Taylor MacDonald: Well well well. Here we are again. ⁓ what is this? Num number six? Number Yeah, this is six. Episode six. No. No, it's not. No, this is seven. This is episode seven. We did six last week. Well, I was doing a bunch of post production this morning, so I had the number six on my mind. Anyway. Is it six or is it seven? Wait, is it six se wait, ⁓ my god. Did you see what I just did? You don't have a teenager. ⁓ Sean C Davis: Hello, Taylor. Is it s is it six? Six? Okay. Is it seven? ⁓ okay, that's what I was thinking. Okay. Seven. Yeah, lucky seven. I did. I do well, no, I don't, but ⁓ it's ⁓ Taylor MacDonald: I'm gonna have to edit this part out post-production. If my son hears this, I'll never live down. All right. How you doing, man? Today is is it, June 11th? ⁓ we're recording this in the morning of June eleventh. This is also the last day of Cincy AI Week. Sean, you spoke at Cincy AI Week ⁓ on Tuesday. Maybe ⁓ do you what'd you think of the conference? How's the conference going so far? Or did you get out there yesterday? ⁓ what's your general what is maybe we start with what is Cincy AI Week? We are also located in Cincinnati, if that wasn't obvious. Sean C Davis: Yeah, we are. It's it is different than I originally expected. I thought it was gonna be kind of like a community sort of thing, but I I've been thinking about this this week and I think it's you know, Cincinnati not really being a a startup sort of hub. It feels like we're more of a Wha what would you say? More of like a marketing and engineering sort of hub. So it's it's definitely I should have seen it coming, but it's way more corporate and less ⁓ less community, which I think can be really helpful for folks in in marketing and, you know, various types of engineering. But yeah, I'd I I wasn't there yesterday. I was down there for a while on Tuesday, kinda in and out, and it felt It felt very sleepy despite being sold out. W what what did you notice? Taylor MacDonald: I think in general okay, so let me say I I agree with you that it's definitely I feel like the corporate overlords have hijacked a lot of the momentum and energy behind this. and and right and I mean I appreciate that. Like there's a lot of good thought leadership going on in that space, particularly here. I do know that ⁓ so I've been going to the Centsi AI f human AI for humans meetup now for probably the last six months or so. And I feel that that ⁓ group Sean C Davis: Yeah, yeah. Taylor MacDonald: So basically it's a a once a month meetup that is hosted up on UC's campus. It's ⁓ organized by Kendra Ramirez and Helen Todd. ⁓ and it has some relationship or backing, perhaps sponsorship or just partnership with a group called the Cinsi AI Consortium, ⁓ which is run by a guy named Carl Freak. And ⁓ Carl has done a fantastic job of like outlining what they call the blueprint to put Cincinnati on the map. in terms of ⁓ I forget there's some ⁓ there's some standards body or some like research institution that is like ranking different cities based on you know a number of different economic factors that would ⁓ position them as like the next gen tech hub, particularly in the space of AI. And so the AI consortium's goal is to get Cincinnati on that map. And part of Cincy AI Week is wrapped up into that thing. So I I'm guessing that the corporate sponsorship and the influence from the corporate partners has a lot to do with the momentum required in order to satiate those expectations to put you on the Sean C Davis: ⁓ this makes sense. Okay. Taylor MacDonald: That said, I agree. It doesn't it feels it I my honest assessment of this event was that it was disor that it has been disorganized at least. I only went on Tuesday though, to be clear. So if if we're being fair, I have fifty percent of the time. I did go last year and it felt a little bit tighter last year, but I think this year it got more complicated because they spread it across multiple venues. So there's three different venues and ⁓ OTR, which is kind of like north of the central business district district in in ⁓ Cincinnati. Sean C Davis: Mm. Taylor MacDonald: is an extremely cool part of the city. The architecture's amazing. There's tons of rich history. heavily gentrified. So that sucks ⁓ for lots of reasons. But Cincinnati's a weird place. Anyway, it's it's walkable and it takes you kind of through some of the more signific like like one specific really cool part of OTR, which is Washington Park. And directly like down the block from one of the venues is is Cincinnati's Historic Music Hall, which is like one