[00:00:00] Andrew: Hello
[00:00:00] Nathan: Hey, how's it going?
[00:00:02] Andrew: We did it
[00:00:03] Nathan: Yes. You did. [00:00:05]
[00:00:05] Nathan: Right before this you were trying to figure out how to download Streamable.
[00:00:08] Nathan: I'm just
[00:00:08] Andrew: We're almost always late, [00:00:10] right? But,
[00:00:11] Nathan: Oh, we are, 'cause we always chat beforehand, and that goes
[00:00:14] Andrew: [00:00:15] Yeah. We had some stream issues, so we're no longer streaming to Twitch 'cause they [00:00:20] suck
[00:00:20] Nathan: Maybe we'll be a little more serious. We felt like the format doesn't really
[00:00:23] Andrew: Yeah. No, [00:00:25] no.
[00:00:25] Andrew: I don't know. I don't really have an opinion on Twitch, but we're trying to do Twitch streams. [00:00:30] Just I don't think we're the right format. It was easy enough to try. We [00:00:35] had some viewers, so if you're watching, hello on YouTube. We will [00:00:40] stream live on
[00:00:40] Nathan: Which they would never, they would never know.
[00:00:42] Andrew: I know, right?
[00:00:43] Andrew: It's just gonna be blank forever.[00:00:45]
[00:00:45] Andrew: Um, and then I always forget, but we have a Discord now, so [00:00:50] Discord's in the below. Yeah, it's, it's weird. Usually we have a topic usually we [00:00:55] talk about. Today I don't know, there's a lot going on. One thing I did wanna, to [00:01:00] mention was the DigiCity saga continues.
[00:01:04] Andrew: So [00:01:05] the, the Ship a Ton competition we talked about on the channel before I've been doing that, [00:01:10] and it's been submitted now for a few weeks, I think. [00:01:15] So it was the end of September was the deadline.
[00:01:17] Nathan: right?
[00:01:18] Andrew: I don't remember. I
[00:01:19] Nathan: been a week. [00:01:20] Yeah.
[00:01:20] Andrew: Store, and it took 10 years.
[00:01:22] Nathan: Yeah. You submitted it before the deadline, which is a smart [00:01:25] thing to do. Yeah
[00:01:25] Andrew: Yeah.
[00:01:26] Andrew: Some, not 50% of the [00:01:30] podcasts submitted before the deadline
[00:01:32] Nathan: My app actually just got approved this morning,
[00:01:34] Nathan: [00:01:35] but
[00:01:36] Andrew: I hate it
[00:01:37] Nathan: But I'm pulling out all the junk I needed to put in there for [00:01:40] the, the, the competition.
[00:01:41] Nathan: It's already halfway done, so I might as well just put it up on the store.
[00:01:44] Andrew: Yeah, might as [00:01:45] well. We could... We'll do a tour when it's up, but yeah. So I, I did the DigiCity app, which is a city builder [00:01:50] game for iOS and Mac, iPad OS, [00:01:55] and, so at the end of September it flipped over, and then they do judging, and so the judging is done I believe [00:02:00] the end of this month, and there's a timer on the site or whatever, and I think there's 12 days [00:02:05] and some hours left.
[00:02:06] Andrew: And so part of the, the competition is to get it [00:02:10] submitted, but then they're also looking at, like, how do you grow the app? H- what kind of user adoption do you get? [00:02:15] I did hard mode on this because I started the competition two weeks [00:02:20] late, 'cause I didn't know about it, and I found out, and then, normally you're, like, hyping [00:02:25] it and you try to get a, like a, a good launch audience so you get a big spike at the beginning.[00:02:30]
[00:02:30] Andrew: I didn't really do that. Talked about it a little bit on the podcast, but I, I wouldn't say that I went [00:02:35] to hype or anything. So, that's out there. I put it in the [00:02:40] description here, so you can go to the website, but it'd be really awesome, I'll pull it up here. If you go [00:02:45] to the Ship It On competition page, there's a little Like button here.
[00:02:48] Andrew: I've already liked [00:02:50] it, h- smash the Like button. There's no...
[00:02:54] Andrew: I know, [00:02:55] right? There's no login, there's no account or anything, so you can visit it. Take a [00:03:00] look. So it's, it's based on views and then likes, and then there's the little App Store button up [00:03:05] here. Please click that, download the app from the Apple App Store.
[00:03:09] Andrew: Please leave a [00:03:10] review. You can also buy the app. That would be awesome. If you don't wanna buy it, [00:03:15] that's cool. So the game is free to play. There's no time mechanic gimmick. There's [00:03:20] no in-game currency. There's no loot boxes. There's no, constant value [00:03:25] extraction, so you can play the whole game for free in the sandbox, and then, if you wanna save the game, [00:03:30] there's a s- a small pay to unlock one-time purchase.
[00:03:33] Andrew: There's some [00:03:35] scenario content in there so you can play with scenarios to give a little bit of a different kind of flavor of the [00:03:40] game. We're, in this version that's out t- or we, we sent it to Apple today. Who knows how [00:03:45] long it's gonna take? It's been taking minutes or hours or weeks, depending [00:03:50] on the sand.
[00:03:51] Andrew: We're, we're putting out challenges, kinda like scenarios [00:03:55] where scenarios bring you through different eras of time in history and give you kind of historical challenges. [00:04:00] The challenge mode can be done any way you want it, and we kinda set you up with a, a [00:04:05] challenge, and then you have to, to fix it, right?
[00:04:07] Andrew: So it's like you... if it's, the city's too noisy, you have to figure out how to [00:04:10] de-noise it. If you're bankrupt those kinda things. So I, I, oh, I should know this. I think [00:04:15] there's 11 going up,
[00:04:16] Nathan: Well, I've
[00:04:17] Andrew: initially, 11 types of challenges
[00:04:19] Nathan: Oh, [00:04:20] okay. Okay. Cool
[00:04:21] Andrew: yeah, I could be wrong, so don't quote me on that. There, it was, it [00:04:25] was four and then I think I did more, so it's, it's seven, maybe
[00:04:27] Andrew: there's 11.
[00:04:29] Andrew: Yeah, [00:04:30] right. Uh, a- a- and those are things we'll keep, we'll keep adding. Scenarios bring you through moments of [00:04:35] time, so probably won't be doing too much more with those. But the challenges [00:04:40] are k- are the kind of the fun differentiator and the goal with those is to eventually make it [00:04:45] so that you can kinda pick and choose what you wanna focus on for the, the [00:04:50] city building.
[00:04:50] Andrew: And so if you don't wanna just deal with all the different pressures and [00:04:55] challenges of scaling a city, you can kinda pick and choose what pressures. If you really wanna have a quiet [00:05:00] city or a, a rail-only city or whatever be able to tune the game engine basically. [00:05:05] So that, that's gonna be a little bit in the future, but that's kinda the road we're on with that.
[00:05:09] Andrew: [00:05:10] So that will go out soon, hopefully from Apple, and then the next [00:05:15] version is already in the works and we're
[00:05:18] Nathan: Yeah
[00:05:18] Andrew: ship [00:05:20] stuff
[00:05:20] Nathan: Yep.
[00:05:20] Nathan: Okay.
[00:05:21] Andrew: stuff
[00:05:22] Nathan: I, I'm gonna pitch a random idea here. I think it'd be really [00:05:25] cool given, in my opinion, you're, you're like a history buff and you, lots of things about [00:05:30] history. It'd be very interesting if you built challenges from things that actually happened in history, and then [00:05:35] those cities have to rec- how do you, how would you, if you were Japan, manage like the[00:05:40]
[00:05:40] Nathan: nuclear,
[00:05:41] Andrew: Yeah, half your city explodes and then
[00:05:43] Nathan: Yeah, exactly. [00:05:45] Or like if you're, if you're a city and you're going bankrupt, how do you come
[00:05:48] Nathan: out of that, right? Like
[00:05:49] Nathan: I from [00:05:50] some anec-
[00:05:50] Andrew: public art collection through Story Detroit. Shout out [00:05:55] Detroit.
[00:05:55] Nathan: Yeah, but I, I think it'd be fun just to kind
[00:05:56] Andrew: yeah, big
[00:05:57] Nathan: things that... Things are harder than they [00:06:00] look, right?
[00:06:00] Nathan: So
[00:06:01] Andrew: No, actually that's a good idea. I kinda like [00:06:05] the, come up with historical stuff to to, to do as a [00:06:10] challenge. Yeah. Th- so yeah, so I, I guess yes, that
[00:06:14] Nathan: So then [00:06:15] derailed it. Sorry. Yeah.
[00:06:16] Andrew: No, no, no. I'm just, I'm thinking about it. It sure. The [00:06:20] version, you know 26 10.3, why not?
[00:06:24] Andrew: [00:06:25] Um, yeah, 10.1 is in the App Store waiting for release, and 10.2 [00:06:30] will be as soon as...
[00:06:31] Andrew: I can only submit one build at a time for approval, which is really [00:06:35] dumb
[00:06:35] Nathan: Yeah, you gotta pull out the other one if you wanna put in a new one.
[00:06:38] Andrew: Yeah, come on Apple. Like I get a new [00:06:40] like anthropic update every couple hours it seems like, and they can't review [00:06:45] stuff for more than like weeks. And there's... They say it takes 48 hours. I've complained about this before, but it's [00:06:50] gotten a lot worse and like they gotta figure it out.
[00:06:52] Andrew: But l- yeah, so challenge stuff we're... And then the [00:06:55] next version though we're doing ship stuff, so there's a port scenario that [00:07:00] you, you... So I guess the scenario is you start with a foundational kind of economy [00:07:05] also and then you kinda go through time. So there is a port one, but there's like no ships or anything.
[00:07:09] Andrew: [00:07:10] Like they're there in
[00:07:11] Andrew: like theory.
[00:07:13] Nathan: it's clicking. Now it's
[00:07:14] Andrew: Yeah, we're [00:07:15] doing ship stuff.
[00:07:16] Andrew: Yeah, so the- they're, we're building some models for ships and so there'll be [00:07:20] actual like shipping and then adding rivers. So right now it's really just kind [00:07:25] of lakes and coastlines, and so rivers will come in and then bridges 'cause you gotta be able to get [00:07:30] over rivers.
[00:07:31] Andrew: So some kinda new challenges and new dynamics [00:07:35] to build in the city, and then all of that will go into the game for free, and if you wanna save and/or support [00:07:40] development, you can buy saving. If you wanna do the whole thing I [00:07:45] forget what it's called, I think it's called Ultra, but you can unlock everything currently for sale [00:07:50] and then everything in the future.
[00:07:51] Andrew: So it's yeah, it's pay once, you're done paying, [00:07:55] and you help support the development of it. Didn't really plan to keep it going [00:08:00] so long, but I'm like...
[00:08:00] Andrew: I, if you talked... Yeah, but I don't know, maybe we could talk about AI [00:08:05] now. One of the things that... I, it's just I keep coming up with kind of stuff that [00:08:10] I, I wanna...
