AI Coworkers, Dax's Twitter Rules, Planetscale Insights, and Adam's Health(care)
We should put it at the beginning of our podcast.
Adam:Oh, probably. So that everyone who hears the podcast because there's probably, like, 10 people that listen the end. Yeah. I
Dax:don't know where I saw it. I think maybe it was just something like a YouTube thumbnail, and I inferred it. But they put it up here, like, where I where I have it now, and they do that thing where they take a picture of their background.
Adam:It's so dark. I can't, I can't see your microphone. Oh, it's like up high?
Dax:Yeah. So it's like, it's like to the side of my head. And they do that thing where they take a picture of their background, and then they just overlay it over the mic. So it looks like there is no mic
Adam:Oh.
Dax:In the frame.
Adam:They can't so they can't move their face.
Dax:They they have to be completely still, and
Adam:they have
Dax:to be, like, right here. Well, I mean, he I so my thing with this mic is I feel like I have to, like, be right up. I feel like I have to, like, make out with it when I talk. Yeah. Otherwise, I don't really get a good sound.
Dax:His was different. It was a little bit further away, so he had some range of movement. Yeah. That's a clever trick.
Adam:It is a clever trick. I've got the I've got a, shotgun mic.
Dax:The shotgun. That's not good for this type of thing. Right?
Adam:It's not good for the pocket. Well, it's it picks up a lot more background noise. So Yeah. The my secret with the podcast is I can get less of my children screaming if they use this mic. Because you kinda do.
Adam:You have to make out with that. I think we have the same mic. Right?
Dax:Yeah. Yeah. It's the same one. The SM 7b. How many of these things do you think they've sold?
Dax:Because I see them everywhere.
Adam:Every single
Dax:It was
Adam:like the Michael Jackson thriller microphone. Right?
Dax:Yeah. Like, as soon as that trivia fact went out, every single person with the podcast just was like, that's the one I need.
Adam:I could be Michael Jackson. Yeah. I could have something in common with Michael Jackson.
Dax:But I see it everywhere. I saw I was walking by a a cigar store, which is a normal yeah. Very common in Miami cigar stores. But, like, half of the retail space was dedicated to, like, a podcast setup. And they had, like, the couch, and then they had all they had, like, 6 of these things on on, Wow.
Dax:On mounts and everything. And I was, like, this is so crazy. Literally, everyone and they're not cheap. They're expensive. They're, like, $300 or something.
Dax:Oh, yeah. And it's just the mic, and and it's so needy. It needs so much around it to sound good.
Adam:Yeah. It's like me. Uh-huh. I need dachs around me to sound good. Can you imagine 6 people talking on a podcast, like, at the same time?
Dax:That's a lot.
Adam:That'd be intense. That'd be a lot.
Dax:I've seen clips like that, you know, where, like, it's, like, 22 sides of people, and they're all just talking bullshit at each other. It's just that the dumbest people in the world is arguing with each other.
Adam:Maybe we need to add some people to this this podcast, some regulars.
Dax:We're going to. Sounds like we're going to it on Tuesday.
Adam:Like guests. Yeah. Yeah.
Dax:Oh, I see.
Adam:Like, one of guests. But maybe we need, like, to expand the hosts. Maybe we need 4 hosts.
Dax:Let's try it. Let's try it with,
Adam:with a
Dax:lot of people at once. Anyone with an s m seven b, you're welcome.
Adam:Yeah. Yeah. Got an s m seven b. Hit us up. I saw an s m seven b on a, Sesame Street, like, album cover, like Cookie Monster is talking into 1.
Adam:No joke. I was like, hey. I know what that mic is.
Dax:I was
Adam:like, you're dumb.
Dax:That's funny. Well, we're gonna we're gonna kinda like, for React Miami, we're gonna have 4 people doing the livestream. Yes. So that's gonna kinda be
Adam:all be on it's same it's you, Prime, Theo, and
Dax:Madison.
Adam:Madison. Oh, so it's the same people that did some other conference, but plus you.
Dax:Con. Plus me.
Adam:NextConf. Yeah. Yeah.
Dax:I'm banned from that one, so I couldn't go to that one. You're banned. I'm just kidding. I'm not banned.
Adam:That's funny. Are you I had a question about that. Oh, did you see oh, it was not about that. It was about, is is Prime starting a podcast?
Dax:I saw him say something about that.
Adam:Yeah. And Pirate Software also said, what's the name of the podcast? Are they starting a podcast together?
Dax:Together. I'm so curious. Yeah. Interesting. Oh, man.
Adam:More competition.
Dax:I'd listen. His competition. Yesterday dotfm.
Adam:Oh, shit.
Dax:Got competition.
Adam:Oh, man. I love it. That was that was great.
Dax:Yeah.
Adam:They didn't even change the background of the thumbnail. It was just it's just like tomorrow.
Dax:I thought I even then, I thought that was a way higher effort than again, for people that aren't aware, we have some competition. There is a competing podcast called yesterday.fm hosted by Melky and and Jay, who is the CEO of SST.
Adam:They're not really recording episodes, are they? It was just a joke. It's
Dax:it it a 100% is a joke. But are they
Adam:gonna record joke episodes? Because that would be great, actually. I would love to listen. Yeah.
Dax:This is like the most niche joke ever. Just the the level of effort to people that were together just
Adam:does not
Dax:make any sense. But yeah. I I mean, I even even the thumbnail they made, I thought was higher effort than
Adam:Oh, yeah. No. They did a great job. I mean, it would have been extra mile for sure to, like, make the actual, yesterday dot FM background. Yeah.
Adam:But good job, guys. That was a good one. Yeah. Got a good checklist.
Dax:You got me.
Adam:You got me. You got me. So I I have technical stuff to talk about. Wow. I I don't need to announce that every time, I guess.
Dax:I think I should because it's like I get why you do it because it's it's infrequent.
Adam:Yeah. Okay. So I have a technical topic and a nontechnical topic. I'll let you choose which one first.
Dax:Let's do the technical one first.
Adam:You might have things you wanna talk about too. I don't wanna just say, like
Dax:I don't.
Adam:We can just we can spend, like, 5 minutes each online, and then you could talk for, like, 45 minutes. We gotta give the people what they want. You know? Some smart conversation. Okay.
Adam:So technical first. Do you ever have one of those nights where it's, like, one long dream, but you also feel like you're half awake and you're thinking really hard, like, in your sleep? Do you ever have that? It's like where you're working through something.
Dax:Extremely rarely.
Adam:Okay. I mean, it's not often for me. I wouldn't say, like,
Dax:the stuff that I I can relate to it. I get what you're talking about.
Adam:You kinda know what I'm saying. The other night, for whatever reason, I think it was the day after the Carmack tweet.
Dax:Mhmm.
Adam:And, Carpathi invested in
Dax:that Magic.
Adam:Magic. Yeah. I think it was the the night after, like, those two data points. And what's funny is, like, saw the tweet and saw you quoted it, saw the Carpathian investment, and kind of read up on Magic. Didn't even think about it that day.
Adam:Like, it just kind of, interesting AI stuff happening. Didn't it didn't really, like, click in my mind or make me have thoughts of existential dread or whatever as a software developer. But that night, it's, like, all night long, I was thinking about this idea that, like, we're definitely gonna have like, in the next few years, we're gonna have, like, coworkers that are AI. Not coworkers, but, like, we're gonna have, like, end to end, like, describe a feature, get a PR from these, like, AI developers. I mean, just large language models, whatever they are.
Adam:Some form of the current version of AI, like, making changes in our code bases, that's definitely gonna be a thing. I don't know. There was, like, maybe 6 months ago or whatever, it was kind of like, oh, that could happen in our lifetime or, yeah, maybe, or it'll just make us all more productive. But, like, it's very clear to me now, and maybe it's been very clear to everybody for a while, that, like, this is gonna be a thing where there are junior developers replaced by AI that's just pushing PRs. And the more I thought about it in my sleep again, I don't know why I can't just If we could, like, lower the cost of implementation to, like, If we could, like, lower the cost of implementation to, like, trend towards 0, that's the part I actually hate.
Adam:I mean, like, I love coding. For fun, that's fun. Sure. But, like, being an individual contributor is tiresome. Mhmm.