of the most, you know, venerable architectural masterpieces in the city. And so I think if you were coming from outside of the C city, there's a lot to like the the multi-venue, the walkable nature of the neighborhood there, a lot of good food options. ⁓ I think that was great. I think some of the Topics I thought were I don't I don't know who they were targeting, like particularly your panel that you spoke on, which I didn't think was supposed to be a panel when you went in. I thought was interesting. but you know, that said, like yeah, it just felt it it's felt disorganized. Like I went to get my when I sh first got there on Tuesday morning. Sean C Davis: Mm-hmm. It definitely was not. Taylor MacDonald: I parked in the Mercer Garage, which is kind of like centrally located between all three of those venues. And and one of the venues, Union Hall, where it was primarily located last year, ⁓ was right across the street. So I walked in there, I was like, Where's my badge? And they're like, ⁓ yeah, you gotta go to this other venue. And I mean, it was fine. It it was what, like a maybe five, ten minute walk all in to get to the other venue. But once I got there, then I had to like search around for the badge. It just was like kind of confusing, and anyway, ⁓ So yeah, I don't know, disorganized, well intentioned. ⁓ also, like if it wasn't clear, go check out the CENSI AI for humans meetup. I that might be the community that you're looking for. ⁓ those seem like a really interesting cross section of different folks that you know, all the way from super beginner to super advanced. And and some of the things that Helen and ⁓ Kendra are doing specifically are you know, establishing different Huddle groups to lean into those different affinities, you know, and or your experience level where you're comfortable, it gives you kind of a safe space to get together with somebody remotely and just talk and ask questions and it it is as well intentioned a community effort as I've ever seen, but I'm not I'm not sure what would I mean, maybe that's a question for you. Like you've attended you you came to one of those meetups, right? Yeah. Well what would your feedback andor criticism be? Sean C Davis: I did. Yeah, yeah, yeah. Mm-hmm. Taylor MacDonald: You know, if you or or maybe the question would do you apply the same assumption or opinion or perspective on the meetup as you did on the conference itself? Sean C Davis: I think there's I I think at first I did, but I've been I've been thinking about that a lot this week. And I I I'm assuming, you know, if folks listening in are not in Cincinnati, they probably have some equivalent in the nearest city. ⁓ or some something like that. But what's really interesting about this, so I the first time I went, I left and I was like Okay, this is a nice community. I didn't learn anything. What was it for? And I think if you are going in for networking, there's a lot of value. But the more I thought about it, I realized that I think there are two opportunities, especially if folks listening in are if if you're you know, you're you're you're here to create for the web or whatever. ⁓ on one hand I most meetups tend to be kind of in our developer world and we don't really break out of that and we're talking about development and patterns. So I think on one hand There's an opportunity to get into those groups and to teach folks, because now anybody can in theory create anything. I mean, they can't necessarily create anything at scale, but the you know, that that idea of like anybody can create anything, which means that if you have any little bit of of history and experience creating something, there's an opportunity to speak and teach, and that's probably a really good front door to get into. And then on the other hand A lot of these folks I noticed going to a meetup like this are they're using AI, but they're using AI in everyday events. And it was it was actually this first this this first ⁓ AI for humans meetup that I went to where I went in and I knew about Claude Desktop and I knew about Claude Cowork and I had spun them up and played with them for 30 minutes and I was like, Yeah, this sucks. This is not for me. And and then there was a talk or two, and then some side conversations at that meetup, and I left being like, Wow, these people are really like they're building their whole workflow. around this tool, maybe there is something in it for me and I that I can, you know, go a little higher level. and I'm still I think I I kind of took this journey, but it was it was seeing how folks that