[00:08:10] Andrew: I don't really have time to play games anymore, but like I really like a good city [00:08:15] builder and when I like take a flight is usually when I have... I'll do some work and then inevitably I'm [00:08:20] just like, I need to just kinda chill out. I'm usually going on vacation. I don't travel for work too much [00:08:25] anymore. But I don't mind traveling either, it's just like I just, I don't know.
[00:08:29] Andrew: I, we've [00:08:30] talked about I was just in Toronto and then I'm heading to I think New York [00:08:35] in a f- month or so for
[00:08:37] Andrew: leisure.
[00:08:38] Nathan: Okay
[00:08:39] Andrew: Yeah, I [00:08:40] wanna get to New York once a quarter. Sometimes it just doesn't happen. And like for folks, I live across the [00:08:45] country, so it's four and a half to six-hour commitment. Yeah,
[00:08:49] Andrew: it's
[00:08:49] Nathan: [00:08:50] time, hour difference, you lose a, basically a day for travel, right?
[00:08:52] Andrew: Yeah, no, you pretty
[00:08:53] Nathan: Yeah.
[00:08:54] Andrew: lose a day [00:08:55] going. Coming back is a little bit better 'cause I can land at the relatively late [00:09:00] hour
[00:09:00] Andrew: here and it's
[00:09:00] Nathan: still-- depending on how long you're on the
[00:09:02] Andrew: Yeah, you get screwed up
[00:09:03] Andrew: time-wise.
[00:09:04] Nathan: [00:09:05] exactly.
[00:09:06] Andrew: Yeah, I used to travel a lot for work too, so it's like I'd get some work done and [00:09:10] then and have a little bit of leisure.
[00:09:12] Andrew: And it's like I don't, I can't watch movies [00:09:15] anymore.
[00:09:15] Andrew: I can, but it's... I just I've just, I don't know, Hollywood needs to step their [00:09:20] shit up. Although we should talk about a movie. Anyway, story being done. So yeah, so I, I, I wasn't [00:09:25] really planning on doing a ton. I was gonna submit for the competition, and I keep having ideas [00:09:30] and then we've talked about it a little bit, but like my AI setup is s- [00:09:35] s- works.
[00:09:36] Andrew: I don't know what else to say. So I've just kinda dispatched an agent to [00:09:40] just take my like brain dump ideas and build a plan for me to review. [00:09:45] And when I have some free time, I'll take a look, give it some feedback, and let it go, and it's been doing a pretty good [00:09:50] job. I'll be doing that. I, I think at least through the end of the year it's gonna get updates and stuff [00:09:55] as I see stuff.
[00:09:55] Andrew: And if f- folks are interested, we have a Discord, we have comments, email the [00:10:00] company. I don't really care. I want I wanna do stuff people want. We've talked about it though, like [00:10:05] game design, like oof, it's hard.
[00:10:07] Andrew: Like, it's really [00:10:10] hard.
[00:10:10] Andrew: Yeah.
[00:10:11] Andrew: and I had no idea, like... And I guess I- maybe a different [00:10:15] type of game would've been less challenging because like I, I don't know enough about [00:10:20] game design.
[00:10:20] Andrew: But what I've learned is that, in like a, a turn-based, [00:10:25] what do they call it? It's a progressive clock. I forget what the term is. [00:10:30] But the, a turn base, like a bunch of stuff happens, you stop, and then you can [00:10:35] basically compute the next s-
[00:10:36] Andrew: n- next turn. That feels really easy after trying to [00:10:40] do a real... It's like a real time strategy is...
[00:10:42] Andrew: I don't know if that's technically what the simulators [00:10:45] are. But that's hard because there's constantly stuff changing, right? And you're,
[00:10:49] Andrew: [00:10:50] and you expect to build a building that it's gonna take a little bit of time to build. Like it just doesn't just pop up [00:10:55] overnight. I suppose it could, but I'm trying to give it some kind of, Part of the strategy is like you can't just build a [00:11:00] building and then instantly you start getting tax revenue, right? It's like you're trying to that's the strat- the, [00:11:05] the financial strategy isn't very robust right now. But in, November's [00:11:10] timeframe
[00:11:10] Nathan: Yeah, that's the fun part of it.
[00:11:11] Andrew: that's, yeah, and like taxation and [00:11:15] yeah. It's, it's all there.
[00:11:17] Andrew: Like there's a generalized taxation right now which you have to fight [00:11:20] against, but it's not super hard. But I w- and I, and again, I wanna make it, I wanna make this [00:11:25] in a way that I present the pathway that I think is fun, but then make it so people [00:11:30] can modify it and have the version. 'Cause if you don't care about taxes and just wanna have a [00:11:35] fun little city builder and build your cool little city and
[00:11:38] Andrew: yeah, why not?
[00:11:39] Andrew: It's, it's [00:11:40] your game. You wanna play it however you want. And, and that makes things a little more challenging. But yeah, the, the real time [00:11:45] cl- so every time a clock ticks, you know, there's a bunch of shit that has to happen And that was a lot more [00:11:50] complicated, I think, than I initially thought. I don't honestly know what I...
[00:11:53] Andrew: I was like, "Ah, this can't be that [00:11:55] hard," and then I start doing it, I'm like, "Eh."
[00:11:57] Nathan: Yeah that's life, right? You think about, "Oh,
[00:11:59] Andrew: Yeah. [00:12:00]
[00:12:00] Nathan: be that bad," until you actually get into it and you're,
[00:12:01] Nathan: "Oh,
[00:12:02] Nathan: This is harder
[00:12:03] Nathan: than I thought."
[00:12:04] Andrew: I've done in [00:12:05] life has been really, really, really, really, really hard, and I'm like, "Ah, it can't be that hard." I'm like, [00:12:10] "Actually, I'd rather I'd rather work on life-saving technology because it's [00:12:15] easier."
[00:12:15] Nathan: Yeah. So in a weird way, that-that's what gives it those things the appeal [00:12:20] is,
[00:12:20] Andrew: But,
[00:12:20] Nathan: you but people wanna do it anyway.
[00:12:22] Andrew: No, and, and, and the, I think part of the [00:12:25] challenge too is like what is fun,
[00:12:27] Andrew: right? 'Cause if you're working in tech and you're working on like a [00:12:30] customer or business application, there's like an objective or there's goals or like aspirations and you can meet those. [00:12:35] But when it comes to like has to be enjoyable like
[00:12:39] Andrew: [00:12:40] fun.
[00:12:40] Andrew: Like something that's enjoyable isn't necessarily fun. Something that's fun might [00:12:45] not be enjoyable. You're like, "Oh, that rollercoaster was fun. I never want to do it
[00:12:48] Nathan: Yeah, yeah, yeah. I
[00:12:48] Nathan: mean, it, it actually [00:12:50] ties in very nicely here with the idea of gamification of things,
[00:12:52] Nathan: right? When you see numbers go up,[00:12:55]
[00:12:55] Nathan: some people think that's fun, right?
[00:12:56] Andrew: Ja it turns into this slog of just getting the numbers to go up, [00:13:00] and then if you look at it on a whole, that process is no longer fun.
[00:13:03] Nathan: You're just trying to see the numbers go up. [00:13:05] Whereas some people
[00:13:05] Nathan: might be enjoying the process even if the numbers don't change too much, but it's more about the [00:13:10] process.
[00:13:10] Andrew: Yeah, and it gives people some, some people anxiety and, and so I'm trying to [00:13:15] take all that into account as I build, and then I thought about it this [00:13:20] morning. We were talking earlier, I can't talk about it yet, but I was like, I really should have built like [00:13:25] a game, map editor kind of idea.
[00:13:28] Nathan: Oh, oh, oh, [00:13:30] yeah
[00:13:30] Andrew: And so that's something that I'm thinking about of what would it look [00:13:35] like to do like sandbox mode so you just...
[00:13:37] Andrew: There's just no goal, just growth and like fun, [00:13:40] and then but make it so that you can kind of like edit the city and [00:13:45] be able to just build for free kind of idea, and then build it however you wanna kinda go, [00:13:50] and then do your own, build your own challenge or something.
[00:13:53] Andrew: Or,
[00:13:53] Nathan: There's a... I don't know if [00:13:55] Apple's Game Center does this, but that it basically opens up the custom content creator angle, right?
[00:13:59] Andrew: [00:14:00] Yeah.
[00:14:00] Nathan: their own scenarios with the map editor
[00:14:02] Andrew: Ex- exactly, and that...
[00:14:04] Andrew: And that would [00:14:05] be kind of the goal of you build a big transportation city, and there's a big water through it. [00:14:10] Or you build a really cool city, and then you flood half of it to try to reconstruct it or [00:14:15] those kind of ideas. We talked about that originally, too.
[00:14:18] Andrew: I would love to make this modular for [00:14:20] folks to, to be able to mod and stuff. It's hard with the Apple [00:14:25] platform, but there's been a couple games that I've found when I was researching this [00:14:30] that you- they have mods. So I don't
[00:14:32] Andrew: know what that's gonna look [00:14:35] like.
[00:14:35] Andrew: But yeah, on iOS
[00:14:38] Nathan: you could do that. I did not know you could do [00:14:40] that. Okay
[00:14:40] Andrew: Yeah, and s- but I don't know if it has to go through them.
[00:14:43] Andrew: It wasn't clear and I [00:14:45] didn't get it far enough into it, but if you have to
[00:14:47] Nathan: Sure
[00:14:47] Andrew: it to the developer and then they approve it, and then [00:14:50] they slipstream it in. But I'm, I have to, I have to dig into the bowels of this, but I [00:14:55] think, you might be able to save some files to your, what do
[00:14:59] Nathan: [00:15:00] Local. Yeah. Yeah, yeah, yeah
[00:15:02] Andrew: you could just load a game save or something
[00:15:04] Andrew: [00:15:05] maybe. I'm hoping. Don't...
[00:15:07] Andrew: Yeah, don't hold me to this, but I'm hoping [00:15:10] that it could just be, like, a save file swap or something
[00:15:13] Nathan: Yeah, yeah. [00:15:15] That would be interesting. Or an asset package swap where you
[00:15:18] Nathan: would download [00:15:20] certain types
[00:15:20] Andrew: When we were talking about I would love to do artwork and make it more open too. It's [00:15:25] not open sourced yet. I, I just haven't got around to it. I'd probably open source it. I don't really have a problem with... [00:15:30] I just haven't had the time. And but yeah, so I might do that, and [00:15:35] then for open sourcing it, it'd be awesome for be- people to be able to swap out their artwork or f- [00:15:40] I would love to do, monuments for famous cities kind of [00:15:45] stuff, and that would be fun to,
[00:15:47] Andrew: to do.