Adam:Like, I've done it for 15 years, and the thing that sometimes I wish, I just wish I enjoyed managing people sometimes because I don't. I don't like it at all, and I've spent periods at my startup being a manager and hated it. But I think I might love it if it were just a bunch of little AI things that I'm just having to, like, give it the right, you know, prompt or whatever you wanna call it, give it the right direction, and then it's just handing me PRs to review. That sounds actually super productive and, like, an amazing thing for society if one developer could kinda have, like, a little legion of or just, like, one instant junior developer that can just literally, like, churn stuff out for you. It just made me think, like, this is gonna open the floodgates to what we can accomplish and how much faster a company that has direction and knows what they wanna do can do it.
Adam:So it's actually like it went from, like, very scary, I'm a software developer. They're teaching robots to do, like it's like the self checkout thing. Like, if you worked at Walmart, uh-oh, and it was that feeling as a developer, like, it's coming for us, to now it's like, no. This is amazing. We're gonna be able to do so much and so fast.
Adam:Like, as someone who runs a company, works at a company, and has developers on my team, I can only imagine what we could all do if we all had access to this kind of thing. Sorry. I've talked way too long, without letting your smart brain react and correct and do the things you should do right now. I think it's 2 AM, by the way. I didn't sleep well last night either.
Dax:Oh my god.
Adam:Not related. Totally unrelated. But, yeah. Anyway, that's that's my thought.
Dax:I totally agree with you. It's this weird thing where I get why a lot of people are anxious. And to be honest, I think a lot of people should be anxious. If I'm gonna be completely honest with, like Yeah. Like, frank about this stuff Mhmm.
Dax:There is the reality is there's a type of person that is trying to do stuff, and they learn a bunch of skills to do the thing. Then there's a type of person that learns the skills because other people are trying to do do a thing. If you're in the latter group, yeah, I see why this is a problem because Yeah. The thing you described where, like, you have all these things you wanna do. And I totally agree.
Dax:Right? Like, we always have ideas, and it's always, like, the, like, process to get to a version 1 of it so you can, like, actually have a tangible opinion of your idea, it's a such a massive cost. So anything that lowers that for someone that has ideas and then someone that's trying to, like, put ideas out in the world and to build things, this is obviously great. And we've been experiencing some portion some version of this forever. Like, the cost of that is slowly lowered over time.
Dax:Yeah. It it required learning new skills, required, like, figuring out new things, and it seems like we're on the cusp of it lowering basically to the lowest it could possibly be without having to, like, really learn too much. Yeah. So I think, yeah, for people that are trying to do stuff and programming is a way to do the thing, and that's if that's the way you perceive the world in your work, I think, yeah, your conclusion is exactly the right one. And it's kind of how I feel too.
Dax:The little details I do wonder about are so that company, Magic, is trying to make it so that, what's the best version of an AI programmer? Right? How can we push this concept Mhmm. To infinity? The thing is this stuff could be so disruptive that that is actually exactly the wrong idea.
Dax:That presumes that you still need software to be built and written. Right? It could there is like a version of real of how this all plays out where the idea of software goes away. Right? If the idea of software goes away and AI is the final software
Adam:Mhmm.
Dax:Final piece of software that needs to be written, Like, why do you need an AI programmer? You get what I'm saying? Like, Sure.
Adam:Yeah. Long term, I see what you're saying.
Dax:If there is an AI thing and it has access to a database and it can read and write from the database and figure out everything that needs to be done just in time, writing software is just kinda like caching that behavior in a way. Right? Like if you can figure out just some time, maybe it costs too much. So then you like codify that into a bunch of rules and you like hard code it as software. But there is like a future where a lot less software is written and it's possible that, you know, there isn't really a business to like make an AI programmer.
Dax:And I kind of believe that because if you look at the SaaS category, which is where a lot of people in our field have jobs, they work at companies that build tech for other companies. Yeah. I'm gonna call them they build tech for real businesses. So there are certain things that need to happen in the world, and tech companies just build software that makes that happen faster. And most of us have jobs in that field.
Dax:That field could just go away, right? That doesn't mean every single industry goes away. It's really just that field of helping companies be more efficient goes away. Mhmm. So, like, let's say you have an oil company that, like, you know, drills oil and does a bunch of stuff.
Dax:That's still gonna be around, like, we still need that at least for some amount of time. But they don't necessarily need to be buying software. They don't need maybe they don't need to be buying, like, a 1000 pieces of software like they do today. So I kind of wonder if the way that people are going to be pushed out of work isn't really that directly that they're going to be replaced by a programmer by the company they work at. Either company might stop making sense at some point.
Adam:Yeah. I remember your tweet now. We even talked about it, I think, where company you work for is more likely to go under than you are to lose your job.
Dax:Yeah. So then if you look at what you were saying, if the thing you want to do is something that exists outside of tech like, it like, the thing you wanna do can't be, like, I wanna build SaaS software. Right? It has to be something different. Like, I wanna, like, I don't know, like, make this good and, like, sell it on this website.
Dax:Mhmm. Then, like, yeah, AI can make that cost, like, basically 0. I think that makes total sense. But if your idea is, like, I wanna build productivity software for this group of people, like, I don't I don't think the AI helps you there. So I think AI would just do that job.
Adam:Yeah. So do you think what you're saying is specific to the the SaaS, like, b to b kind of stuff, like productivity software?
Dax:Because Yeah. I think so.
Adam:I'm trying to think like, so, I I mean, I can't not talk about Savvy's. I try not to, like, get too deep into, like, my specific situation.
Dax:Mhmm.
Adam:But, you know, if you have a consumer facing website, it's kind of like, website slash media thing slash just information, news. It's think like Wikipedia, Yahoo, Finance. I don't know. If you have a site like that, I guess AI is already serving up a lot of statmuse stuff. So perplexity being an example.
Adam:Like, they're kind of a new aged AI based search engine. But they do what Google does, which is when you ask sports questions, they reference our stuff. So we still have this kind of, like, resource that's valuable to the AI, I think, and that we've curated this very, like, specific database. But I guess also like the long tail here, I feel like you still got, like, most of the world on WordPress and, like Stuff
Dax:is very slow. Yes.
Adam:It's very slow. And, like, if you think about how slow things go, do you like, I guess what I'm getting at is I could see the AI programmer thing being useful for 5 to 10 years while, the the in state that you're kinda describing
Dax:Yeah.
Adam:Takes shape and and and becomes a thing.
Dax:Yeah. I think practically you're right. Like, that's what I'm describing is, like, an overly simplified scenario where usually I'm just, like, talking about what the final thing is and there might be, like, millions of opportunities. It might take 50 years, who knows, like, on the pathway to that. So I'm not saying, like, magic can't work out, but right now, if I'm gonna make, like, a wild ass bet and start a company like that
Adam:the big if you're a venture like, looking at it from a venture perspective, like, the big giant bet, then maybe that's not the best. It's kind of an intermediate step that you're betting on.
Dax:Yeah. Like, if I could build anything right now, it probably wouldn't I probably wouldn't be trying to do what Magic is doing. It it it kinda works in 2 ways. Right? Like, what's great about their business, it's they're such an incremental step.
Dax:Like, people companies are already producing a lot of code. They already have, you know, people that, they hire to do that. They have workflows and processes, and they can, like, incrementally adopt this stuff. So from a business point of view, like, makes total sense. And, like, I get why there's a lot of opportunity there.
Dax:But again, if it's like a moonshot type of bet, that one doesn't seem that in in a lot of ways, it seems, like, too practical and, like, too likely to work to actually work. Do you know what I mean? Yeah. It's like a real paradox. Yeah.
Adam:But if I look at it as, like, a a potential customer, I'm very excited about it. I hope they succeed because I I would love to use it and pay for it. And and, like, across multiple thing like, not just at, you know, my day job, but also, like, every side project I ever think of. Just being able to, like, leverage something like that and not spend so much time implementing, sounds very attractive.
Dax:Yeah. A lot of things change. Right? Let's say that thing is really good, and you never have to directly manipulate or read the code. You just have to Mhmm.
Dax:You would do is still iterate on it, and you maybe iterate on it in a very familiar way. It's a way the way we iterate on software right now. But you you don't have to do the direct part. What does the best programming language mean at that point?
Adam:Yeah. So I think we've kinda talked about this, and and I think the it's the in state thing too again. Like, with the in state, it's like, sure. It should just be spitting out binary or assembly language or whatever. Right.
Dax:I don't know.
Adam:Maybe WASM has a a point in the fur in the after all. Like, there's a reason for WASM to exist. I don't know. But in the intermediate state, it it cut it kinda be nice if it was spitting out TypeScript or whatever you're used to reading so that you could actually review it and and catch stuff. Like, as it's getting better, I could imagine there might be things that it does that you're like, oh, I didn't understand exactly what I was going for here.