are not writing code every day or having agents write code every day interact with AI that actually kinda opened my mind up to trying to use AI in a little bit of a different way, which I think was a benefit I wouldn't have gotten if I was just going to some software engineering meetup. Taylor MacDonald: That's interesting. yeah, so I had ⁓ I'm in this round table with a bunch of executives every month and this guy mentioned something really fascinating which I hadn't really considered. So like first of all, AI week kinda crept up on me. I mean I knew it was coming, but to be honest, I don't know how I missed the C F. Like I the window to speak at this conference. I mean, that's an initiative that I personally have at Ample this year. And so I'm I'm kinda disappointed in myself that I I wasn't more on top of my game, right? That said. Sean C Davis: You didn't want to do an impromptu panel that you didn't prepare for? Taylor MacDonald: Honestly, I you know what's funny about that, Sean? This podcast is kind of an example of that. Whenever I have to like prepare when I have prepared remarks and I need to go slide by slide, if I read the shit off the slide or if I like overindulge the preparation for that, man, I totally fucked that up. Or screw I need to be careful about my language. We're getting marked as explicit these days on on ⁓ the podcast. Yeah, so sorry, that's my fault. ⁓ I need to be ⁓ the the Sean C Davis: Yeah, yeah. Okay. Are we? ⁓ that's amazing. Taylor MacDonald: The more organic or the less prepared I am. Sometimes I think the more ⁓ c value I think I can bring to the conversation. That's it. Anyway, you said something hilarious that made me chuckle. Anybody can do anything. Isn't that the plot line of ratatouille? Did you ever say that maybe? About the rat the rat that can cook. We're all rats that can cook now. Sean C Davis: Mm-hmm. Is is that? ⁓ it's yeah, there's the ⁓ there's the ⁓ Goos Gousteau had written a book and that the title of the book. It's that's what it is. Anyone can cook. And we're c we're just cooking up code over here. Yeah. Taylor MacDonald: Yeah, any anyone can cook. I literally called anyone can cook. I it's true. Anyone can do it. ⁓ if you have the passion for the code, you can anyway. Yeah, so in this round table, I was talking to this ⁓ this guy mentioned or or what I was my preface was and I am all over the place today. My preface was that I was not prepared coming into AI week. I did not review the speaker list, I forgot to block off my calendar, and so on I think Monday night when I sat down to buy my ticket, I'm like, ⁓ crap. I've got meetings all over the place. I I'm not gonna be able to go and just completely engross myself in the conference experience. Which in a way is good because that can be a a fairly exhausting experience too, in and of itself, right? And most three day conferences I'll make it about two days in like an hour or two. And then I'm like, I'm done. I can't. I I'm not going to the last day. Well, so anyway, so I didn't really have much of a plan going into this. And so on Monday night I opened up the schedule and I started flipping through all the speakers and it was interesting. I started I recognized more of the speakers on the lineup than in previous po like I've never had that experience where it's like, ⁓ I know that person, ⁓ there's Sean, there's Alan, there's Ken you know, like ⁓ and so I thought that was interesting. And then I started debating, well, should I go see the for example, your panel? It was interesting. Sean C Davis: Mm, mm. Taylor MacDonald: probably wouldn't have attended that particular session if it weren't for you, but I I wanted to like hear the way that you know, I wanted to hear your words and support you and all that kind of stuff. but this guy on the round table mentioned to me, he was like, listen, you know, at this point in the game, like kind of when you are an expert or a senior or you know what the real question is what value you're gonna get out of the talk. So if you go and they're just talking about really rote level details like how to install clawed code and and how to use that, that's obviously not that valuable for you. And that in a lot of ways at these conferences, it's it's a lot like that. Each of these sessions, you know, you get a little bit here, a little bit there, but for the most part, it can be kind of a drag if you sit 30 minutes into a session and find like you get up at the end, you're like, well, that was a waste. And so he told me that he often will prioritize the speakers. So his goal