[00:15:47] Andrew: I have all these ideas, I just... It's like I don't... [00:15:50] Th- this isn't my focus. This was done for a competition, and the competition technically ends [00:15:55] the end of this month. I guess it, the build part ended. Now it's like the grow part. [00:16:00] So again, download the game, rate it, go like it, pay me money, don't pay [00:16:05] me money.
[00:16:05] Andrew: I don't know. I don't really care.
[00:16:06] Nathan: Yeah.
[00:16:08] Andrew: I think it's worth it, to be [00:16:10] honest, and I guess maybe an argument for folks too is that If you don't [00:16:15] pay for stuff, it doesn't exist
[00:16:17] Nathan: Yeah.
[00:16:18] Andrew: it doesn't continue
[00:16:19] Nathan: Yeah, [00:16:20] we've grown accustomed to not having to pay for things, and
[00:16:22] Nathan: unfortunately that's,
[00:16:23] Andrew: and, you know, and, [00:16:25] and,
[00:16:25] Andrew: yeah, oh, here's business kind of thing. So I- I have, I have an apps business, [00:16:30] and there's other apps that we put out, and I get feedback from [00:16:35] them, and it's almost exclusively great feedback, which is an anomaly in product
[00:16:39] Andrew: management. Anyone [00:16:40] who's...
[00:16:40] Nathan: Almost keyword.
[00:16:41] Andrew: Yeah. Yeah, there's a few of you out there.
[00:16:43] Andrew: No. It, [00:16:45] it, it's, it's actually strange. I've worked in tech for too [00:16:50] long and haven't really [00:16:55] ever had a product or something I was working on where I was getting just [00:17:00] mostly good
[00:17:00] Andrew: feedback most of the time.
[00:17:02] Nathan: Yeah, yeah,
[00:17:02] Andrew: yeah
[00:17:04] Andrew: And so that's [00:17:05] refreshing, and it's nice, and I really appreciate when people just if you email in support and be [00:17:10] like, "Hey, this is awesome.
[00:17:10] Andrew: Thanks," that's cool. I know some people feel weird 'cause they think they're [00:17:15] like they, they, "I don't need support." And it's like, "Oh, you can share through there. That's totally [00:17:20] fine." I, I, I guess that's so that's probably, I don't know, 30% of folks are just like, "Thanks [00:17:25] so much. This, this app's great, this cool functionality," whatever.
[00:17:28] Andrew: I gotta be careful with math [00:17:30] here to make sure it's, the math is right. I'd probably say 40% of folks are really [00:17:35] happy, and then they have either, like, constructive feedback or a [00:17:40] what abouts that what if it did this or it'd be really cool if it did this or kind of [00:17:45] suggestions, but also really well thought out suggestions.
[00:17:48] Andrew: This is clearly a pain [00:17:50] point and you've gone through enough, like,
[00:17:52] Nathan: Nice
[00:17:53] Andrew: thinking about it to relay it [00:17:55] in a way and, and a lot of times if I see it, I'll just do it or [00:18:00] I'll pop, pop it over to an agent and be like, "How hard is this?" And it'll be like, "It's done," and I'll be [00:18:05] like, "Great, just push to test flight build."
[00:18:07] Andrew: And I actually then I get feedback from people and there's a couple reviews in [00:18:10] the store even where they're like, "It's amazing turnaround time." And I'm like, "Yeah, the, our, [00:18:15] our DevOps are solved, Like-
[00:18:18] Nathan: our engineers work 24/7.[00:18:20]
[00:18:20] Andrew: Seriously. You
[00:18:21] Andrew: know the agents don't sleep and I, I, [00:18:25] they basically run 24/7. There's human work that's involved, but it's, it's [00:18:30] becoming so much less.
[00:18:32] Andrew: So, that I, I love it, yeah, if you have ideas, [00:18:35] feedback, like I'm not gonna promise we're gonna do everything, but we do most stuff people ask for as long as [00:18:40] it's not crazy and it fits in the, the realm. Your pods will not have [00:18:45] iCloud support. I'm really sorry for... It's either one or two people who keep writing [00:18:50] in.
[00:18:50] Andrew: I, I just, I don't even know how that would [00:18:55] work, right? I, I just, it, I... The only thing that I [00:19:00] could think of to do for that would be to like, do almost like a [00:19:05] backup, which would just be like an export
[00:19:07] Nathan: to the, yeah, back up to the cloud and then how would [00:19:10] you
[00:19:10] Andrew: And then you would have to import it on your other device.
[00:19:13] Nathan: device. Yeah.
[00:19:14] Andrew: [00:19:15] Oof. And
[00:19:15] Nathan: I don't know how the sync would work for
[00:19:17] Andrew: that's not a sync, right?
[00:19:18] Andrew: It, it would...
[00:19:19] Andrew: [00:19:20] And,
[00:19:20] Andrew: and the, there's an open format for this that [00:19:25] everyone supports, and it's pretty basic, right? It's, it's like the podcast name, [00:19:30] URL, feed, data, that kind of stuff. There isn't a lot of other stuff there, and I guess I theoretically could write [00:19:35] something for that, but it's, it's it's a say no moment.
[00:19:39] Andrew: [00:19:40] And we've talked about this too. I struggle a little bit 'cause I'm just like, "I'll just give it to an agent and they'll figure out some crazy way to [00:19:45] do it." The problem
[00:19:45] Nathan: like learning how to say no kind of thing. Yeah, Yeah.
[00:19:48] Andrew: And, and, and I, [00:19:50] I've noticed this a lot with AI stuff is it's like I can just have the AI do it and it's fine, but then eventually you [00:19:55] just get feature bloat and your software just crashes 'cause it's just too complicated or too hard [00:20:00] or doesn't make sense.
[00:20:01] Andrew: And so something like that, like theoretically could work, and I've had people pitch [00:20:05] ideas and like that... They're great ideas, like, but I, I just won't do it. It is open source, so if you wanna [00:20:10] write an iCloud connector and submit it, I'll take a look. But you'd have to probably [00:20:15] push the whole database over, and then Apple has all these weird rules about stuff, and I just...
[00:20:19] Andrew: I... And it [00:20:20] feels fragile too. And, and I guess like my [00:20:25] other piece of that is that you can use the GPodder sync for [00:20:30] free. You can host your own for free. You can use YourPod sync which is also free, which is something we [00:20:35] host. Or you can pay for a more robust sync for... It's a few bucks a [00:20:40] month, right?
[00:20:41] Andrew: So i- it's like I... There's other solutions that I think are [00:20:45] actually better for it instead of an iCloud. And I know... I'm not trying to pick on people. I just wanna [00:20:50] be clear. Like it's a really good example of like theoretically this makes sense. I can [00:20:55] see a few ways of how it would actually work.
[00:20:57] Andrew: But I think it would be a [00:21:00] tremendous amount of work for a very not good experience, and it just... I don't think it would [00:21:05] work well. And there's already, I would say, better solutions out there. So if you're totally in [00:21:10] iCloud and you're tot- And, and there is like an iCloud sync function [00:21:15] where, it's like iCloud drive is a file sync, but you can also do like application [00:21:20] state stuff through it.
[00:21:21] Andrew: But there's a lot of... Yeah. But there's a lot of like [00:21:25] asterixes on
[00:21:26] Nathan: What you can actually...
[00:21:27] Andrew: Yeah.
[00:21:29] Andrew: It's not [00:21:30] designed to be a real time file stream, yeah
[00:21:33] Nathan: As well as the way that they've [00:21:35] designed their sandbox applications.
[00:21:37] Andrew: Yeah.
[00:21:37] Nathan: weird to have to figure out on
[00:21:39] Nathan: their [00:21:40] side how
[00:21:40] Andrew: and this... Yeah, and th- and that's... And then it's like, app state stuff kind of [00:21:45] works, but then trying to get it ported, like if you have a Mac or something. I guess spoiler alert, [00:21:50] we're gonna release a... I'm t- I'm talking for folks who don't know, I'm talking about Your Pods, which is a podcast app that [00:21:55] we, we have, in the, the app portfolio.
[00:21:58] Andrew: We are gonna release a [00:22:00] Mac app, in I think the next version. So there's a version that we're waiting on [00:22:05] Apple for, to approve. That's 26.9. [00:22:10] 26.10 will come as soon as I can get it out the door. It's like I'm waiting for Apple to approve the [00:22:15] current version so then I can submit 26.10. 26.10 will have a native Mac app.
[00:22:19] Andrew: So [00:22:20] we've, we've had, I forget what Apple calls it, but you can use an iPad app on [00:22:25] the Mac.
[00:22:25] Nathan: On the map. Yeah, yeah
[00:22:26] Andrew: You know, and it works fine. It's just like you don't get media keys, [00:22:30] you don't get the native Mac experience. It's a little kludgy, but it works. And so that's been [00:22:35] kind of the Mac way for a while. But we will cross the Rubicon and have a [00:22:40] full...
[00:22:40] Andrew: It works today. There's a few kind of weird bugs that I'm encountering [00:22:45] that I'm, I'm trying to... With the UX that we're trying to like... And like me specific stuff, so I don't know if it's [00:22:50] just me or if it's what's going on.
[00:22:52] Nathan: No, there are some things that you would want, I think, [00:22:55] that you would try and build that I, I feel like a lot of people wouldn't use.
[00:22:59] Andrew: Yeah [00:23:00] Yeah, I know, right? So and it'll be parity with the, the ma- the iOS and [00:23:05] iPad app, so you'll be able to just, it should just work, right? It just... It'll look a little [00:23:10] different. We're trying to adapt it to the Mac, so all the stuff that you would expect, you [00:23:15] know, Mackie land to be like, it'll do it.
[00:23:18] Andrew: And so that, that'll be [00:23:20] out, hopefully in the next week or so, soon. And then there's some [00:23:25] Mac-specific features that I don't wanna talk about quite yet 'cause I don't know [00:23:30] how well it's gonna actually work. But we've had a lot of folks asking for, AI [00:23:35] features, and so there's a couple different pathways for AI features in the podcast.
[00:23:39] Andrew: [00:23:40] Yeah, no, a lot of folks have asked.
[00:23:42] Nathan: Oh, interesting. We haven't talked about this. Okay, cool
[00:23:44] Andrew: Yeah. [00:23:45] Surprise. So we're, we're doing some really cool... You're testing one [00:23:50] of them, so we're doing- we're going that
[00:23:51] Andrew: route. And then we're also looking at,
[00:23:54] Andrew: [00:23:55] we're also looking...
[00:23:56] Andrew: Yeah, we've asked the question of what does local AI look [00:24:00] like?
[00:24:00] Nathan: Yeah. Yeah. Okay
[00:24:01] Andrew: And, and so for the Eurpods crew, [00:24:05] we're gonna have that solved in a really easy way where you don't have to have a [00:24:10] bunch of bullshit technology strung together.