Adam:And maybe you fix it directly or you give feedback and it responds to the feedback. I don't know. I could see that intermediate state benefiting from it being a language you're familiar with.
Dax:Yeah. That's at the immediate inter intermediate state, but you can see how you might be more willing to choose a language with different properties. Right? If you if you feel like I guess you start using it this way, it turns out you're, like, really just editing and tweaking the stuff it's outputting. You might be more than willing to give up a familiar language.
Dax:You might say, like, I'm down to use a more, like, performant language. I mean, it's a little harder to iterate traditionally. Even with the new SSC CLI, when I start and we wrote it in Go. And I'm familiar with Go. Like, I did it for years, but I had it in a while, so it was rusty.
Dax:And one of the things I try to do is okay. I'm starting a new workflow. Basically I'm in a new language effectively, not what I've been doing for the past couple of years. Let me try to like overuse AI. Even when I feel like I can do it, let me, like, try to do it through AI first.
Dax:And just to, like because I just kept I just keep feeling like I'm not using it enough, so this is a good opportunity for me to use it more. And, yeah, you know, AI wrote a lot of the CLI, a lot of the code that's that we're shipping. Yeah. And and it did a good job, and I mostly just, like, babysat it and, like, pointed it in the right direction a lot of times. And, I do feel like sometimes it found better ways to do things than I would have.
Dax:I think it did do things faster. So I I don't like to use the word faster because I think whenever the concept of faster comes up, people are always like, it's only, like, 3 seconds faster. Who cares? What actually matters is it took me less energy. It definitely takes me less energy to tell chat gbt something and have it spit it out.
Dax:Yeah. It's a little awkward at first because you're a little more familiar just typing the code versus describing the problem. But once you get used to it, it is easier. So yeah. And then so if you take this one step further, okay, now you're at the stage where people are willing to relax their immediate feelings about languages because they're not the ones directly manipulating it.
Dax:Now if you're a language designer, what does designing a language for AI mean? Like what does that would you do things differently? Would you care about certain things? AI needs a lot of examples, so, like, that also maybe makes it hard to create a new language now because
Adam:It's like a cold start problem.
Dax:Yeah. Exactly. And it's funny. So we're doing new docs for SST for Ion, and we're putting a shit ton of examples everywhere, not for people, but so that AI
Adam:can learn.
Dax:So we're already changing our behavior.
Adam:That's awesome. Oh, I definitely think of SST when I think of, like, building stuff really fast with AI. It'd be really great to have this sort of integrated deployment picture in there with the code that you're also building. Like, just the idea of using serverless stuff, and be able to, like, literally stand up a startup in, like, such little time Yeah. Relative to the old barriers of entry, kinda goes hand in hand with that whole movement, I think.
Dax:Yeah. And that's why I think what is so that's really exciting. The part that I think we all have to adapt, at least for me, because I've always worked, like, in tech tech. Like, I never worked at, like, a real company. I've only worked at a company helping real companies.
Dax:Yeah. So my whole Yeah. Exactly. So my whole mindset I think I need to shift it a little because I don't think selling shovels is gonna be as viable. I think I need to, like, do the actual digging now.
Dax:Yeah. Yeah. Because like the ultimate trouble has been made in some ways. And even with SSC, you know, I think about this where our whole thing is creating a nice abstraction for people that is simple and progressively discloses into something more complicated. That takes a lot of work and a lot of constraints and a lot of trade offs, but if an AI can just in time always generate the right thing, like what level of abstraction does it need?
Dax:Like, does it need all that? And if the user isn't really directly manipulating this stuff, like, is it okay to sacrifice a whole bunch of things that we currently are not willing to?
Adam:Yeah. I
Dax:I can't
Adam:remember if it was your tweet, or I saw somebody tweet something along the lines of, like, things that don't matter when it comes to AI in a in a programming language. And they mentioned, like, type safety.
Dax:Like,
Adam:it's an interesting thought that, like, things that we value so much for refactoring and for all these things, like, those concepts, those concerns do just kinda, like, It It's super interesting. I can't remember what else was on that list. Was that a tweet of yours, or was that something else?
Dax:Mention that I did have a tweet about how type safety is something less important. The reason that one clicked for me immediately when I started using copilot because copilot's autocomplete sometimes does stuff you, like, get from type safety.
Adam:Yeah. Yeah. You know, like,
Dax:it figures out, like, the right thing, and it's, like, it's actually doing a better job than the type safety.
Adam:I just had the thought today that, like, my Neovim configuration, I can't remember how it's prioritized, but it does feel like my LSP and, like, Copilot are competing, and often I'm choosing Copilot. Like Right. It's it's just interesting that the LSP's already been diminished somewhat because my first tab down, like, it's a Copilot option, and it's it does exactly what I needed to do. It's something the LSP couldn't have predicted. It's just that level beyond.
Adam:Yeah. It's interesting. Yeah. Exactly. Sometimes sometimes I do wish, like, I could just go with the like, sometimes I'm frustrated when I choose the Copilot thing, and it doesn't get the last bracket or something.
Dax:Right.
Adam:Do you have that issue? Like, it does that stupid stuff where then it's like, I gotta figure out what's missing.
Dax:Yeah. Yeah. Yeah. Yeah. I just, like, randomly delete parentheses and it really got the missing words.
Adam:Until it's until it's good. Yeah.
Dax:Yeah. So
Adam:that was nice about LSPs. They're a little more precise or a lot more precise. Right. But they can't do thing. They're just in a totally different class.
Adam:Like, they're completely outclassed in other ways by things like Copilot.
Dax:And there's someone in the in the Twitch chat saying, I think you're exaggerating a bit on how good AI will be in the future. So my my response to this is actually everything I'm saying doesn't require AI to get much better. Even if you take what it can do today in terms of raw capability, every problem I have with it, at least today, is a coordination and workflow problem. It's like, I actually need it to do these, like, 6 different, like, do this and pipe it into this program and then take it from this program and do this other, like, combine it's like, to me, it's an it's like a very much so an an orchestration problem, which is very solvable. So if AI gets, like, 20% better and the orchestration stuff gets a lot better, a lot of what we're talking about, starts to become like a reality.
Dax:So, yeah, in a lot of ways, like, AGI in a form is kind of here. It's just we don't know how to, like, connect it together in the right order.
Adam:Yeah. Yeah. I don't I don't feel like the actual models themselves need to get better for, like, so many things to open up. It it's when we started Samuels, I mean, that was, like, the whole thing was just, like, natural language processing tech was really advanced, even 10 years ago. But, like, there's so few consumer applications of it, and it was just like, hey, we could apply this to the thing we care about.
Adam:And a lot of people love that. And it was just like it's it's the same thing today. I feel like with AI, it's like all the pieces are there. It's just now people putting them together into products that are really useful. And, yeah, that last mile, that orchestration bit, I feel that pain too.
Adam:But, yeah, I don't I don't think we're, like, banking on some huge step function of complexity or stuff that's opened up and the accuracy of the models or whatever. Like, I feel like they're good enough to do, so much of what I do day to day.
Dax:The other thing is okay. So given all this, you know, what actually matters now as like, what does this all require if the input isn't being able to write code well? One of the ways I've been thinking it's funny because Satmuse I think fits perfectly into this. If an AI can do everything, can be useful, and you want to build a product where an AI is involved, it still needs a clean source of data. But whatever the thing you're doing, the cleaner your data is, the better it can do the work.
Dax:So, you know, me and Liz are working on Radiant, which is, you know, for personal finance. We are the way we're thinking about this is we want to make it the first step is to make it stupid easy to, like, clean up your data and create rules that, like, just human created rules that, make that, like, the best source of financial data that's ever existed for you. Mhmm. We can layer in some, like, general AI stuff to help you clean that up. That's new.
Dax:Hasn't been able to haven't been able to do that before. It works really well already. And what we end up with is, like, a really interesting clean set of events, financial events for you. But the way we've built things, we're not limiting it to financial events. We'd love to get all sorts of stuff in there.
Dax:So just go take something completely random. Let's say your, like, browsing history. I don't know if we would actually do this, but, like, every website you visited, that's also now ingested in there. And you can see as AI gets better, we now have this amazing set of data that is like your personal data that this can now tap into, and we can build probably something really cool out of it. Like Yeah.
Dax:Ken Willett was coming with, like, really funny examples on Twitter, but it's the exact type of thing that we could enable. Like, you know, it's, like, 5 PM, and then your AI messages you being, like, hey. It's dinner time. You haven't ordered Indian food in a couple months. Maybe you should have Indian food to come and order that for you.