was to seek out the or at least I'm I'm probably breathing a lot more into his statement than than what he actually meant, but that Sean C Davis: Mm-hmm. Taylor MacDonald: The goal would be to seek out the good motivating dynamic speaker, either the ones that you're already familiar with or ones that really like make you kind of think about it slightly differently. And so this experience for AI coming in unprepared and having limited windows of which I could attend, I basically just sat down and watched the conference or the sessions that were available to me. As opposed to like trying to get to the venue on time and do all the stuff and manage all that. I just sort of like, you know, rolled with it. And I had A more positive experience this year than I've had in past years. ⁓ particularly for this conference, but also at other conferences that I've attended over the years. Sean C Davis: Mm, okay. Yeah, you know the you know the people, you put less pressure on it. Yeah. I could I could see that. Taylor MacDonald: Well even yes, but even just sitting there when you're like, I don't there's I'm not that interested in this one, it's sometimes fascinating how it can push you into a different concept or a different way perspective or it kind of forces you to think from somebody else's vantage point. And when you're going and self-selecting, like, well I won't I really want to go to this talk for this reason, if the speaker isn't Sean C Davis: Yeah, yeah, yeah. Taylor MacDonald: Engaging, you know, it's it's kind of a waste, right? ⁓ whereas if you go in cold, it's like going to watch a movie, you know. I went to see the ⁓ dude, I saw this fascinating movie ⁓ last weekend with my son. It's called ⁓ No, no, no, no. ⁓ god, no. Wait, is that the one with ⁓ gosling on the moon or something? ⁓ no, this is called ⁓ backrooms. Have have you heard about us? Sean C Davis: You're not gonna talk to me about Project Hail Mary, are you? Yeah, yeah, yeah. ⁓ yeah, yeah. This this ⁓ this trailer looks amazing and I have a I have a coworker who said it was it's ⁓ she she loved it. She loved it. Taylor MacDonald: So Atticus has been all over me. I don't know how or where he found it, but he's really in this horror thing right now. Like him and Crimson have watched I don't know, like seven different movies that were all derivatives of the Conjuring. Which I didn't even know it was that kind of a franchise. He was like, ⁓ we're watching the conjuring. ⁓ we're watching the conjuring too. ⁓ we're watching this. We're watching this thing about this doll. And it's all just like, what? ⁓ anyway, him and Crimson have been watching all that stuff, and I have not been paying attention to it. I mean, I like a good horror. Sean C Davis: Okay. Yeah. Ha ha ha. Taylor MacDonald: What anyway, he's been bugging me to go see this movie. And my goodness, it was it was quite the mind f ⁓ trip. Sorry. It was ⁓ anyway, I I know I won't spoil anything for for you or any of the listeners, but it was a weird it's funny you mentioned the trailer because I guess this is some YouTube thing some guy dreamt up and for years, maybe on 4chan or some crap, he was putting out these videos. Sean C Davis: Yeah. Yeah. Taylor MacDonald: about this concept. I don't even know if any of that's true. Actually, I just know that it was on YouTube. And Attica's like, we need to watch this movie. I'm like, great. Like I really want to go watch some crap. You know, YouTube, you know, produced like production level quality junk. And ⁓ no, it was extremely well done. And going in cold, which I think was the original point of bringing up back rooms. Going in cold helped me Sean C Davis: Okay. Taylor MacDonald: Like i it freed me from the ⁓ constraints that I would have brought into the movie if I had a a preconception about what the heck a backroom was or what the plot line, you know, foretold. If that makes sense. Sean C Davis: Yeah, yeah, yeah, yeah. I know what you mean. Taylor MacDonald: Anyway. All right. well, ⁓ so the conference was just okay. are you going today? Sean C Davis: Conference was just okay. No. No. I've ⁓ I mean I've got a I've got a sick kid at home and I'll use that as my excuse, we'll say. ⁓ yeah, I got some catching up to do, and a lot has 'cause a lot has come up this week and ⁓ I was trying to think about even where to start with that beyond the conference. You know, have you Okay, I ⁓ there's there's a handful of things, but I think it should all probably start with. Taylor MacDonald: Yeah, that's a good excuse. I'm sorry to hear that. Sean C Davis: Have