he said, you know what? Not that he brought it into prominence, that sounds wrong. Like, I'm sure lots of people use Rust for lots of great reasons and they ship stuff on satellites and doing whatever they're doing.
Speaker 3:It definitely created a lot of interest in Twitter Like the
Speaker 2:people we hang out with, the YouTube and Twitter and like very online community got very obsessed with Rust and that was a lot of primes doing and then Yeah.
Speaker 3:And I'm sure that there's like genuine cases where someone wouldn't have heard about it if not for that and they ended up using it in a very good situation. Like I think about my years with Elixir that were so formative to like me understanding distributed systems and like just so much impact like functional part and a bunch of stuff that impacted me. And it was because it just showed up on Hacker News like three times in one week. I was like, okay, now I have to look into it. So I'm sure that that like definitely happened.
Speaker 2:Interesting. This one's been long. Have we gone a long time? I can't really read the numbers, they're so small.
Speaker 3:It's been one twelve. Oh, We've been off for for two weeks. We gotta
Speaker 2:know, let's just make let's make it three hours because it's been three weeks. We're gonna make up for the other two episodes. I got nothing going on. Let's do it. No, I'm just kidding.
Speaker 3:That's what's going on with that and he's like looking into Zig now. And I've been looking into Zig a bunch too, just coincidentally in the past couple months. And, yeah. I think the one thing you can make fun of Prime before, which is what everyone should focus on is he falls in love very quickly and gets everyone tells everyone that they should fall in love too. Yeah.
Speaker 3:And then and then he like, maybe falls out of love.
Speaker 2:Yeah. It's the it's what's that cycle? The hype cycle chart? It's
Speaker 3:Yeah, like exactly.
Speaker 2:He does that very fast.
Speaker 3:Well, TJ made a really funny meme today. He he used that Grim Reaper meme, the one where
Speaker 2:it like
Speaker 3:goes into each thing. And his prime like killing each language with his love for it and then eventual hate
Speaker 2:for it. Yeah. Hopefully, Zig can withstand the prime train.
Speaker 3:Yeah. We'll see.
Speaker 2:Yeah. We'll see. Okay. Well, I do have to pee and that's it's the new year but I still have
Speaker 3:So I know first because you told me that you peed right before we started. Yes. So you'll literally have to pee. It's been one hour
Speaker 2:Way more frequently. You understand me going an hour and thirteen minutes is like long. Like I pee every thirty minutes in the morning after I've had coffee, it's
Speaker 3:I peed this morning when I woke up.
Speaker 2:That's it? I probably have like a prostate issue or something.
Speaker 3:And it's it's been like five hours and I won't pee
Speaker 2:That's it.
Speaker 3:Probably for the rest of the day.
Speaker 2:Oh my goodness. I get up every thirty minutes to an hour. Yeah. Well, not second half of the day. It's like something cools off, maybe I just stop drinking so much water or the caffeine.
Speaker 2:I don't know. Something happens and then like second half the day, I don't pee much. First half of day though.
Speaker 3:Full of urine. Just urine everywhere. Just lots of pee. Oh, yeah.
Speaker 2:No. Because I totally stand when I pee like a normal US adult man. I don't. I sit. I don't care.
Speaker 2:Whatever.
Speaker 3:Is it is it is your are your toilet seats heated?
Speaker 1:They are. I knew it. I love it so much. Oh my I knew it.
Speaker 2:I'm not even No
Speaker 3:one would know to ask that question, but I knew to ask that question because I know you.
Speaker 2:I think my toilet seats are one of my favorite possessions. Like, sitting on my toilet seat is such a moment of joy every single time. Because it's like every single time I forget. And it's cold right now in Missouri, it's like 20 degrees, and I I'm going to the bathroom because I gotta pee and I sit to pee like a normal European male. And as I'm getting ready to sit down, it's like, oh, yeah, the toilet seat is heated.
Speaker 2:Oh, I love it so much. It's the best feeling. Recommend Yeah.
Speaker 3:So my toilet seats are never cold except for this week. So this week, I have cold toilet seats.
Speaker 2:Yeah. That 50 degree toilet seat. You just buy one, man. They're the best. Mine's it's Toto.
Speaker 2:It's good.
Speaker 3:Franklin's telling me to buy a heater for my room. And I'm like, I don't think it's worth I don't want like some random ass heater just in my room Yeah. For like
Speaker 2:Just get like a just get like a big GPU rig and start like doing local LM stuff and it'll be heating everything up. There you go.
Speaker 3:I I just really want I want like just a heat lamp over my keyboard so my my hands are just Keep
Speaker 2:your hands Yeah. That would be I actually did have like a desk mat that was heated. My wife got me once trying to solve the hand thing. I don't know if it solved it or not. I guess not.
Speaker 3:But They should make keyboards that are hot.
Speaker 2:Oh, the keyboard should be hot. Of course. Of course. Great point.
Speaker 3:Yeah.
Speaker 2:Okay. I really
Speaker 3:can't Alright. Alright. See you. See you.