Dax:Today, AI could do that. It can totally do that. The problem is there's no pipeline to, like, it have that information about you to do that. So we're making a way to, like, connect all your financial stuff, but also your Amazon account, your Uber Eatsum account, connect everything. Right now, under the pretense of, like, just getting a good financial picture of your life, but you can see how if this data set becomes very clean, it just feeds into whatever end interface, AI interface is going to come out at some point.
Dax:And so that means it's the same thing. Right? Just having that clean dataset and the way people access it might be different in a couple years.
Adam:Man, yeah. You're you're just reminding me that, like, everything I was thinking about up to this point in the conversation was, like, the the the what we've come to know is the traditional way to interact with AI where, like, it's generating a bunch of text, but there's all these other ways AI is super useful, like summarizing and cleaning up data. And that's whole other vectors that companies will get better, new products will emerge. Yeah. It's just I think we're in for, like, a pretty great few years if you're into technology.
Dax:Yeah.
Adam:This this episode's turning into the AI episode, and I feel like it'd be really cool if at some point we're, like, this whole episode was generated with AI. This isn't actually Adam and Dax.
Dax:Yeah. Yeah. No. It's, it's it's exciting. Yeah.
Dax:Some of that stuff, like you said, like, outside the generative stuff, because we've been playing with it just to clean up data, and it's it, like, it solves things that I've seen Yeah. Over and over and over in my life, and finally it does it correctly. So a classic thing with financial data is sometimes it'll be the merchant name and for some reason the location. It'll be like, Starbucks, Miami, Florida. But, like, there won't be spacing between these things.
Dax:It'll just be Starbucks, Miami, Florida. So it's impossible to really write a parser that is, like, give me the counterparty for this. It should just be Starbucks. Yeah. Because as a human, you can you have a sense of, oh, this is the name and this is probably a location even if you don't immediately recognize location because it has a location sounding type name, right?
Adam:Mhmm.
Dax:And you can very easily split those apart. Historically very hard to do, but AI does this like near perfectly. There's another thing I've been kind of trying to figure out because I think this is going to be something that comes up a lot. There's like a set of things we can ask AI and it's gonna give you back a response. And oftentimes, we want a deterministic response, right?
Dax:So again, Starbucks, Miami, Florida. What is a counterparty? That kind of only ever needs to be asked once ever and then it can be cashed for, like, all time. Right? Because we don't wanna every time a transaction comes in, we don't wanna, like, ask you.
Dax:We're gonna build some kind of caching layer. So it's kind of like yeah. Like is OpenAI like caching every single question that people ask and then Oh. You know what I'm saying? Like, eventually, we'll have asked all the questions.
Dax:I mean, not literally, but we'll have asked, like,
Adam:it's like a I understand. Algorithm
Dax:thing. Right? Like
Adam:Uh-huh.
Dax:Most questions have been asked before already. And if you don't want a creative response, like, you know, you can set the temperature to to 0 or whatever. Yeah. Yeah. If if you're asking a question with a temperature 0, if someone else like, if someone else asks basically the same question, like, it can give you the exact same response.
Dax:So that's kind of interesting too.
Adam:Yeah. I think but you hit the one of the key sticky points, which is that there are other parameters, like and, like, I have custom instructions in my g JetGPT. So, like, how does that impact I was presumably, you can't just cache. Unless there's something some intermediate thing that they can cache that Yeah. Then runs through.
Adam:I don't know.
Dax:It depends on the use case because I don't think chat gpt could, but this case where I'm I'm, like, filtering and cleaning up data. Yeah, I'm basically 100%. Yeah. And then but then I also wonder, can it introspect and create deterministic rules for me? So when I say, give me the counterparty, can I tell it to write code that gets me the counterparty for this?
Dax:And it writes in, like, crazy ass, like, thing that doesn't make any sense, but it does work, and I can just run that code every time instead of directly querying the OM. It's really weird. It's such a weird, like, dynamic.
Adam:Yeah. It just opens up so many different things that you think, like, when we started our careers, there were problems that you just wouldn't even approach. Like, there there were things that you just couldn't even think about solving. And it's almost like it's like things you 5 years ago, things you would have reached for Mechanical Turk.
Dax:Right. Exactly.
Adam:It's like, oh, this is a human, but it's just it's still all within a system. But it has, like, human reasoning skills. It's pretty wild.
Dax:Yeah.
Adam:I'm very excited. I hope to sleep through the nights and and not have another night like that. But it it was a turning point for me. Somehow, that night of sleep, I went from, like, anxiety over AI and what does it mean for our profession to, like, oh, this is gonna be awesome. Yeah.
Adam:And, yeah, maybe the key is, like, being entrepreneurial. If you're not entrepreneurial in any way, shape, or form, maybe it's terrifying. But if you got any bit of that in your if you think you could, I think it's a huge opportunity.
Dax:The one other thing that you touched on that may last last point I'll make is, because I think the key thing in John Carmack's post was he talked about how he doesn't like managing humans, and he probably will have a better time managing AI. And there is a really funny thing where I've already experienced this, and it's been kinda crazy to realize. I don't know if we talked about this already. When I ask chatJPT for stuff, sometimes I talk to it for, like, some kind of complicated things. I'm, like, trying to get it to design me something or, the feedback you can give it is so much more precise and direct because you do not have to worry about a long term relationship damaging a long term relationship with this thing.
Dax:Yeah. So if it does anything bad, I assume, like, that sucks. Like, here's exactly why it sucks. Mhmm. Whereas the person, you, like, have to, like, give them a few tries and then, like, kinda, like, layer in feedback slowly.
Dax:That is an insane productivity boost when you can just be, like, this sucks, you're being too corny, like, need to be more like this. And you just can't interact with a person that way. So it's quite different managing
Adam:You can't. There's so much more that comes into play when you hire an a human being and you are or would you just put in risk in charge of them? Like, when they're your direct report, there's 1 on ones every other week. There's I mean, there's just all these things that an AI developer, you you don't have to deal with. And we just we sound like the evil empire right now.
Adam:I'm sure to some people that are listening, they're like, oh, the this is these are the jerks. They're gonna hire some AI developer instead of me. Yeah. I don't I I think it's the it's the what's the abundance mindset? It's like, this is gonna make, is that just a trope that it's yeah.
Adam:It's gonna take jobs. It's gonna make more jobs. I I really feel like there's just gonna be more opportunity because of all this. But I can't say that, like, in individual circumstances, there won't be people that don't have the opportunity they had. I don't know.
Dax:Yeah. And I and I I think that's the case. And at this point, my mindset is I can really just worry about my own path and, like, how it affects my own path. I don't know how it's gonna affect everyone else's. Yeah.
Dax:But I'm trying to Remember
Adam:himself. Sorry.
Dax:Yeah. I mean, it's just like people, I mean, a part of me, if I'm being honest, it's some of these differences are not what's the word? There are there are they're not, like, clean differences. Like, I've had differences with people in my life where I felt like their view of the world was getting in the way of mine, and they made it harder for, like, me to do the version of my life that I wanted. So in some ways, yeah, I will say, I will admit that this type of thing is way more in my favor than this other thing that I've kind of been fighting against.
Dax:So in some ways I, like, don't really feel that bad. So, yeah, that that is like a little bit there as weird as it is to admit. Like a like a Schadenfreude, I guess Yeah. Is the word.
Adam:If if anybody if anybody's listening to this and they're feeling anxiety now because they feel like they're gonna lose their job, everything moves slow. So just remember that. Like, things move slow. It's not an overnight thing. Now getting laid off at a FAANG company, that's overnight.
Adam:That happens all the time. It's been happening en masse, and it has nothing to do with AI, I don't think. So, yeah, could you get laid off at your cushy job or your not so cushy job? Yes. That is a fact of life.
Adam:But the AI thing, it's like long term trends and shifts in the way we work.
Dax:Yeah. This is not, like,
Adam:you need to be looking over your shoulder.
Dax:Yeah. There's such like a it's it's weird because I like I've been made so aware of how on the edge of things that I focus on. Like, just because of and I think for a lot of us that are friends, like, we're kind of all in the same place where we're just, like, always right on the edge, so we get hyper obsessed with this stuff that takes so long to make it to the vast majority of the world. Like, there's companies that are just moving to AWS and then, like, just figuring out the cloud. Yeah.