you played with this Claude Fable Five model? Taylor MacDonald: I haven't. only because I've been so busy that I haven't even been doing much other than like talking to other humans and AI is, you know, only so helpful in that. Although I will tell you that I used my my Pocket, which is this and I we're not being sponsored by them unfortunately, but I was no I'm endorsing this product but only because I love it, not because I'm getting paid to. But ⁓ there's this little physical device called Pocket. You've probably seen it. Sean C Davis: Mm. Yep. I just saw this somewhere y for the first time yesterday, I think. Taylor MacDonald: ⁓ it's awesome. So check it out. I'm actually so speaking of CFPs, I'm debating whether or not I mean I submitted a conference talk to Momentum, which is another Cincinnati like tech specific conference that's been around for a while. And one of my talks is a lightning talk on how ⁓ on how I have automated, you know, my my brain and my life through all these transcripting devices. And part of that talk is I'm trying to build my own. Sean C Davis: Uh-huh. Taylor MacDonald: version of the pocket. Check out. I'm in the middle, this is my breadboard, but I've got a little microphone and ⁓ this is a little Raspberry Pi right here. And this button is going to turn it on and turn it off. And so the goal is to emulate the functionality of the pocket myself, but then use the Whisper library directly on the Raspberry Pi so it never comes off. It never goes to the cloud. So I can basically transcribe the conversation and then Sean C Davis: ⁓ what? Mm-hmm. Taylor MacDonald: stitch it back into my you know, my Hermes and open brain style, you know, mind map. ⁓ but Sean C Davis: Now what's the benefit of having the the little device instead of just using your phone? Taylor MacDonald: So unfortunately I don't have it with me, but yeah, that's a great question. So ⁓ Kendra actually asked me that question. And I guess it's subjective. You know, if if you don't want something else to manage, then maybe the value proposition isn't as good. But what I like about it is that I can like ⁓ here's an example. So I'm on this board and sometimes I have to take Lachlan with me to the board meetings, and I want to record the meeting, but then Lachlan inevitably will be like, Can have your phone? Because I'm bored and I want to play Roblox or whatever. And so I can give her my phone and then this is like a separate device that literally can sit in the middle of the table and just record the conversation. That anyway, the the point of the pocket is that you record in person conversations. And when you're done, it will ⁓ take that MP3 file and ship it up to the cloud and that will transcribe it and then it will basically give you like a rag interface where you can query the conversation. Sean C Davis: Mm, mm. Okay. Taylor MacDonald: It will generate like a mind map, it'll generate a to-do list and summary output of everything that was discussed. ⁓ and it's great. It's amazing. And of course it's I think we talked about this last week. I mean, this has spurned a number of philosophical debates in my house about whether or not I'm destroying humanity or or this is the path to the destruction of humanity, which it may well be, but I don't think I'm responsible for that. I'm just along for the ride here, man. Anyway Sean C Davis: Yes. Yep. Yep. Yeah. Taylor MacDonald: I had the pocket with me all week at the conference and so I was able to just turn it on in the middle of the talk. And so I literally have like full summary output information of at least the sessions that I attended, which is really cool. ⁓ let me see if I can find one. I'll show you an example of what it looks like. ⁓ but yeah. Sean C Davis: So you're y you're using the pocket, you're not using your your custom thing yet. Taylor MacDonald: Not yet. My my breadboard is I'm still working on it, man. I don't know much about ⁓ electrical engineering. ⁓ but in this this is what I think is so amazing about AI. Like I am a tinkerer. You are a tinkerer. Like I know this about you. Like every time I talk to you, you've got like a new thing you're working on. And I'm the same way. Like whether I'm like seeding grass or replacing toilet or building a frickin', you know, AI companion device or Sean C Davis: Mm. Taylor MacDonald: Right now I'm in the middle of ⁓ one of my latest pet projects is ⁓ I want to build an iron forge. ⁓ like I've been talking about it for a decade and I I think I'm gonna finally do it. Like I realized it dawned on me. I finally have all the