[00:24:12] Andrew: I think it's gonna just work, knock [00:24:15] on wood. That'll have some caveats to it, but that's gonna be part of the Mac app. You can [00:24:20] do some cool stuff, like on device on the phone, but because iOS [00:24:25] is so resource constrictive, like if you're not plugged in to a power
[00:24:29] Andrew: outlet, it, [00:24:30]
[00:24:30] Nathan: Yeah, it,
[00:24:30] Nathan: destroys the
[00:24:31] Andrew: life.
[00:24:31] Andrew: iOS... Yeah. And, and it's a shitty experience, so it's [00:24:35] like maybe we'll shim it in and put a big old disclaimer of "This [00:24:40] is not great to use, and it sucks, but if you want to go for it, go for it." [00:24:45] but whereas the Mac you have a lot more less, or I should say you have a lot less constraint, a lot [00:24:50] more freedom to do some stuff.
[00:24:51] Andrew: So we have some really cool features planned for that.
[00:24:54] Nathan: Oh, I
[00:24:54] Nathan: think I [00:24:55] see.
[00:24:55] Nathan: Okay. Okay. yeah.
[00:24:56] Andrew: It's, it's like the thing you're testing, but [00:25:00] local
[00:25:00] Nathan: Yeah, yeah. Okay.
[00:25:01] Andrew: Yeah.
[00:25:02] Andrew: It's a variation 'cause there's, Yeah
[00:25:04] Andrew: [00:25:05] And then, there's some folks who have asked about running local AI through the app and stuff, and [00:25:10] so I think the Mac will be the conduit to that.
[00:25:13] Andrew: We'll see. I like talking [00:25:15] about stuff that don't hold me to it 'cause it's... That's just designed. It's... We haven't even [00:25:20] implemented that kind of stuff yet. But, Ollama is, is floating out there in my mind of what would it [00:25:25] look like to have that as a back
[00:25:26] Andrew: end and,
[00:25:26] Nathan: what I, think you're, you're going
[00:25:28] Andrew: Yeah.
[00:25:28] Nathan: to Ollama for [00:25:30] running
[00:25:30] Andrew: Yep. yeah.
[00:25:31] Andrew: S- And then, and the problem with that, though, is your mileage varies based on your [00:25:35] hardware, and a lot of people don't really understand that, and so that'll be a whole kind of how we go [00:25:40] about that. But I'm really excited. It's been stuff I've been wanting to do for a while. It's just hasn't [00:25:45] gotten, gotten there yet for various reasons.
[00:25:48] Andrew: And I don't love to talk about [00:25:50] future-looking stuff. Maybe CEO brain kicks in and it's like I'm not a public company, so I don't really [00:25:55] care, but it's like I also, start promising shit and you don't deliver on it, you look like an idiot. But we're, we're [00:26:00] definitely gonna do it. It's just I don't know how it's gonna just shape out yet.
[00:26:03] Andrew: That's a really long way of [00:26:05] saying... And then there's 5 to 10% of customers who are just unreasonable, which is just business.[00:26:10]
[00:26:10] Nathan: Yeah.
[00:26:11] Andrew: You know who you
[00:26:11] Andrew: are.
[00:26:12] Nathan: out there.
[00:26:13] Andrew: But no, I've, I've had... I, I,
[00:26:14] Andrew: had...
[00:26:14] Nathan: they [00:26:15] actually listen to the cast?
[00:26:16] Andrew: Maybe.
[00:26:17] Nathan: Possibly.
[00:26:18] Andrew: There's one person [00:26:20] who left a comment, and I was just kind of like, "That's just factually incorrect."
[00:26:23] Andrew: "I don't know what to tell you." [00:26:25] And as a business or as a you can't be like, "You're an idiot," but [00:26:30] sometimes I want to be.
[00:26:31] Nathan: Which you can. Yeah, yeah,
[00:26:32] Andrew: But I like I'm... It might be a [00:26:35] misunderstanding, so I was like, "Hey, this is how it actually works," in a nice way. But it, it's when people come [00:26:40] at you with factually incorrect stuff and then slam you discourage other people from [00:26:45] pursuing.
[00:26:45] Andrew: If you have criticism, I'm all for it. Write into support. I'd rather you just write [00:26:50] into support. We'll take a look at it and we'll probably fix it. Leaving an app comment or [00:26:55] social media or whatever that's not a great way to get stuff solved. So if you wanna just complain, I [00:27:00] guess go for it.
[00:27:01] Andrew: If you wanna actually have something fixed or have someone take a look at it, [00:27:05] just send an email to support
[00:27:06] Nathan: Unfortunately, I think people are starting to be trained that if you just @ [00:27:10] someone on like Twitter or X or
[00:27:11] Nathan: whatever, then that's one way to get the attention [00:27:15] or-- And then if
[00:27:16] Nathan: they have a-- Yeah, yeah, exactly.
[00:27:19] Nathan: I don't think we [00:27:20] should be moving to a world like that, but
[00:27:21] Nathan: unfortunately, that's kind of the
[00:27:22] Nathan: space that we're in
[00:27:23] Andrew: some people who email me [00:27:25] directly, and like also not the best way. I do respond to my email, [00:27:30] but if it's not expected or critical, like it could take weeks, maybe [00:27:35] months. I get a lot of crappy email lately, especially like all these AI like sales [00:27:40] bots. I, I'm so sick and tired.
[00:27:42] Andrew: Like I'm getting to the point where I almost need an AI to [00:27:45] s- to work my
[00:27:47] Nathan: Yeah, you should create an AI to handle those AI, and then [00:27:50] you have an infinite AI loop of
[00:27:51] Nathan: just
[00:27:51] Nathan: managing each
[00:27:52] Nathan: other.
[00:27:52] Andrew: just losing the, the [00:27:55] feed, right? Like I, I can't... I get hundreds of emails a [00:28:00] day sometimes and like the company doesn't use email, so it's like it's not noi- [00:28:05] typical company noise. It's just literally all external communication and most of, like I'd [00:28:10] say 99% of it is just total garbage.
[00:28:13] Andrew: And I didn't sign up for it. [00:28:15] Just like people are trying to pitch me shit. Like that's a strategy, I just don't think that... I don't know if that [00:28:20] ever worked to be honest, but it definitely doesn't work.
[00:28:22] Andrew: And
[00:28:22] Nathan: a weird numbers thing where people just hope that[00:28:25]
[00:28:25] Andrew: Probably. Yeah,
[00:28:27] Nathan: of the people
[00:28:28] Andrew: If your strategy is to, [00:28:30] to mislead me in the subject line to get me to open it to then try to
[00:28:34] Andrew: like...
[00:28:34] Andrew: [00:28:35] That's not... I, I ban you, I block you, and I just hope you're never a customer again 'cause [00:28:40] nobody in the company will ever see your email. You get shadow banned if you send me dumb [00:28:45] shit.
[00:28:45] Andrew: But yeah, I've had people...
[00:28:46] Andrew: Yeah, I've had people reach out to me and like that's... I love feedback, [00:28:50] I love... And I don't want to know that my email's hard to guess, but like also support's [00:28:55] better if you actually want timely feedback.
[00:28:58] Andrew: I feel bad 'cause there was one person who emailed [00:29:00] me like, I don't know, maybe a month ago. It was actually a really good idea, and they had a problem that I think we [00:29:05] fixed anyway or we like patched or something. And it was like they sent [00:29:10] it to support and then support was like, "Hey," " Sorry, but here you go."
[00:29:14] Andrew: But [00:29:15] yeah, I don't know. So that, that... There's another thing I was gonna talk about with that kind of [00:29:20] stuff. That's, that's all coming, crap, it was a really good point. I was like we could talk [00:29:25] about thing It's the No Script Show
[00:29:27] Nathan: That's what happens. A,
[00:29:29] Andrew: Yeah
[00:29:29] Nathan: no, I, I [00:29:30] think the... Yeah, how did we get here? We, we were talking about yeah,
[00:29:33] Andrew: People feed back [00:29:35] and complained, and then we got into the brainstorming stuff.
[00:29:39] Nathan: Yeah, [00:29:40] yeah, yeah, yeah. No, I, I was, I,
[00:29:41] Nathan: think the roadmapping. Okay.
[00:29:43] Andrew: Yeah,
[00:29:44] Andrew: [00:29:45] Oh,
[00:29:45] Nathan: the thing about it, I think they just said... Oh, no, go. If you had the, the point that came back in, go ahead.
[00:29:49] Andrew: Yeah.
[00:29:49] Andrew: [00:29:50] It's just like working with AI stuff, and oh, I remember now. I have a blog [00:29:55] now.
[00:29:55] Nathan: Oh, that's what you wanted. Okay
[00:29:56] Andrew: That's what it was. And we were talking about or maybe, I [00:30:00] don't know, maybe I was thinking about it. But, I, I th- I think I mentioned it last week. I'm not gonna [00:30:05] plug it. I'm talking about it, but I'm not gonna plug it.
[00:30:07] Andrew: If you wanna find it, it's not hard. But [00:30:10]
[00:30:10] Nathan: Okay.
[00:30:11] Andrew: I wrote a post, I don't know if you actually read it, but I wrote a post about,[00:30:15]
[00:30:15] Nathan: Yeah, yeah
[00:30:15] Andrew: like, how to be successful working with AI, and [00:30:20] that might be worth talking about. I don't remember how much detail you, [00:30:25] you've got or, or how much or well, how well you read it. But the gist of it,[00:30:30]
[00:30:30] Andrew: yeah, the, the gist of it was, You have to be a player coach with [00:30:35] AI, like at least right now.
[00:30:36] Andrew: Like maybe it'll get better. And when I say AI for the purpose of this [00:30:40] talk, it is large language models, generative AI, what most people [00:30:45] think AI is. There's a whole world of AI that is other stuff, but that's what we're [00:30:50] at. And you have to be a player coach and, and you have to be a true player coach.
[00:30:54] Andrew: And I [00:30:55] think I guess to define player coach is a skill set in which you have [00:31:00] the ability to, to be an individual contributor, to execute, to do the thing. So play the game, so [00:31:05] to speak, like you know how to do it. And then the coach part is you can take that [00:31:10] experience and, and help translate that into actions or words for other [00:31:15] people to then improve.
[00:31:17] Andrew: And it's not teaching and it's not [00:31:20] controlling, and I think people really get that part screwed up where it's like sometimes a coach will have to [00:31:25] teach you something, but at, especially at a professional level, like you should know what you're doing and they're [00:31:30] there to help you get better. And so they might teach you a different technique to, to [00:31:35] run in a different way to be faster, more efficient.
[00:31:37] Andrew: But like they, they... You know how to run if [00:31:40] you're a, an athlete, right? And, and they're there to help you. And then they're also there to make sure y'all gel as a [00:31:45] team and, and have a strategy to, to try to win the game and, and make the [00:31:50] team comfortable with that strategy to, to execute it.