Dax:It's gonna take them a while to figure out all this other stuff. So, yeah, things do move really slow. Go ahead Matt what you said about hiring, where, like, the jerks are gonna replace people and hire I already made a tweet about this. I don't know if you remember, I was, like, I can't wait for GPT 5 so I never have to hire one of you idiots again. Oh, that's good.
Dax:I like people. I'm not one of those people that's like, I hate people, and I just wanna, like, be by myself and talk to AI. I love people and I, like, wanna continue to work with people. I just don't think it needs to be a lot of lot of people. And I've been this way pre AI.
Dax:Like, this has, like, been my dream for a very, very long time, which is, like, me and, like, a small group of people that I love, and we're just doing we're having massive impact. Like, I've always wanted that, and it's just, you know, step more in that direction.
Adam:Yeah. It just further enables it. What does e slash a c c mean?
Dax:Oh, man.
Adam:I I don't know what that is. It's just But I I keep I thought I'm sure
Dax:a lot.
Adam:What what does it mean?
Dax:I think Ken Wheeler's roasting of them was probably the most accurate.
Adam:Yeah. I think I did see one of his tweets. Is it is it like it's AI related? Like, it's people who are in the AI?
Dax:So there there's this idea, which is which I think I roughly agree with, which is we wanna accelerate AI as fast as possible, and that's what's best for humanity. Like, we shouldn't be, like, wasting time asking all these questions. Our primary goal should be to, like, boost this because any, theoretically, it fixes so much, and it helps Mhmm. The world so much, and you believe in that input. And, like, they've codified that into, like, a thing, e, a, c, c.
Dax:I don't know what something accelerate
Adam:stand for something?
Dax:Like I think the effective acceleration is, like, something like that.
Adam:Okay.
Dax:This, like, comes from, like, the there's been a bunch of things in the in this shape and so it's kind of in that category. But so, you know, some of that makes sense, you know, the way I described it. But when people put this kind of shit in their Twitter bio, it's just like a weird identity thing. They're like, I am someone that, like, likes AI. But what if you, like, dig into that, it's, I'm someone that's afraid of AI.
Dax:But if I put this in my identity, then it feels like I'm gonna be one of the people that benefit. You know?
Adam:Don't spare me. The AI overlords.
Dax:Yeah. Exactly. It's like, I'm gonna be one of the people that benefit. And you're all gonna get left behind, and I'm gonna be and it's like, someone that just puts that in their profile is not doesn't fit the persona of someone I would actually like Yeah. Benefit that from that.
Dax:It's it's I think, like, you see this with, like, crypto too, like, a bunch of crypto people put stuff in their profile that, like, signals that something, so yeah, it's purely an identity thing at this point and whenever I see them, it's like, yeah. Okay.
Adam:So we'll know we'll know how many of our listeners are, into this stuff or crypto based on how many fewer viewers we have after this episode. Just occurred to me, like, we have enough people to listen that I'm sure we offend someone every single episode.
Dax:Someone has EACC in their profile, and I'm hoping after this episode, they, they reconsider it. Because I am judging
Adam:Yeah. That's what I'm
Dax:very directly.
Adam:It's funny. There were other tech things that happened, like news, if you will. The White House says you should write Rust. This is why I didn't read anything beyond. I didn't read it either.
Dax:Did they did they literally say Rust, or did they say that they don't we should encourage memory safe languages? Is that what
Adam:they say? Memory safe. Yeah. I don't the literally, what I know about this is from, like, a couple of YouTube thumbnails. That's literally it.
Adam:That's how I learn about tech these days.
Dax:To be honest, I I think there is literally nothing more than what you've learned from YouTube thumbnails. It makes sense. Like, you know, cybersecurity is a huge thing, and a lot of the exploits are for memory safety issues. So, of course, this makes a little sense. And of course, the best thing to take away from it are a bunch are a bunch of jokes, which I think is what happened.
Dax:Right?
Adam:Did Russ ever make a comeback after, like, all the foundation stuff? Like, did it come back in favor, or are people still upset?
Dax:A lot of stuff is being written in Rust. That's so I don't know if that stuff really made a dent in, like, the actual usage of it.
Adam:It was just the Twitter perception of it.
Dax:Yeah. I think I think some energy was lost.
Adam:It hit all 50 of us in the Twitch, dev community. That was it.
Dax:Yeah. Yeah. And then the 50 of those people that never really ship anything anyway stopped writing Rust. So
Adam:Yeah. Okay. Cool. Cool. Cool.
Dax:Yeah. What else?
Adam:I feel like there was other stuff on Twitter. It's been it's been semi active in the tech world.
Dax:Has it?
Adam:Maybe not. Maybe I made that up. What have you been tweeting? That's always fun.
Dax:I feel like we actually have segments, and one of the segments is, reviewing my tweets. Yeah.
Adam:Yeah. We do. It's it's looking at your Twitter. Yeah. You I just saw your tweet from an hour ago.
Adam:You've all had CO 2 monitors for over a month now, and I haven't seen it
Dax:improve it yet. Oh,
Adam:Oh, I love it. I open my window a lot more now. I do have one sitting right here.
Dax:You say you open your window a lot more now? A lot more.
Adam:It's 1193 right now. That's not good. It when I turn on my studio lights for some reason, it's probably because I talk more when the lights are on, But the c o two levels, they climb. It's just too cold right now. I can't open the window.
Dax:Yeah. I mean, I like I said, I live in, like, a very container. But, again, my AC is always on and I think that basically does it for me. But I don't know, I don't have a CO2 monitor so I could be dying.
Adam:By the way,
Dax:this is gonna be my last episode from this location. So this
Adam:is Sorry. What? Oh, you're moving.
Dax:Yeah. Next door. To an identical like, this room is identical.
Adam:It's gonna be identical, like, the houses are similar.
Dax:Yeah. So I'm gonna set everything back up, and we can check, like, a frame to frame thing and see how much it actually changes.
Adam:That's awesome. Yeah. I love it. So there's no way anyone will notice if they don't hear us say
Dax:that you just moved. Did you see Jay posted a picture of the of my Yeah.
Adam:Was that the back of your house? Yeah. Did he come over? No.
Dax:No. No. I know. I I I sent him that picture, and then he posted it on Twitter. That's what we do.
Dax:Anytime someone sends us a picture in our team chat, someone else posted on Twitter and
Adam:made fun of them. Funny. I still see pictures of you that pop up all the time, like the one I generated in, DALL E, and then there's the one you posted of you, like Looking insane? Yeah. You're like, you look insane.
Adam:Yeah. Oh, man.
Dax:That's gonna be the face I make when I replace everyone with the guy. That's the exact energy of that photo.
Adam:That's what you should be, yeah, anticipating if your job's on the line. Can you guys
Dax:just switch my Twitter profile picture to that? Like, what kind of energy would I give?
Adam:It'd be weird if you didn't have this avatar. You've had this avatar as long as I've known you.
Dax:No. I'm never gonna change it. I do have a I I did okay. So when I started being more public, I made a rule to myself, which is I'm gonna pick an avatar, and I'm never gonna change it. Mhmm.
Dax:Because when you change it, like, you never really are. People yeah.
Adam:I've changed my life.
Dax:Committed to this forever. And to be honest, I think my avatar, even if I'm, like, 30 years older, I think it still works.
Adam:It'll still work. How did you make that, by the way?
Dax:It was actually a gift. Someone gave me a watercolor. Someone knew someone that did watercolors and for I forgot what it was. It might have been for Valentine's Day or something. They gave us, me and Liz, a watercolor of, like, me and her together, and that's just, like, my head nods.
Adam:That's cool. So Liz could have one that's of the other part of the watercolor. Yeah.
Dax:And I think she should. But I think She should. I think what what it's different with women because, like, her hair is different now. So, like
Adam:Yeah.
Dax:That doesn't match, and she feels like it's a different person. Whereas me, you know, not to worry about that.
Adam:Yeah. I was gonna say, I could do a watercolor review. That's a little a little easier than
Dax:It's long lasting.
Adam:The average person. The the your banner photo, on Twitter, is that like a Zuiko kind of
Dax:That was that was, like, when, like, AI generated images first came out, and I was, like, what am I
Adam:AI generated. Yeah. I was, like looking right at your head. Like, he's pointing right at you. Yeah.
Adam:Unless you notice that.
Dax:I I do wanna change it, but I'm very uninspired. I don't know what to do with that.
Adam:I forget what mine is half the time.
Dax:Did you
Adam:realize you've tweeted 15,000 times?
Dax:And Is that true?
Adam:How how bad do you feel about yourself now?