materials. Like, okay, so I'm getting old. I'm like bordering on my fifties here in a few years and ⁓ Sean C Davis: Yes, you've been talking about that for years. Yes. Taylor MacDonald: There is this moment that I remember back in like my twenties or my early thirties where I hit critical mass on my pantry. You know what I mean? Like I had all the spices and I had all the things and it's like in that moment you're like fuck or crap. I don't know what I'm gonna make for sorry, sorry. I I don't know what I'm gonna make for dinner tonight, but then you can open the pantry and be like, ⁓ well, there's like twelve pathways to conclusion here. And it's because I hit critical mass. Well, I finally hit critical mass on tools. And crap in my basement that could be repurposed to solve a problem. And it's a weird it's a weird place to be. So I've I realized the other day that I've got all the materials, including like the propane ⁓ apparatus and the bricks and like I can literally build my own ironforge this afternoon. But I don't want it to be super janky, so I'm gonna like plan it and do all that stuff. But anyway, the fact that you can do that shit, that you can Sean C Davis: ⁓ yes. Okay. Taylor MacDonald: stuff that you can wake up on a Sunday and say, All right, I'm gonna go build an Iron Forge, but we know where to begin. And you've got YouTube at your disposal and then you can open up AI and AI can literally like craft the output or the solution based on the tools that are available to you or the spices that are in your pantry. It's freaking amazing, man. ⁓ Sean C Davis: Yep, yep. I know and I'm bringing it to the I'm building all these tool digital tools, but I'm I love bringing it into the physical world as well. It's like I just walked around the yard, took in in the winter and took a bunch of photos and I was like, Okay, here's the layout of my yard. This is the year I'm gonna transform it. And I have been working on the landscaping in this house for twelve years and it is like it is it It tw it looks twice as good as it did any other year. Cause it was finally like, all right, we're gonna do it. Yeah. Yeah, yeah, exactly. So I think that's that's really cool. And something I wanted to ask you about was the okay. what is what what still comes up a lot and actually what I was going to if I was gonna if I was gonna actually be able to give a talk at ⁓ since the AI week was gonna talk a lot about Taylor MacDonald: Finally pa paid off. You hit pay dirt, man. Literally. Sean C Davis: the build versus buy proposition and and how I feel like it has totally shifted within the context of both personal tools and internal apps for Teams. ⁓ and I think there is still an inflection point in some cases, but I I am very regularly now, if I hit, if I start to stumble or hit some friction, I think, is it worth it to Is it worth it to try to work around this in this other tool? And or how hard is it going to be to rebuild this? Like it's something with ⁓ Excalibur. You know they got an MCP server and it's amazing. You also have to pay a bunch of money if you want to use it and use it with the team. And it's like, it's just Canvas rendering SVGs. How hard is that to re can I rebuild that in a week? So I'm curious from your perspective, you know, where's that? Taylor MacDonald: Yeah. Sean C Davis: Where's that inflection point for you? And I'm I know your time's really limited and and you don't have as much coding time, but ⁓ I don't know. I has it changed how you're approaching that? Taylor MacDonald: Absolutely, man. I mean look at we've we've been talking about Contentful's acquisition. We we have we have a client right now that sought us out specifically because they can't afford the rising cost of the vendor lock in imposed on them by frankly Contentful. it's a decision, right, for Contentful to go that route. But yeah, yeah. So anyway, so yes, I think about it a lot. You know, five years ago If you're like, let's build a CMS, I'm like, you're an idiot. We did that in 2010. Like that was a terrible idea. I don't know if I feel that way anymore. And and you know, you might argue that, well, that's, you know, reckless or risky or whatever. And the truth is, when you're talking about a piece of software that has to breathe and live and evolve over the next ten years, it's supporting an enterprise. It's like, you know, got a ton of different concerns and a ton of different feature requirements. Sean C Davis: Yeah. Yep. Taylor MacDonald: The what you're paying for is the stability. You're paying for a team that knows that software