[00:31:54] Andrew: But they're [00:31:55] not jumping on the field and shooting the basket for you or hitting the ball for you or whatever. Whereas like [00:32:00] that's where I think a lot when you translate it into business, you get like overbearing [00:32:05] executives or overbearing middle management who jump in and try to just control everything either because [00:32:10] they are capable individual contributors in a role that they don't quite [00:32:15] understand, or they are really, really bad at the role that they're in, and that's the only way that [00:32:20] they know how to make stuff happen or that's what they like doing.
[00:32:24] Andrew: [00:32:25] Um, and that, that doesn't work. Like it will work with AI, but you're [00:32:30] not going to unlock the, what I would say the full potential is. And we talked a [00:32:35] little bit about it last week or the maybe the week before, where there's there's a lot of people right [00:32:40] now claiming some crazy shit about AI and I...
[00:32:42] Andrew: All I can say is like everything that [00:32:45] I've read or seen from a lot of people is the stuff I experience every day. [00:32:50] Like it's not sentient, it's not gonna take over the planet, it's not alive. But if [00:32:55] you are a true player coach and can enact those skills with [00:33:00] AI, gains for some things that are like inexplicably hard to [00:33:05] explain And it's mostly coding.
[00:33:07] Andrew: There's some other stuff that... But I would say especially in the [00:33:10] realm of coding it, it, it makes you superhuman. I, I, I definitely [00:33:15] experience on a daily basis doing things that would have normally ta- would, that would [00:33:20] take me and a normal human team of three or four people a few hours to [00:33:25] discuss, let alone implement.
[00:33:27] Andrew: And with AI in the right, working in the right [00:33:30] way, it's done
[00:33:32] Nathan: Yeah. Yeah. Yeah, yeah.
[00:33:34] Andrew: and in [00:33:35] production. And then I've moved on. And the reality is it's done, and then [00:33:40] I wait for the AI to finish while I do other stuff, and then by the time I [00:33:45] forget about I even did it, it's like, "Oh, hey, it's pushed," and go check it out.
[00:33:49] Nathan: Yeah, [00:33:50] yeah. I think one, one of the elements with the player-coach idea is, is I think people overlook the idea of [00:33:55] the domain. So usually when there's a
[00:33:56] Nathan: player-coach, they're coaching in the same domain. So they have some [00:34:00] way of, of being able to add value back that's relevant to whatever the task is.
[00:34:04] Andrew: Yeah
[00:34:04] Nathan: not gonna [00:34:05] have great results if you don't exactly know how to direct the agent or whatever it is to do [00:34:10] certain types of things. so the other thing I think a lot of people seem to either not consider [00:34:15] or jump into enough is sometimes even as a player-coach, you'll need to do to [00:34:20] show. And with AI
[00:34:22] Andrew: Yes
[00:34:24] Nathan: providing a [00:34:25] template or sketching something out that actually
[00:34:27] Nathan: works a great deal. Even [00:34:30] if you don't know exactly what you need, but you have some semblance of what you're hoping for, right? So [00:34:35] if it's a mock-up of some idea, just draw boxes or
[00:34:37] Andrew: Yeah.
[00:34:38] Nathan: right?
[00:34:38] Nathan: Just to give
[00:34:38] Nathan: some frame of [00:34:40] reference
[00:34:40] Andrew: i- impromptu, oh God, I draw all the time. That's a huge hack. Like we talked about this [00:34:45] probably years ago now on the podcast that the vector image space is far more [00:34:50] dense than language
[00:34:51] Nathan: Yeah, write down the words, right?
[00:34:53] Andrew: It's a huge [00:34:55] hack. What does that mean? When you decode information visually, [00:35:00] it is processed in a very different way and interpreted in a very [00:35:05] different way than the, the chatty language that you do.
[00:35:07] Andrew: And it's more information dense [00:35:10] literally, so you can actually have more information per compute [00:35:15] cycle, but you can also communicate more abstractly
[00:35:18] Andrew: with, with the model as well [00:35:20] than, than using language. You get different variations of how well that c- the [00:35:25] abstraction can work. But yeah.
[00:35:28] Andrew: And, what was I gonna make? I was gonna make [00:35:30] the point that, prompt engineering has basically now been reduced to [00:35:35] task, and then how do you know success? Which I would
[00:35:39] Andrew: argue [00:35:40] is outcome-driven development or just outcome-driven anything in general. [00:35:45] And a lot of people are still catching up to prompt engineering.
[00:35:48] Andrew: It's dead as far as I can tell. There's [00:35:50] very small instances where it matters, and it can be helpful, [00:35:55] but you're far... With the newer models like, literally latest gen [00:36:00] and maybe previous gen, you don't... Like I, I... You [00:36:05] might get a percent more adherence. Like it's very, very minimal. And I think you [00:36:10] could just be faster by just getting really good at [00:36:15] outcome-driven, this is what we're trying to accomplish, if there's relevant context that you [00:36:20] need, and that's where that player-coach really comes in, is you still are not outsourcing your judgment to the [00:36:25] model, but you're using your judgment to try to filter in the right level of [00:36:30] abstraction and the right level of content to feed it, to give it content.
[00:36:34] Andrew: And a [00:36:35] lot of people say just let it read everything." there's only so much context you can get,
[00:36:39] Andrew: and the more [00:36:40] context you try to cram into a prompt or a project or a session, the worse [00:36:45] results you can get. Sometimes it's better, sometimes it's worse. Most of the time it's worse. [00:36:50] And to really, I think, unlock the power of the model, you have to focus [00:36:55] it in the right way on the right problem with the right context, and then you [00:37:00] get this like massive, productivity gain because it, it'll just nail [00:37:05] it.
[00:37:05] Andrew: And it, and, and it'll reason better too, with it what it... And it's a [00:37:10] little bit, they're a little bit better now about knowing what they don't know, and when they are assuming, they let you [00:37:15] know. yeah, yeah. So, that's the kind of the player-coach. And I guess I just, I broke it [00:37:20] down into key points too.
[00:37:21] Andrew: So like player-coach mandate, the illusion of competence. So you hit on [00:37:25] that where if you're in a high-level domain the models probably don't [00:37:30] know as much as you and they definitely don't have a freshness factor of data is a big problem. Like there's [00:37:35] a training lag. And so I see a lot of really smart people that I know [00:37:40] like poo poo or anti-AI because it's not as competent as them about a subject.
[00:37:44] Andrew: [00:37:45] Okay? But it is competent in a lot of other ways, or it can help you [00:37:50] get to things faster,
[00:37:52] Andrew: Right.
[00:37:52] Andrew: And if then it's, then it's your creativity on how [00:37:55] you're using it to augment you, and then again, that's the player coach. Whereas if you think you're the [00:38:00] best academically-minded person in your thing, cool.
[00:38:03] Andrew: But if you can't work with other [00:38:05] people, then you're not gonna do really well, and you work with other people and you collaborate, so then use AI [00:38:10] as a collaborator. And you have to figure out for you and your, edge of [00:38:15] humanity's knowledge what that looks like, and that's gonna be different for everyone in every [00:38:20] capacity and different on tasks.
[00:38:22] Andrew: Most people aren't hitting this. Most people, the model's [00:38:25] smarter than you are about whatever you're trying to do. It's better at me than code, 100%, like [00:38:30] undeniable,
[00:38:31] Andrew: right? Yeah.
[00:38:32] Nathan: and I trained for it,
[00:38:33] Andrew: Yeah.
[00:38:34] Nathan: so
[00:38:34] Andrew: You [00:38:35] have to shift to abstraction. So a lot of people are task-minded or task-focused, [00:38:40] and sometimes that's okay with AI.
[00:38:42] Andrew: You, you know you need a specific thing done. But [00:38:45] I, I see a lot of people fall into the trap of they they like have to manage it. [00:38:50] And
[00:38:50] Nathan: e
[00:38:51] Andrew: yeah, it's just too much of a micromanage. And, and you'll get [00:38:55] really interesting stuff, and it'll work really well for a lot of things. But then your cognitive [00:39:00] load and your, your ability to exercise [00:39:05] magnitudes more capacity is, is hampered by your own [00:39:10] doing.
[00:39:11] Andrew: And it's hard, I think, for people to understand that because it, it's, [00:39:15] it's an abstraction problem of you have to be able to abstract the task enough. And this is where [00:39:20] transitioning from an individual contributor who, "I'm gonna do the work. I know what I need to do," [00:39:25] to, "I now have a team of people, and maybe I'm gonna be doing some of the work with them.
[00:39:29] Andrew: Maybe [00:39:30] I'm gonna be doing nothing, but I need to relay enough context to them so that they know what we're trying to [00:39:35] accomplish. And maybe I have a solution, maybe I don't. And then we need to figure out [00:39:40] how to make it happen in too short a time." And, and that's where a lot of really high-level [00:39:45] individual contributors are seeing, I, I, in the the post I, I talk about [00:39:50] 10X and h- and 1,000X.
[00:39:52] Andrew: And these, these,
[00:39:53] Andrew: these...
[00:39:53] Nathan: or the, yeah,
[00:39:54] Andrew: Yeah, [00:39:55] and these measurements
[00:39:55] Andrew: are... Yeah, a- a- and it they're just illustrative [00:40:00] numbers in that the point is that it's an unfathomable [00:40:05] amount of, of ability that you can lever
[00:40:09] Nathan: It's hard to
[00:40:09] Nathan: measure too, [00:40:10] right?
[00:40:10] Andrew: if you're doing it right.
[00:40:11] Andrew: And, and coding is always hard, right? It's lines of... If we're going, when a lot of people are [00:40:15] going back to lines of code is work done, and it's just no, it's not, that's not where we need [00:40:20] to go. But it's this abstraction problem. And then the other one was, like, organizational friction. [00:40:25] And traditional management structure is really designed to be top-down, and then we pretend [00:40:30] that it's bottom-up, and then sometimes you have mechanisms for, working through [00:40:35] meet in the middle kind of
[00:40:35] Nathan: Yep
[00:40:37] Andrew: it's really more around buy-in than anything. But it, at the end [00:40:40] of the day, it, it's like that's just not how the AI works. You need to be in control [00:40:45] still, and you are responsible for ultimately the product that you're building. But like every other [00:40:50] team that I've ever managed, it's yeah, but, the workers have autonomy and, [00:40:55] and then and a big thing I see organizations doing poorly right [00:41:00] now is that they're holding themselves back from these productivity gains because they're trying to shim [00:41:05] AI as a tool into a really inefficient and ineffective process.
[00:41:09] Andrew: And that's where you see a [00:41:10] lot of the middle management getting laid off
[00:41:12] Nathan: Yeah, yeah. I've seen the opposites too, where when you [00:41:15] give people too much agency because they have the tools now you have to rein, rein everything [00:41:20] back. So it's
[00:41:20] Nathan: like this weird TikTok thing where, where, uh
[00:41:23] Andrew: Well, no, and, and, and that's just, but that's [00:41:25] the judgment and then that's the player coach, right? Where, I just commented [00:41:30] on a thing today, actually no I think I commented on a thing yesterday, where [00:41:35] just 'cause you can doesn't mean you should.