Dax:I don't feel bad. About all
Adam:the tweets?
Dax:I feel good. Okay.
Adam:That makes me feel good.
Dax:So a lot of thoughts that I have documented that I can go back and read. You
Adam:do. You crank them out. That's why we can always talk about tweets of yours because you always have new tweets to talk about.
Dax:Yeah. The c o two one today was good.
Adam:Oh, what did you do with the Sentry? Well, what's his name?
Dax:I wasn't there. Okay. So that was the other way. Jay sent me that picture, and I posted on Twitter.
Adam:Another photo that ended up in your Slack and then reposted.
Dax:He was in David was in Toronto, so Jay met up with them.
Adam:Oh, nice.
Dax:Yeah. I see them. David, Armin, Ben.
Adam:What's what's the PlanetScale Insights feature?
Dax:Oh, man. This is awesome. Someone just asked about it, so I was gonna I can talk about it. Someone asked how long have I had my plans get recommendation. I've been using it for over a year now, and I've just always had a great time with it.
Dax:So I've been recommending it since then, I guess. Yeah. They they turned it on for my account last week, and it basically Flex? It was
Adam:a nice nice flex. Yeah. I
Dax:got it early. I got insights early to my to my MySQL database, coveted. I wish I feel bad. So I Yeah. It's actually a very straightforward feature.
Dax:I'm sure it's, like, complicated to do, like, a lot of details, but it's straightforward in the sense that it's obvious. They can see all of your queries happening. They can see how long they take. They can see how many rows it scans. And it basically gives you recommendations of being like, hey, you're querying by this field a lot.
Dax:Consider adding index to it because you're currently scanning a lot of rows. And this is like a frequent query you're doing. And it gives
Adam:you Yeah. It's Aaron sitting over your shoulder.
Dax:Exactly.
Adam:Hey, Dan.
Dax:People are joking. It's just him, like, typing out recommendations for everyone.
Adam:He's manually going through all the queries, and I found one.
Dax:But there's things that it caught. There's one thing that it caught that was, like, a really stupid mistake. Like, most of my tables are the same and they're queried in the same way. But There was one table that was slightly different. It was queried in a different way, so I forgot to put in like, I needed a secondary index on on a field, but that was, like, the way I queried it.
Dax:And I totally didn't notice. And it was it was, like, one of the most frequent queries that happened in my system. And it called it out. It was, like, here's a table. Here's what's going on.
Dax:Here's a recommended index that you can add. It didn't match my naming scheme, so I didn't you can just, like, click a button there and, like, update your schema right there.
Adam:Oh, wait.
Dax:Hang
Adam:on. So PlanetScale. How do how
Dax:do how
Adam:do you codify your schema?
Dax:That's why you would never use that feature because Okay. It doesn't make because I the source of truth
Adam:for that into your yeah.
Dax:Yeah. Well, my source of truth for that is a drizzle definition. So I went updated the drizzle definition and then, did my normal
Adam:You just kinda reference what they say, and you kinda know where you should at. Yeah.
Dax:Okay. And then, and then I did it through Drizzle. Even though the index I added wasn't exactly the same as what they recommended, they understand that it fixes that problem Yeah. And they automatically detect it, and they close the issue, which is amazing. And then
Adam:That's amazing.
Dax:Yeah. And I found a bunch of others where, like, I had extra indexes from stuff I was playing with that I forgot to remove. So it's great. It just, like this isn't even AI, but it gives you that same feeling where it's because it's just completely heuristic based, but it gives you that same feeling where you don't have the time to, like, babysit this stuff and it can stuff and easily fall through the cracks. Yeah.
Dax:But it just does it for you. Like, yeah, you can give people good observability tools to find these issues. But, like, just find it for them and then just tell them.
Adam:Yeah. And just suggest, here's how you
Dax:fix it. Yeah. I wish more tools did that. That's awesome. That's great.
Dax:And I'm just, like, why hasn't RDS shipped this? Like, they've been dominating SQL deployments for so long now. Yeah. And then they just started to expose, like, slow query stuff.
Adam:But I'm sure it's not fun. I'm not I've not used it, but, like, do what do you have to go to CloudWatch or something? Oh, no.
Dax:It's a it's huge pain.
Adam:It's like Yeah.
Dax:So even even turning on that feature is a whole set of steps. But, yeah, PlanetScale is great to continue to this is another one because in my tweet, I was like, this is another sticky feature. I called it out in that way because I'm also really into them as a business because I feel like they do things in a way that makes a lot of sense. Like, their when they first launched, their big thing was a whole merge request flow. And these are features where they're very different.
Dax:You probably haven't seen it before, but, like, as soon as you try, you're just, like, fuck. I never wanna give this up. And that's great as, like, a dev tool business to really, like, focus on, like, really small things like that that that once you experience it, you'll bend over backwards to, like, you know, keep having that in your workflow. So this is another one. I would say this is their second one where I'm, like, okay.
Dax:I like it would be hard for me to get this up now.
Adam:Yeah. I wanna I wanna talk about databases for a second because Yeah. Until, like, the data API came out for serverless v 2.
Dax:Mhmm.
Adam:I think we talked about it at one point. Yeah. How do you have you, like, embraced that? Are you still just too is PlanetScale's too too sticky and you can't leave? Is there anything that you've discovered with data API that, like, oh, it's not as good as I thought it was on v 2?
Adam:Or how do you feel about that?
Dax:We use it a little bit for our movies demo because, like, the vector database was backed by that. Yeah, it worked. There's no issues and I think the Drizzle adapter for it works fine. There's some, like, weird quirks where it can't do certain data types, which I don't fully understand, that we've just seen before, and and they're still present in this in this version. Also, they only do Postgres right now, I think.
Dax:If the MySQL version is still not
Adam:Oh, right. They haven't they haven't rolled it up for MySQL. Okay. So that makes sense why you would still you'd be in PlanetScale until that, at least.
Dax:One of the next things we have to do is make our console self hostable. So that would mean conditionally swapping out PlanetScale for RDS, but we're kinda waiting for MySQL support in the data API.
Adam:Yeah. That makes sense.
Dax:But yeah. Okay. Like, there's some things the PlanetScale model is is it's just a model and there's, like, trade offs and negatives with it. So Yeah. They have a very traditional deployment.
Dax:You have a MySQL instance accessing data on disk that's local to the instance. Mhmm. And to scale it, you, like, shard potentially if you, like, need to scale it that way, and you have these, like, live instances. Whereas Aurora and Neon do that thing where they separate everyone loves this phrase. We separate compute from storage.
Dax:Yeah. So, like, the storage layer is completely independent, and that can scale and replicate on its own outside of like the MySQL or Postgres replication protocol. And they can just spin up compute as needed to run your queries. And there's some benefits to this because, like, the whole auto scaling thing can work a lot better under that model. That's why, like, serverless v 2, you know, can scale up and down.
Dax:There's it's a little bit BS because when it scales up, its caches are cold. So you kind of have this cold start thing that you probably don't even realize you have because it's, like, hard to observe. But technically, the separation of storage and compute is better, for that type of thing. So if you're like if you're adding an index on a giant database, you need a ton of compute to compute it, You can like scale up just to add the index and then scale back down. So there's cool stuff like that that RDS can do.
Dax:Yeah. But it's just like it's a different model, but there's thought a lot about the plans to get a model that's simpler plus all these, like, nice features they threw up threw on top. Okay.
Adam:Can I talk about health insurance?
Dax:Yeah, please.
Adam:Or generally health care in America? I'm gonna try not to make this too ranty, because sometimes I get a little fired up over some of this stuff. But, it's been almost a year since I canceled my health insurance.
Dax:And I
Adam:don't know if I've ever talked about this on the podcast. I've definitely talked about it on the Internet. I just canceled it. I was just one day, I was like, this is stupid. Not even just the money, but just like the health care system in America is stupid, and I don't wanna be involved with it.
Adam:So it was kind of like going off the grid. It was like, I just wanna be as separated as I can from what I think is a very broken system. Like the way the billing works and the way health care providers have to charge so much more because the insurance companies will only pay, like, a certain percentage and all this big game and everybody just knows it. It's just stupid. So I stopped paying my health insurance.
Adam:In that time, I picked up jiu jitsu and went to the ER twice. Yeah. I I've not gone to the ER in my adult life. So in high school, I once, had to stay in the hospital overnight for an infection. But since then, in my adult life, I hadn't really had any serious medical situations, and I paid for health insurance for those 10 years as an adult and never really used it.