inside and out and you're paying for the promise that it's gonna stay online and it's gonna continue to evolve and continue to be updated. If you could build a piece of software today for a f twenty bucks of tokens or whatever, and it did effectively what you wanted it, like it solved a problem for you, I totally think there's an argument to be like, this is a better value proposition than it is to buy the software because you could just delete that code base and resurrect a different version tomorrow for twenty bucks. Sean C Davis: Yeah. Yeah. Taylor MacDonald: If that makes sense. I it it's I know that sounds reckless, right? To be clear, I would heavily caveat this statement to anybody I was talking to, but if you're marginally savvy, why would you you have ultimate control? Like now telling Claude, say, hey, I want you to fan out a swarm of agents, and I want you to effectively recreate Contentful's interface and backing. There's some complexity there, right? And I mean they have a really solid, amazing product. And and I'm not saying that through this lens I would say, well, we're building, you know, CMA tokens and C D A tokens and yeah, like a full on API and and ⁓ webhook support and all this crap, although you could. but there's a lot of concerns that they bring to the table. You know, single sign on I mean if your if your business requires this level of fidelity and support and reduced risk, then it still makes sense to pay for that. But if you're a small business If you're the construction company, I mean this is the analogy we used, you know, a month ago. If you're the construction company that is generating fifty to sixty million dollars worth of revenue and yet you're still operating your project management out of a s a spreadsheet, and it's just the way you've always done that, well, you don't have to do that anymore. You know, and it's not gonna cost you a million dollars next year in order to solve that problem. Will it be the perfect thing, the most polished thing? No. But that's not the point. The point is a ⁓ you know. Sean C Davis: Yeah. Yep. Taylor MacDonald: You've you've automated the the process that was holding you back. You freed up resource hours and labor to go focus on a different problem with the organization. And you didn't do it with this storm cloud of vendor lock in hanging over you so that, you know, the minute that company decides to sell out to Salesforce, you're locked in, baby. And they're you know, it's gonna cost you more to get out than to you know, than to to stay. I I don't know. Does that answer the question? I I think we're in a throwaway society. We've been primed for this, man. How many printers have you pitched because one little rubber gasket crapped out or a print head dried up or something like that? We just throw it away and we go buy a new one. If you apply that same evolutionary concept to technology, we're in that same space. I'm not necessarily saying it's a good thing, but it's an economical thing if you think about it, assuming you can control your risks and your stakeholders are comfortable. Sean C Davis: ⁓ man. Yeah. Taylor MacDonald: That approach. Sean C Davis: Yeah, but the the th the okay, but a printer has a job that I think generally everyone can ⁓ who needs to print something can agree on, right? I mean there's different versions of a printer, but you're like, all I'm gonna need is black and white printing, so I'm gonna pick this printer and it perfectly solves my problem. But with software, especially SAS Like this is I need to draw this out because what I'm thinking of is like if you could visualize the the problem space for that that you know, let let's even it doesn't have to be a big number, it could be like a dozen different organizations have around content management. You graph this out, you kind of end up with this Venn diagram, right? And and then what a company like Contentful does is they say, okay, cool. I can serve all of you with this overlap in the middle here. And this is the solution I'm going to bring. Or in some cases, you're like, I'm going to cover all of this. And in either case, right, you if if you are one of those customers, you're like, I'm down here in the corner. And here's my problem set and here's what I need to do. Contentful is going to offer one, probably way more stuff than I actually need. And it's going to be built in a little bit more of a generic way. And so even even with your example, if you're like, yeah, go rebuild Contentful, you start talking to that agent and you're you and and you put your product