[00:41:37] Nathan: Yeah. Exactly. Yeah,
[00:41:38] Andrew: And that's what I was talking about earlier with [00:41:40] the iCloud thing is yeah, I could totally dispatch an agent and it would find some crazy way to do a [00:41:45] sync over iCloud and it would like probably work-ish.
[00:41:48] Andrew: But we're not going to do [00:41:50] that And,
[00:41:50] Nathan: think that's one of the challenges with people working with AI especially, lots of engineers [00:41:55] who feel like, "Oh, because it's technically possible, we should just
[00:41:57] Andrew: Yeah
[00:41:57] Nathan: and build it." But [00:42:00] there's a priority or a hierarchy that you should be thinking about [00:42:05] features when you have AI that can go and do it, a lot of people's first instinct is, "Okay, let's just go do it."
[00:42:09] Andrew: [00:42:10] Exactly.
[00:42:11] Andrew: And, and and it's funny, we talked a little bit about this, Ingolst, with the [00:42:15] blog, but like I could not find a blogging platform that
[00:42:17] Andrew: was minimal enough, Right?[00:42:20]
[00:42:20] Andrew: So I built one.
[00:42:21] Andrew: And I, I see the irony in that, [00:42:25] but I think I also built it in a way... I'm trying to think [00:42:30] a little quickly about how it actually works.
[00:42:33] Andrew: It's really more... [00:42:35] Excuse me. It's really more, I built kind of a middleware
[00:42:39] Nathan: Okay. [00:42:40] Sure.
[00:42:40] Andrew: Right, so the actual build piece, right? So it's like I'm using Astro for all [00:42:45] the HTML. It's a static site generator, and the problem I had with [00:42:50] Astro is that if I wanna make a post, I have to format it and program [00:42:55] it into a page or something. I don't wanna do that. Sure, I can have AI do it, but then...
[00:42:59] Andrew: yeah
[00:42:59] Andrew: [00:43:00] but then every time I do it, and then you have to compile it technically, right? And then you push the file to the web server, and there's [00:43:05] processes to mediate this, and like that's hard. It's not hard, but it's, it's... That's not,
[00:43:09] Andrew: that's not...
[00:43:09] Nathan: the [00:43:10] ass.
[00:43:11] Andrew: As, as like as a user of the blog, I wanna be able to write a [00:43:15] thing, be done with it, and then push it, which WordPress, right?
[00:43:18] Andrew: But WordPress is this heavy [00:43:20] service. You need to run it on a database, and y- there's a bunch of vulnerabilities and security, [00:43:25] and it's slow, and like you can make it work and, and anyway. And I, I run the [00:43:30] podcast website on WordPress. It's great for what it's good for, but it, it's heavy.
[00:43:34] Andrew: No,
[00:43:34] Andrew: [00:43:35] I've, I've been run WordPress for 15 years or however long it's been out.
[00:43:39] Andrew: Yeah [00:43:40] on and off, but yeah, no, the, the podcast has been there the whole time. It's wonderful, [00:43:45] but again, it's like it was way too heavy for just a
[00:43:47] Andrew: personal blog that I'm gonna
[00:43:48] Nathan: choose the right tool for the job
[00:43:49] Andrew: [00:43:50] Yeah. And then there's all this other stuff in between.
[00:43:52] Andrew: The other thing I wanted is I w- I want it to [00:43:55] be automatically shared to the, the Fediverse, which is a, a decentralized protocol [00:44:00] for social media. So it goes to that stream and people can pick it up or [00:44:05] not. It's just a way of just sharing, and also a lot of blogs will do that, and there's [00:44:10] feed managers that'll suck it in, and so a little bit for discovery and, and that kind of thing.[00:44:15]
[00:44:15] Andrew: I have RSS and Atom supported as well as, as feed subscribers. [00:44:20] Surprising how many don't support that very easily.
[00:44:24] Nathan: It feels like those [00:44:25] are dying quotes
[00:44:27] Andrew: Yeah, but the, the blog, the peop- the [00:44:30] people who are into reading personal blogs are all about feeds.
[00:44:33] Nathan: Mm-hmm
[00:44:33] Andrew: It's r- it, it's... I- and I'm [00:44:35] not one of these people. I will read them and I have ways to read them, but I don't use a [00:44:40] feed manager. So that I wanted, and then I also, I also wanted [00:44:45] it to post to social media, um, as much as I could.
[00:44:48] Andrew: There's a lot of social [00:44:50] platforms that are locked down, but of the ones that aren't locked down, I wanted... It's write once, go [00:44:55] everywhere. And so you can use things like Buffer and there's all these other third-party platform, and then I [00:45:00] have to
[00:45:00] Andrew: Write a connection and now I'm running, again, a big service.
[00:45:03] Andrew: I borrowed a [00:45:05] couple, of those services. There's like plugins and stuff. [00:45:10] And it
[00:45:10] Andrew: sounds sketchy. And I mean it, it does sound sketchy.
[00:45:12] Andrew: There, there's pe- you know, pe-
[00:45:13] Nathan: confused. I didn't know what you
[00:45:14] Nathan: meant,
[00:45:14] Andrew: [00:45:15] yeah. People have written libraries for that, right? So it... So I, I put all that in, and [00:45:20] then instead of writing my own hosting stack software, whatever, it runs in a container [00:45:25] with NGINX, which is really standard.
[00:45:27] Andrew: And then the back end is [00:45:30] NextCloud. We talked about this before. So I write the draft in my [00:45:35] NextCloud or wherever else I wanna write. I use Obsidian and then that connects into NextCloud. And [00:45:40] then when I move it to a specific folder, there's a job that runs every minute to just look at it and [00:45:45] say, "Oh,
[00:45:45] Nathan: make up everything new and then
[00:45:46] Andrew: It runs the Astro compiler pipe [00:45:50] step, whatever the hell happens, and then it spits it out on the website, and then it goes to the feeds and everything [00:45:55] else. So it's like it sounds complicated. It's not. I'll probably open source it too. It's just I haven't got [00:46:00] around to it. But it I...
[00:46:03] Andrew: But don't write your own s- [00:46:05] bespoke thing except for when you have to. It's also really easy, right? 'Cause this took [00:46:10] maybe an hour of tinkering and doing a little research to try to figure out... 'Cause there's a lot [00:46:15] of different options, and AI had some really strange ideas to how to do stuff. So this is where [00:46:20] you totally get into trouble with a Frankenstein's monster kind of problem.
[00:46:24] Andrew: [00:46:25] So yeah, I, And the monster was not named Frankenstein. That still makes me a little
[00:46:29] Andrew: mad.
[00:46:29] Nathan: [00:46:30] it's the doctor.
[00:46:30] Andrew: It's the doctor is Frankenstein, so it's, it's Frankenstein's [00:46:35] monster.
[00:46:35] Andrew: Yeah, Yeah.
[00:46:36] Andrew: I think, I think it's actually referred to as monster in the script
[00:46:39] Nathan: Sure. [00:46:40] Yeah, I don't think it's proper noun though, right? It's not capitalized M, which
[00:46:43] Andrew: No, I, I don't... Yeah, I think it's just [00:46:45] Monster. Anyway,
[00:46:46] Nathan: yeah
[00:46:47] Andrew: So but, so it's so but I responded to a, a [00:46:50] comment somewhere, and I was just like, it- I'm like, kind of like, "Ugh everyone's [00:46:55] writing everything to solve their own problem," but also everyone's able to write [00:47:00] specific solutions to problems they have, which I think is awesome.
[00:47:03] Andrew: But again, that's a judgment [00:47:05] thing of you have to know when to say no. You have to know when something might be good [00:47:10] enough you can, you can use. And I tried Gycle, I tried Astro Native. There's some weird shims and [00:47:15] stuff you can do. I also didn't really want it like an online [00:47:20] interface with like spam and like hacking and stuff.
[00:47:23] Nathan: Okay
[00:47:23] Andrew: to have a web [00:47:25] interface for this, like I'm getting
[00:47:27] Andrew: away from like having headless
[00:47:29] Andrew: [00:47:30] everywhere.
[00:47:30] Andrew: yeah,
[00:47:30] Nathan: Okay. Yeah, yeah, like headless CMSs where yeah, you don't have
[00:47:33] Andrew: Yeah.
[00:47:34] Nathan: for it, you [00:47:35] just
[00:47:35] Andrew: Yeah, a-a-and I wanted it so that it was, like, editor agnostic because [00:47:40] it, it has to be in Markdown. That was my one thing is Markdown is the way now.
[00:47:44] Nathan: Mark down all the way. [00:47:45] Yeah
[00:47:45] Andrew: but...
[00:47:46] Andrew: And there's a lot of Markdown editors, but I was just like, I, I just, [00:47:50] I didn't wanna have to write a stupid editor or borrow an [00:47:55] editor and then shim it in and then have to maintain it, and then there's...
[00:47:58] Nathan: those are very complicated, by the
[00:47:59] Andrew: [00:48:00] No kidding.
[00:48:01] Nathan: Yeah, what is what you get
[00:48:02] Andrew: A-Andrew did a game and has learned his [00:48:05] lesson. Andrew will not do an open source or any sort of file editor even though, my [00:48:10] God, I have thought about it a lot.
[00:48:13] Nathan: Many people have dry, dry[00:48:15]
[00:48:15] Andrew: yeah.
[00:48:15] Nathan: yeah.
[00:48:16] Andrew: No, it's, that's really hard. So anyway, so yeah, and maybe [00:48:20] I'll, I'll push it out, but I thought that was, like, a, an exact specific problem of... And if there's a [00:48:25] better solution that will
[00:48:27] Nathan: Yeah, let
[00:48:27] Andrew: take a mark- a Markdown file [00:48:30] easily and then push it out to social media and the Fediverse and publish it to a website [00:48:35] and run on 150 megs of RAM in a container, let me know
[00:48:39] Andrew: 'cause... [00:48:40] And I'll, like I said, I'll,
[00:48:41] Andrew: I should, I should open source it. I, I don't think s- go make money on it. [00:48:45] I don't really care. And the other thing was, like, I didn't wanna host it on, A lot of [00:48:50] people are asking me why not Substack. I have a project that's on Substack, which I might just [00:48:55] kill.
[00:48:55] Andrew: I haven't decided yet. I think it's more important than ever to [00:49:00] own the means of production much as you can. And so your data, your [00:49:05] content, however you want, like what you give up on that is discovery. So if I were to have a, a Substack, [00:49:10] theoretically I would get more coverage. However, then they can [00:49:15] editorialize.
[00:49:15] Andrew: They say they won't, and then everybody always changes. They say that their algorithm is fair [00:49:20] until it's not. They say that whatever. And [00:49:25] we're on YouTube right now, but I do that for distribution, [00:49:30] and discovery is frankly not very good. We don't have nearly the subscribers we [00:49:35] do versus audio only, which I think is fascinating.