Adam:And then I gave it up and promptly, ended up in the hospital twice. There was something else that's happened health wise. I've had some knee stuff. Anyway, it's the most I've, like, been a participant in the health care system, and I don't have health insurance and can't believe how much money I've saved. Even with 2 emergency room visits, like, that's been, like, 4 months of premiums.
Adam:And I know, like, insurance is to, like, protect you from the big one and, like, that I know, people are listening and they're like, yeah. Well, you haven't had, like, a terminal illness or, like, some terrible thing that's gonna cost all of your net worth or whatever happen, and that's why you get insurance. There's insurance for that, though. Right? I know I I keep
Dax:That's what I have.
Adam:To look into this. Yeah. Like like a catastrophic insurance or something.
Dax:Mhmm.
Adam:Yeah. So I don't even have that. I should probably look into it. But I don't know. The health care system's stupid, and it just bothers me.
Adam:Why does it bother me so much that, like
Dax:It's really annoying. It's extremely annoying. Part of it
Adam:is just, like, we're such we're so we're in such bad health, and we spend so much on health care. And it just feels like that dichotomy, like, as a country, I'm speaking specifically to the US, like I mean, that's I guess that's the whole thing with insurance is you spread around the risk. And the risk of, like, dying of heart disease in America is super high, so I don't wanna pay that shared collective risk when I feel like there are measures I can take that can prevent or lower my risk to, like, avoidable diseases. Okay. Now I'm just gonna stop because I've not slept enough, and I don't know where I'm headed with this.
Dax:No. But okay. So on this topic, so, I do have health insurance. I do pay for it myself. Every year when we have to, you know, sign up for it, we go through the same process where we're like, okay, maybe things have changed, like, what's going on?
Dax:We do the research, we type in the thing, and we always pick the lowest premium option because we don't need Yeah. I don't need any of the stuff that they try to build around it. I just need the most basic. I will pay for everything myself. Yeah.
Dax:Give me some number if I go over, you take care of for something catastrophic. That's all I need, cash out insurance. And it's always so annoying because that even for that, the price of that has doubled since I started paying for it like years ago when I was like, you know, I was too old to be on my parents. Yep. It's doubled And they keep, like, trying to, like, kind of, like, make it, like, real insurance, but I don't really want that.
Dax:There's, like, no such there's no thing out there where it's, like, they just do the most bare bones stuff. Like, they don't try to Yeah. Give you a network. They don't try to do any of that. They give you literally 0, and they just try to minimize premiums.
Dax:Like, it doesn't that doesn't exist.
Adam:Mhmm.
Dax:And the problems with our health care system because, like, people are always like, oh, yeah. It should just be, you know, public, like, other places. Like, there should be a government option. The thing is, like, I hate when people, like, don't look at problems specifically. It's kind of look at generally.
Dax:They're just, like, oh, other companies are have, like, a government option, so it'll work here. But if you look at some of the details, we have, like, a mess up system in so many little ways that making a government option, I can easily see how that's just going to make things even worse. If you look at a per capita spending, I think we literally outspend the next country by 2 x.
Adam:Yeah. On health care costs?
Dax:So we're already spending a crazy amount per person, and we're not we don't necessarily have, like, double the outcomes. It's also really complicated because a lot of this stuff gets invented here. And, like, you know, the whole world gets a benefit. So that that's another dynamic, but there's also this dynamic where hospitals have consolidated like crazy. That's not a very competitive, market and the prices have gone up because of that.
Dax:There's laws that are out there that make it so if you wanna start a hospital, you have to prove that you are not competitive with an existing one. It's, like, complete completely backwards.
Adam:You've mentioned that to me before.
Dax:Yeah. So there's, like, all there's, like, this whole clusterfuck of of problems. And, yeah, at this point, like, I think what has gotten better is I think providers have recognized that people just want to opt out of insurance and there are like great doctors that have like somewhat reasonable prices that will treat you without insurance and they have an upfront, here's what it costs, you pay me, you know? But, yeah, you still I still feel like I need that disaster insurance just in case. Yeah.
Dax:And the cost of that is way higher than it needs to be. It's really annoying. And for you, like, you have a family of 4, so it's probably, like, 1,000.
Adam:Oh, yeah. It'll be it'll be a lot. Yeah. Yeah. I don't know.
Adam:It it frustrates me every time I think about it or especially if I talk about it.
Dax:Yeah. So so people that try to say, I stopped my forum in half, and it was, like, $10,000. Without insurance, it would have been a 100,000. So here's, like, another stupid thing. Like, if you go to the hospital and you ask, what's the cost with insurance?
Dax:They say 10 k. What's the cost without insurance? Did they say a 100 k? The 10 k, your insurance is not paying the other 90 k. It's just that's the insurance rate that they've negotiated.
Dax:Yeah. When they say you have to pay a 100 k, you don't. You just tell them, you get treated, and you're like, I'm not gonna pay a 100 k. They will make you pay 10 k. So it's just it's just annoying for a stupid reason.
Dax:And we had this with, Eliz got her gallbladder out, and they get they caught us some, like, insane number, like, 100 and something $1,000. And we just said we're just like, no. We're not gonna pay it. You can do funny things too. You can, you can send if there's if they keep sending you a bill, you can send them $1 every time they send you a bill, and that, like, resets whatever thing that they have.
Dax:It's just so dumb. Eventually give up. It's so stupid, but you know what? It's a stupid system, and you gotta act stupid in it.
Adam:I guess that's what it is. It's a stupid system. You just gotta act stupid to not go crazy because oh, my word.
Dax:Yeah. So you just gotta play a stupid game with the hospitals where these procedures are not expensive. I mean, they're expensive, but not, like, outrageously expensive.
Adam:Yeah. I mean, don't they have to charge like, they have to charge insurance companies this huge rate because the insurance companies only pay, like, a fraction of what they bill. So it's just this game where they keep billing
Dax:higher numbers. It's it's really yeah. It's bad for everyone. The hospitals, like, any service provider that takes insurance, you literally have no idea what you're gonna get paid when you treat a patient. It's, like, so confusing.
Dax:Yeah. Liz's uncle, he owns a giant health care business in Miami, like, billboards everywhere. Everyone knows what it is. It's crazy how big of an operation this is, and he treats, like it's like a network of doctors and pharmacies and, like, everything you need for someone elderly to, like, take care of all of their needs. Yeah.
Dax:It's even gotten to a point where they have, like, recreational stuff and just it's like a whole community. But it's massive. They have, like, so many so many people enrolled in this. This giant complicated business, like, they put all this together over the last, like, 30 years, so much work. They make, like, $70 a patient per year.
Dax:It's something that like, in net, that's how that's what they get. Yeah. And they're facing an issue right now because some middleman somewhere increased something by, like, roughly that amount.
Adam:Oh, my word.
Dax:But now they're making, like, near 0. Yeah. So it's, like, this big complicated difficult business that he spun up just for, like, this this small I mean, he makes a lot of money. I mean, they have made a lot of money because they're so big, but Yeah. It's just crazy how complicated it is compared to anything that I've ever done.
Adam:Yeah. We're spoiled by the software margins.
Dax:Yeah.
Adam:Or at least what used to be software margins. I don't know.
Dax:Maybe those are gone. They're going away, for sure.
Adam:Yeah. Are interest rates climbing again? Are they
Dax:Are they?
Adam:Or, I mean, lowering again. Are are they dropping interest rates? Did I hear that?
Dax:They definitely committed to pausing hikes for
Adam:a bit. Pausing hikes, which is as good as lowering
Dax:Yeah. Some people think in
Adam:a lot of people's eyes.
Dax:Some people think they might be they might be forced to raise again, depending on how pausing goes. But you can see the market, they're like the market is basically priced in rate lowering. I think so because my portfolio is not even higher, which is amazing.
Adam:Interesting. I
Dax:feel great. My emotions are good again.
Adam:It's all tied to the markets.
Dax:Yeah. I feel great. And I I need the rates to go down so I can so I can find someone that's by my New York apartments. I can buy the house that I'm moving into. And
Adam:Yeah. You're stuck with that situation. My parents are trying to sell their house. And they've had it for a while.
Dax:I'm gonna I haven't even tried selling it. I'm gonna I'm gonna list it. I'm gonna list my apartment this summer, and I'm gonna see if someone buys it. I just can't I, like, can't fathom who would buy this, but I can't fathom who would buy a lot of things and then they always sell.
Adam:So Yeah. Maybe a fan of the Dax's house.
Dax:Yeah. I want them to pay for an apartment in New York City in a great location, really quiet location. Not the location that you probably would think to go to, but it's a great location. Let me know. I have an apartment for you.
Adam:Okay. While we're talking to our our listeners, we are at an hour here, and I need to get off here before I just need to stop talking. My c o two monitor is getting out of hand. It's 1360. I'm gonna die over here.
Adam:I don't have health insurance. But if you're listening to this podcast still after an hour and whatever it's been, maybe hop into Apple Podcasts. Leave us a rating. We don't we don't get a lot of ratings. I think you're supposed to ask for them, and maybe we'll get more if I ask for them.
Adam:And I don't even honestly know what happens if you give them to us,
Dax:but it makes you feel
Adam:as I look
Dax:at them. Clarify.
Adam:A good oh, good rating. Yeah. If you if you don't like the podcast, why are you still listening? And don't leave us a rating. That's mean.
Adam:But, you know, if you wanna leave a good one, it it warms my heart. I check it every once in a while. We've got we've got some, but not not a lot. If you wanna leave a review, even better. Tell us what you like.
Adam:If you don't like stuff, maybe DM it to us. Don't put it on the public record. Dax, do you have anything to add to this meta, asking of the listeners?
Dax:Did should put it at the beginning of our podcast.
Adam:Oh, probably. So that everyone who hears the podcast because there's probably, like, 10 people that listen to the end. Yeah. But maybe one of those 10 people.
Dax:Yeah. That's
Adam:true. We
Dax:only have a one per episode will add up.
Adam:Yeah. Hey. We've done, like, 81 episodes. We're coming up on a 100. We gotta think of something special.
Adam:I wonder if we could record
Dax:saying we've been coming up on a 100 ever since we got over 50.
Adam:I wonder if we could if it would time out, I could do the math. But could we do could we record episode 100 in, Miami together when I come for Miami? Probably way off. No. That'll be like episode 86.
Dax:Wait. How many weeks is it? There's only 4 more weeks. Right?
Adam:It's not a lot of weeks.
Dax:6
Adam:weeks. Yeah. Something like that. Okay. Well, maybe we'll record episode 100, but not release it until we record the other 15 episodes.
Adam:And then Okay. That'll be episode a 100.
Dax:I don't know. We'll see. We can do it in this office or in, I guess, in my new house.
Adam:In the one that looks just like that.
Dax:Identical office. Yeah. Did you see what I said about the lights in that house? No. So we go we got the keys yesterday, and they're still fixing some stuff up.
Dax:So we're not, like, living there yet. But, we get the keys and I turn on the faucet in the sink because I'm like, okay, what's the water pressure like here? It was good, which I'm happy with, But then the lights start flickering and I'm like, no, no, no.
Adam:When you turn on the water When I
Dax:turn off the faucet, the lights stop flickering. And I did it like 3 or 4 times to really make sure
Adam:Oh, no.
Dax:And it's like it's like strobing. It's like Yeah. I'm like, what the hell is this? And, yeah, so it has a tankless water heater, which is great, But Yeah.
Adam:We have 1. Love
Dax:it. Yeah. I think it's drawing too much power. Because I think it's a new one. They added that in recently.
Dax:Oh,
Adam:so it's not like water touching something.
Dax:Short That's what I was worried about. So how are we gonna track that down? How are we gonna track down, like, where this is happening? Well, you'll
Adam:know when you see the the fire. That that's where it was.
Dax:Remember the water?
Adam:I guess the water yeah.
Dax:That would have caught fire. I don't know.
Adam:Anything electrical just freaks me out.
Dax:But then I found the water heater, and I turned it down, and then it stopped happening. So this is, like, a thing I gotta go figure out. I probably will have to call an electrician and figure it out. I'm assuming they put it on the same circuit as the rest of the house instead of, like, having a separate circuit for it. There were some weird fixes that people online had.
Dax:They, like there's some kind of, like, ring that you can buy that's like this it's made out of some, like, metal, and if you, like, loop certain wires through in the right order, it'll, like, stabilize whatever is going on. What? And it's just, like, completely this passive thing that I don't know. I gotta look into it.
Adam:That sounds dangerous. Did you know that, like, electricians get shocked all the time? I was naive and thought, like, if you're if you see an electrician, that means they've they're really good, they've never died, and they've never gotten shocked. But, like, they get shocked all the time. It doesn't kill you.
Adam:I would have thought, like, you touch,
Dax:I don't know, an AC
Adam:wire, yeah, in the US, and it would just fry your heart. You'd fall over dead. But they get shocked all the time.
Dax:You know that some people get a magnet embedded in their finger?
Adam:What?
Dax:And for people that work with electricity a lot, they get they, like, they, like, augment themselves by putting a magnet in their finger and they start to be able to feel currents and your brain, like, integrates that to, like, a new sense that you have. What? And helps them, like, quickly realize that wires lag.
Adam:Like a void. Yeah. That's so cool. It's like Spidey senses or something.
Dax:Exactly. But just for, like, this very one specific scenario.
Adam:I'm trying to think of how I could get some kind of an implant that would help me with my work. Can't think of anything. I mean, Neuralink, I guess. I could just think and stuff would happen on my computer. Will you get Neuralink if you if it, like, becomes a thing where you can just go in and get, like, an outpatient procedure and have this thing implanted in your brain, would you do it?
Dax:Percent. Yeah.
Adam:Well, really, it it okay.
Dax:It depends what I get out of it. Like, if I'm just, like, controlling shit in my mind, like, that's not that interesting. What else
Adam:would you get out of it?
Dax:If it's, like, backing up every, like Oh. Memory I have to somewhere, so then if I die, it, like, restores it into a clone of me? Yeah. I'll do it.
Adam:Oh, interesting. Okay.
Dax:You you weren't expecting that specific of an answer, were you?
Adam:No. I wasn't. I thought it was just, like, move your mouse with your brain. I don't know.
Dax:That's funny. That was enough for you. That's like the end state of this technology.
Adam:Yeah. That's what I thought it was. Okay. Anyway, alright. I'm gonna stop talking.
Adam:So I don't die of carbon dioxide.
Dax:Oh my god.
Adam:Is that a thing?
Dax:There's carbon monoxide. It's a new we just we keep inventing new things to be worried about.
Adam:I bet I'm gonna bring my when I come to Miami, I'm bringing my monitor.
Dax:Oh, can you please bring it? I do wanna know.
Adam:I'm going to.
Dax:I wanna know.
Adam:Your office is probably like 25100. You're probably just like, that's why your tweets are so dumb. You're just, like, half a brain so less. Your tweets aren't dumb. I'm just I can't even be mean.
Dax:My tweets are working, so I I I I should change it. I should I should hurt.
Adam:I should Just keep doing it.
Dax:Accept this damage to my brain.
Adam:I remember when you had, like, 400 followers on Twitter. It's so funny. Like, I remember thinking, this guy's way too smart to not have people listening to him. What's going on?
Dax:Tweeted that once, and I that made me feel good.
Adam:I did. I tweet I could I could pull up the receipts. Yeah. And now look at you. Everybody's hanging on every word, what Dax has to say.
Adam:Anyway, I also remember when, like, Theo had a 1,000 followers, and I'm trying to think, who else? Trash?
Dax:He's only a 120 something now. Yeah. Or 20 something 1,000. Yeah.
Adam:Remember when Trash just had a couple thousand?
Dax:Trash grew, like, impossibly fast.
Adam:He did. Crazy fast.
Dax:And, like, he just never grew again. He just hit. Yeah.
Adam:Hit a ceiling with memes?
Dax:Yeah. Oh, now The trailer never posts.
Adam:Is trash coming to Miami? It's coming around Miami? No.
Dax:No. He keeps every time I ask him, he says the same thing to me. He says, Miami is a wasteland or he says, Miami is not good for me. I was like, okay.
Adam:Which is it, trash? I mean,
Dax:I guess both could be true.
Adam:His name
Dax:was really trash. So, you know, he belonged in the wasteland.
Adam:I will never forget, TwitchCon in Vegas. Like, the moment he touched down, I think. He had, like, cigarette in the mouth, and he was gambling in the airport. I mean, just, like, immediately turned on Vegas mode. Oh,
Dax:man. Yeah. I
Adam:thought of I thought of something we could do at the end of every episode. What was it? I don't remember.
Dax:A cheer?
Adam:No. A cheer, but it was something. I don't know. Maybe I'll think of something later. Maybe episode 100 will unveil, an outro.
Adam:Does it feel weird to not have I don't know if everyone see it.