[00:49:39] Andrew: The numbers aren't [00:49:40] even comparable
[00:49:40] Nathan: Audio-- Did audio only start first for... Did it have a head
[00:49:43] Andrew: [00:49:45] Yes.
[00:49:45] Andrew: Yeah.
[00:49:46] Andrew: So
[00:49:46] Nathan: good to look at, so
[00:49:48] Nathan: it's know, maybe. I, I [00:49:50] and I think for f-
[00:49:50] Andrew: I think for podcasting, like it's [00:49:55] frustrating
[00:49:55] Nathan: mode is, is probably
[00:49:56] Andrew: Yeah, and it's decentralized, right? You do have [00:50:00] Apple and Spotify who are trying to collapse the market down for their own [00:50:05] purposes, but it's a, a decentralized federalized service that anyone [00:50:10] can set up an RSS stream and then your podcast player can download.
[00:50:13] Andrew: If you wanna get [00:50:15] discovered there's feeds that... Or there's, podcast hosts that you can submit [00:50:20] to, kind of run that central. But there's other non federal or federal- [00:50:25] federalized, federated, sorry, not federalized, federated, services that [00:50:30] you can do disc- that'll do discovery and just pull it- pull you in.
[00:50:33] Andrew: There's a lot of, there's a [00:50:35] lot of services that 38's on that I was just kinda
[00:50:37] Andrew: "Whoa, I didn't... I had no idea this world [00:50:40] existed,"
[00:50:40] Andrew: and it's cool. But I, I had no idea. I was like, "That's
[00:50:44] Nathan: Yeah. I [00:50:45] mean the internet's vast, right?
[00:50:46] Andrew: freely available and the, the [00:50:50] protocol's not locked down. And so I, I wanna...
[00:50:52] Andrew: I think we need to keep it that way. And so for a blog or [00:50:55] something, it's never been easier and cheaper to host your own thing. I can say whatever I want [00:51:00] and no one's gonna shut me down. Whereas there's constant editorializing going, [00:51:05] and if they don't shut you down, they don't promote you or they make it harder to find [00:51:10] you.
[00:51:10] Andrew: Shadowbanning, right?
[00:51:12] Andrew: And I'm not saying anything crazy, at least I [00:51:15] haven't yet. Maybe I will. I don't know. It's me.
[00:51:18] Nathan: I was gonna say maybe not [00:51:20] intentionally.
[00:51:21] Andrew: Yeah. No, and I, I'm not trying to be inflammatory on the [00:51:25] blog. I, I guess the other thing with the blog too is I have essays, which are an essay, [00:51:30] but then I also had this idea of like incomplete thoughts and thinking in public
[00:51:34] Nathan: Yes, [00:51:35] I have, I have a personal repository of incomplete thoughts, of
[00:51:38] Andrew: Yeah.
[00:51:39] Nathan: dr- [00:51:40] drop notes on the internet. I'll
[00:51:41] Andrew: Yeah. Yeah. it like I say fragments or something, [00:51:45] and I, I think it's important 'cause that was one thing I liked about Twitter back in the day is people would just post random [00:51:50] shit that they were like, "Ah, I was walking to work and I had this idea," or " What if this was like this?"
[00:51:54] Andrew: And [00:51:55] then a little bit of context. And so I wanted to keep that, and I would love to see that brought [00:52:00] back more where people are just like having kind of idea fragments
[00:52:04] Andrew: And sharing that [00:52:05] context to folks instead of just all the social media flame bait crap that we have [00:52:10] now.
[00:52:10] Andrew: So anyway. I know. Yeah, it, it's an interesting experiment. I don't plan to post [00:52:15] that often. There's a... I have a lot of stuff that I've been thinking about for years that I [00:52:20] gotta clear out of the old head, but, I don't expect to get an [00:52:25] audience. I don't care about subscribers.
[00:52:27] Andrew: If you wanna subscribe, great. There's RSS feeds. If there's [00:52:30] other ways you wanna subscribe or think I should support, let me know. I'm only familiar with [00:52:35] Fediverse or federated kind of stuff, and then Atom and RSS. [00:52:40] There's a million and one different little feed systems, but those are the ones I'm supporting.
[00:52:44] Andrew: I'm not gonna [00:52:45] do every single flavor of the day
[00:52:47] Andrew: FeedBurner kind of thing. It's just [00:52:50] a RSS feed and then it's XML feed, so it's it's really standard
[00:52:54] Nathan: Yeah, basic [00:52:55] technology, right? That, that has been work-
[00:52:56] Andrew: Yeah.
[00:52:57] Andrew: No, and,
[00:52:58] Andrew: and, and, and and a lot [00:53:00] of the platforms have made stuff worse
[00:53:02] Nathan: Oh yeah, that happens.
[00:53:04] Andrew: And, and we've talked [00:53:05] about this a little bit too, but and, and I don't really have anything against Substack, but I do know that there [00:53:10] have been people that have written things that are, politically [00:53:15] unfavorable or, the popular opinion doesn't support, and they've been [00:53:20] either outright cut off the platform or shadow banned.
[00:53:24] Andrew: And I'm not talking about [00:53:25] extremist, like crazy stuff. It's just stuff that, that people are writing that is not the [00:53:30] populist opinion. It's Anti-populist or it's fringy, whatever you wanna [00:53:35] call it. And I don't have a problem with that. Like that, I, I'm a free speech... I live in [00:53:40] America. We have the First Amendment.
[00:53:42] Andrew: You're supposed to be able to have ideas and not [00:53:45] necessarily be punished for them. I don't like hate speech. I don't like extremist [00:53:50] rhetoric or speech or whatever you wanna say, but the... It is an absolutist thing. Yes, [00:53:55] we have some laws on you can't threaten people, that kind of stuff.
[00:53:57] Andrew: But it's like, when you start, when you [00:54:00] start giving people the ability to, to, to keep people from saying or doing or [00:54:05] thinking things that you don't like, the opposite can happen. [00:54:10] 'Cause there's a group of people out there who are opposite of you, who have different ideas about what you can [00:54:15] or, or can't say, or what you should or shouldn't say.
[00:54:17] Andrew: And it's a slippery slope argument, but the one thing that [00:54:20] makes America, like pretty much unique, I don't... I think there's only one other country that has [00:54:25] anywhere close to a free speech law that like we do, is it's free for [00:54:30] all, for better or worse. And, and with... We've talked a lot about social media [00:54:35] craziness, and it's like that's the platform incentivizing the speech, [00:54:40] right?
[00:54:40] Andrew: There's always crazy people. There's always fringy people. There's always polarizing [00:54:45] people. There's always people who are saying shit you don't wanna hear or don't like. The social media [00:54:50] companies then show you that because they know you're gonna either share it with someone else or complain about it, which [00:54:55] creates engagement.
[00:54:56] Nathan: Hmm
[00:54:57] Andrew: I, so I think people have a really poor impression [00:55:00] of free speech when you're online all the time because you're literally [00:55:05] confronted with the stuff that you probably don't want. But that's not how most life is, right? [00:55:10] That's just not how it is. So anyways, so like that, I, that's really important, and I, I...
[00:55:14] Andrew: With [00:55:15] AI and, and auto, in a way of robot comment banning and shit now, and like people are mad at [00:55:20] Reddit and, and it's we need a way to keep people [00:55:25] accountable online, and I think the easiest way is to just map your identity. I believe you should [00:55:30] be able to be anonymous online, but I also then don't think that that necessarily needs to [00:55:35] translate to everybody in the world hears you now
[00:55:38] Nathan: Yeah. Oh [00:55:40] like in, in the sense of you can't or you shouldn't... Th- there's this idea of the, the cloak of [00:55:45] anonymity, right?
[00:55:45] Nathan: You are more
[00:55:46] Nathan: likely to say or do things. Yeah, yeah. Right, right. [00:55:50] Yeah. accountability.
[00:55:51] Nathan: right?
[00:55:51] Andrew: But I fully... I don't think you should have to [00:55:55] be under your identity on the internet. I'll be very clear.
[00:55:58] Nathan: sure. Yeah, yeah
[00:55:59] Andrew: you [00:56:00] wanna be anonymous on the internet, all for it. Absolutely. However, then I don't think you [00:56:05] should necessarily get blasted through social media and have millions of people view you [00:56:10] either.
[00:56:10] Andrew: And I, I'm not saying shadow ban either, but I'm saying I don't think you're entitled to your [00:56:15] opinion being heard. Now, if you have your name behind, like my blog is my name. It's pretty obvious who the [00:56:20] I am.
[00:56:21] Andrew: care. But I think more people need to do that. This is me online. I'm [00:56:25] just writing some shit down.
[00:56:25] Andrew: I'm sharing ideas. I don't have comments technically. It's [00:56:30] Fediverse, so you can like comment back. I guess that's one thing maybe I should add is if people [00:56:35] do feed back, then I should display it, I think is how that works.
[00:56:39] Andrew: [00:56:40] yeah,
[00:56:40] Nathan: I think w-- I think Twitter's one of ear-- Twitter's early policies was [00:56:45] something along the lines of like freedom of speech is not the same as freedom of reach or [00:56:50] something like that, where, Yeah, yeah, so I think what you're mentioning is,
[00:56:54] Andrew: [00:56:55] Yeah,
[00:56:55] Nathan: directly align, but it's very
[00:56:56] Andrew: it and, and there's a lot of bots online, like 20% or more of the internet is [00:57:00] bots, so it's like the people you think are real are not real either. You're just arguing with technology. [00:57:05] And, and we gotta get
[00:57:06] Andrew: rid of that shit too, but I, I know.
[00:57:08] Andrew: But no, but, but I, I do [00:57:10] think and this is not a new idea, right?
[00:57:12] Andrew: But a lot of people have thought about what does it look like to [00:57:15] verify your identity on the internet
[00:57:17] Nathan: Yeah. Oh, yeah.
[00:57:18] Andrew: I just don't want it required, [00:57:20] right? 'Cause again, I, I, there's all sorts of places in the world you don't have the protections I do as [00:57:25] an American citizen in America,
[00:57:27] Nathan: Yeah, yeah,
[00:57:27] Andrew: right? And, and I want [00:57:30] everybody to be able to interact with the world, and there's a lot of places in the world right now you can't speak freely online.
[00:57:34] Andrew: Either [00:57:35] you literally can't or if you do, someone comes to your house and you don't get, you're not seen [00:57:40] from again.
[00:57:40] Nathan: Yeah, yeah, yeah. Yeah, there are
[00:57:41] Andrew: But
[00:57:42] Nathan: ideas around civilizations that could be built
[00:57:44] Andrew: [00:57:45] Yeah. And I'll probably write, I'm, I'll probably write about this 'cause I, I think it's an important thing, especially with AI and, [00:57:50] and them sucking up all your data and selling it and, if you wanna monetize my...
[00:57:54] Andrew: I guess I, [00:57:55] the AI probably crawls my blog. I don't really care.
[00:57:58] Nathan: it, you-- Yeah, I was gonna say you could, you [00:58:00] could prevent
[00:58:00] Andrew: Yeah, but
[00:58:01] Andrew: I it's public, right? A, I, I don't want [00:58:05] you taking credit for my work or my, my [00:58:10] ideas and plagiarizing. But if you wanna use my ideas and my work to build [00:58:15] and make something better, you then have to contribute it back, right?
[00:58:17] Andrew: All the so- all the open source software is under [00:58:20] GPL3.
[00:58:20] Nathan: Licenses,
[00:58:21] Andrew: has opinions on licenses. The one reason we chose that [00:58:25] method was it forces you to put the contributions back in public. And [00:58:30] I, I, I'm, that's I'm a, I love FreeBSD. I really [00:58:35] truly believe in that model. However, I think just with the way economics of software has gone,[00:58:40]
[00:58:40] Andrew: I wanna ha- I wanna have to force people, if you're gonna use my intellectual [00:58:45] property, then you need to contribute your intellectual property back to the rest of the world so we can all [00:58:50] benefit.
[00:58:50] Andrew: Which people have problems with and are uncomfortable with, but then don't use my [00:58:55] intellectual property, and you don't have to abide by the rule. But th-this, this podcast [00:59:00] is Creative Commons licensed, right? Just don't claim that you had the idea first or you said the [00:59:05] thing, and build on it or tear it down.
[00:59:06] Andrew: I don't really care. The blog isn't technically licensed. I [00:59:10] d- thinking about I should throw the, a Creative
[00:59:13] Nathan: Before you open source it.
[00:59:14] Nathan: Yeah. [00:59:15]
[00:59:15] Andrew: No j- just the content. I just wanna be clear with folks.
[00:59:18] Nathan: oh, oh, oh. Oh, the [00:59:20] content is added. Got it.
[00:59:20] Nathan: Okay
[00:59:21] Andrew: It, it... And then that gives people the ability if you wanna cite it [00:59:25] and, and build on it, great.
[00:59:26] Andrew: I just attribute, right? Creative Commons attribute is the [00:59:30] basic license. They have problems with it, but it the intention for me is that if you're gonna use [00:59:35] my work, you cite it, and you build on it or tear it down, but you don't get to claim that it was [00:59:40] yours ' cause my ideas aren't unique, right?
[00:59:42] Andrew: No o- there's no such thing
[00:59:43] Andrew: as a unique [00:59:45] idea or perspective. No, th- there never r- I don't think there ever really was. Like, you're [00:59:50] always building on someone else's stuff,
[00:59:52] Andrew: right?
[00:59:53] Nathan: everything's a remix.
[00:59:54] Andrew: Yeah, and l- [00:59:55] language is inherently that, so unless you come up with your own language and you're doing your own thing, then it's maybe [01:00:00] unique.
[01:00:00] Andrew: But anyway. We did it. No script show
[01:00:04] Nathan: I think the funny [01:00:05] thing was before we went into it, we're like, "Okay, let's not talk about AI.
[01:00:08] Andrew: I know, right?
[01:00:09] Nathan: [01:00:10] But I think, I think it ended on a very interesting note actually, to be honest. There, there's some very interesting topics there. [01:00:15] Coming into it, we were like, "We have no idea what to talk about."
[01:00:17] Nathan: And then yeah,
[01:00:18] Andrew: yeah, it, it,
[01:00:18] Andrew: maybe next week,[01:00:20]
[01:00:20] Andrew: maybe next week, if I, see if I have another blog. But I can maybe we could [01:00:25] to talk about like data and like digital freedom and those kind of [01:00:30] ideas 'cause it, it,
[01:00:30] Andrew: I... In the AI debate I'm not hearing... [01:00:35] I'm hearing like pieces of that, but the... And I guess maybe for folks to [01:00:40] mull on at home over the week is the AI companies have effectively [01:00:45] trained their models on, this is Anthropic and [01:00:50] OpenAI specifically, Google and everyone else is culpable too, but they've trained their models on the [01:00:55] entirety of human knowledge that's digital and then some not.
[01:00:58] Andrew: And then they're like [01:01:00] renting what they learn back to us.
[01:01:01] Nathan: yeah. Yeah.
[01:01:03] Nathan: Effectively
[01:01:03] Andrew: bullshit because they're [01:01:05] not paying people for the knowledge that they used to [01:01:10] create that intention. They did it at a scale that isn't human doable, right? If you [01:01:15] read all the works of Shakespeare and then come up with a play, everyone is kind of like, "Yeah, you can [01:01:20] do that, that, and then build on it," and then you gotta give them credit but like [01:01:25] sure.
[01:01:25] Andrew: But to, to literally soak up the entirety of human knowledge and then rent it [01:01:30] back to us or rent the derivative learning back to [01:01:35] us, I think is like abhorrent.
[01:01:38] Nathan: Okay
[01:01:38] Andrew: And then to [01:01:40] hear all about this like empathy and we have to do the right thing and for, for society, but you're [01:01:45] gonna charge me? I don't expect you to be a charity, but like that's [01:01:50] really
[01:01:50] Andrew: Up.
[01:01:50] Nathan: discussion
[01:01:52] Nathan: In
[01:01:52] Andrew: Yeah.
[01:01:53] Nathan: where you have people [01:01:55] who are contributing the data to train on models, but the owner [01:02:00] is the people who ended up providing the data to train on it
[01:02:02] Andrew: Yeah. Well
[01:02:03] Nathan: something that people are talking [01:02:05] about
[01:02:05] Andrew: and I know they've trained on the blog. I don't have it turned off, right? And I, I know I think the [01:02:10] podcast, we ha- it's, they, they definitely do. I see them. [01:02:15] And I'm okay with it because I'm okay with the free exchange of [01:02:20] information. I'm not okay with what they're doing in the derivative, but I'm not gonna shut it [01:02:25] down.
[01:02:25] Andrew: Because I think access... I think it just has to be free and open, and this is one of the [01:02:30] ramifications of free- just like free speech. People say stuff I disagree with all the time and, and I don't know if [01:02:35] I really get upset, but I don't like. That's part
[01:02:37] Nathan: I, sometimes I can't tell.
[01:02:39] Andrew: right?[01:02:40]
[01:02:40] Nathan: I'm joking. Yeah.
[01:02:41] Andrew: But that it's just i- in order for me to, to be able to [01:02:45] do that, that's just part of it, and then you have to also give the privilege to other people.
[01:02:49] Andrew: [01:02:50] I guess that's a right, but and, and But I also am very increasingly [01:02:55] uncomfortable with the way that this is going and, the of AI
[01:02:59] Nathan: [01:03:00] Yeah, yeah. Yeah, for sure
[01:03:00] Andrew: and more... We've talked a little bit about this offline. So I think
[01:03:03] Nathan: Yeah
[01:03:04] Andrew: interesting to maybe [01:03:05] talk through that a little bit and maybe I'll have some, some thoughts for the blog [01:03:10] that
[01:03:10] Andrew: if...
[01:03:10] Andrew: I, yeah. A- and I get...
[01:03:13] Andrew: Yeah, and I didn't really answer why [01:03:15] blog, but there's a lot of things that if you don't try to write it down, you don't [01:03:20] really understand either how far off you are or maybe how brilliant your idea is. [01:03:25] The process of writing and thinking about what you're writing
[01:03:28] Andrew: is is really...
[01:03:29] Andrew: And not just [01:03:30] typing like a quick text or something. You
[01:03:32] Andrew: know, it, it's...
[01:03:33] Nathan: like flushing it
[01:03:34] Nathan: out. Yeah, and this is [01:03:35] why I don't like using like voice prompts when, when people are talking about AI, like they
[01:03:38] Nathan: talk to their AI and you get... [01:03:40] No, t- the act of typing a- forces you to think about what it is you're
[01:03:43] Andrew: Yeah.
[01:03:43] Nathan: done. So
[01:03:44] Andrew: Yep, [01:03:45] no, 100%. So, if you don't, you don't have to write a blog, but write a journal. Like we've had Michelle on in the [01:03:50] past and she's talked a lot about journaling. It's been a while now. We've taken a [01:03:55] little bit of a break from interviews. I just don't have time to, to be honest.
[01:03:59] Andrew: [01:04:00] Um,
[01:04:00] Nathan: too hard.
[01:04:02] Nathan: Honestly, setting it all up, I, I thi-
[01:04:03] Nathan: it's one of those things where it [01:04:05] sounds simple, but there's a lot of work that goes into it, right?
[01:04:07] Andrew: Yeah. Well, and [01:04:10] I, I, don't prep in a sense that I don't write questions. It's, it's very much a- [01:04:15] an organic conversation, but I do do background on people or if they have a project they're [01:04:20] working on or something, I, I wanna be familiar with it so I can at least like talk intelligently or ask [01:04:25] intelligent questions as we're going.
[01:04:27] Andrew: But yeah, we ha- we have one coming up that will be out pretty [01:04:30] soon that is a little different than what we normally do. But
[01:04:34] Andrew: I think it's gonna [01:04:35] be... It's a, it's a really, really cool episode in a lot of ways. It's a, it's [01:04:40] an interesting episode in a lot of ways, too. It's a guest that's been on before, so I'll tease it as [01:04:45] that.
[01:04:45] Nathan: I'm looking forward to it
[01:04:47] Andrew: Second occurrence, I think. Maybe third. Second, I think. [01:04:50] Yeah.
[01:04:51] Nathan: I think second, yeah
[01:04:52] Andrew: I think, yeah. So anyway. Yeah, Nathan knows. You can get it out of [01:04:55] him. I'll never s- I'll never tell
[01:04:57] Nathan: Yeah, come at me. No, in the Discord. No, I'm [01:05:00] just kidding.
[01:05:01] Andrew: Yeah, keep plugging the Discord. There's,
[01:05:03] Nathan: it...
[01:05:04] Andrew: you
[01:05:04] Nathan: I'm just
[01:05:04] Andrew: [01:05:05] it...
[01:05:05] Nathan: the lines. Follow the orders
[01:05:06] Andrew: Yeah.
[01:05:07] Nathan: me
[01:05:07] Andrew: I know, right? Yeah, no. With the Discord, again, it's a [01:05:10] way for people to collaborate. It's not super active. It's not very busy. It's, it's a critical mass thing. You need a [01:05:15] bunch of people in there doing stuff, so join it and eventually we'll get, we'll get there.
[01:05:19] Andrew: It's just a slow [01:05:20] build. And I frankly don't promote it, and I should. I, I don't... It's not I don't, I don't do it [01:05:25] well. I need to get better at it
[01:05:26] Nathan: Yeah, yeah.
[01:05:27] Andrew: Cool. All right.
[01:05:28] Nathan: Now
[01:05:29] Andrew: Happy Thursday. [01:05:30] Yeah.
[01:05:31] Nathan: All right. Have a nice weekend,
[01:05:32] Andrew: All right, yeah. Yeah, take care. Bye