manager hat on, you're gonna end up building something completely different because you're not actually trying to recreate contentful. You're trying to solve the problem that you have, and that's just your inspiration for getting there. Taylor MacDonald: Yeah, ⁓ completely agree. I I mean that's that's the so on this board. I ⁓ it's a nonprofit organization, the Forest Conservancy here in Fort Thomas. And obviously the challenge with any nonprofit is you gotta figure out how to raise money, whether through membership or events or or grants. And so particularly in the case of the Forest Conservancy, the board was interested in surfacing Grants. And ⁓ you know, there's a million different ways to think about that. And one of the board members went out and just I think he like vibe coded like an HTML, like he had Claude do the research and then return, you know, the data of that research as an HTML file, and he sent it to me. And it was all I mean, it was great. It was a good source of information, there was good information there. But that is obviously now you're constrained. That's not scalable, you know, that that r would require Brad to constantly regenerate this file every day. separately I was working on solving the problem too and I ended up ⁓ going a little bit farther. I built like a platform that uses like an NADN workflow to just constantly consume, you know, feeds and then you can add the sources. And then I built like a little Kanban swim lanes, you know, to keep track of the progress of different, you know, applications that would have been filed for a grant ⁓ or proposals that would have been submitted for the grants. ⁓ And the I the thought that like both of those are are valuable a avenues and how do we go from here? Like how do I take Brad's stuff and mix it into my stuff and all that kind of stuff. I mean that that speaks to that customization, right? Like if we were forced to live like let's say we had adopted a sauna to do this, ⁓ just using that as a rough example. It's a swim lane, you could have a card for every grant. It still requires a whole lot of glue around there to get the same output. Right. So in a in a few days worth of, you know, iterative like working through this, I now have a full end application. It will authenticate against our Google Workspace. You can now only every everybody on the board has access, but nobody else does. They can get in, they can see all the different grants, they can request new sources, they can do all this stuff. And it's fully customizable for the board. The board it didn't cost the our nonprofit a penny. And now they have this whole new apparatus. Sure. Sean C Davis: Mm-hmm. Mm-hmm. Taylor MacDonald: Will it crap out? Does it need maintenance? Are there ⁓ potentially like vulnerabilities that underpin some of the dependencies there? Yeah. If not today, there will be in a week or a month or a year. But I don't know I don't know that I care. Like now, I'm saying this in a position of privilege. It doesn't matter. There's no risk to this board if that application went offline. It's just like, ⁓ well, net loss, one penny. I don't there was no upfront investment. Sean C Davis: Yeah, right. Taylor MacDonald: I it would be like an expectations problem, but it wouldn't it wouldn't result in a loss of revenue or anything like that. And then meanwhile, the board now has this fancy new tool that can theoretically further in in a significant way further the financial interest of the organization for the betterment of our mission and values. I don't see how you can argue with that. Like you know? There's no amount of free Salesforce you know, licenses that can necessarily replicate that same level of power. Sean C Davis: Yeah. Yeah. Yeah. I was thinking about the this in context with ⁓ Okay, so I'm spinning up all of these different projects and it what I would do back in the way my brain worked when I was writing code manually would be like, man, keep solving this problem in the same way on every project. And then you have to decide Am I building some starting point or am I building some abstraction layer? And the abstraction can be really challenging to maintain, distribute out, ⁓ updates and things like that. And one of the okay, so one one example. They end up being developer tools, right? One of the examples is ⁓ w we'll call it logging. Logging or, you know, event logging, maybe. is then you wanna You want to capture events from users. You might want to push notifications of specific events to Slack or whatever. And I'm like, I have a slightly different need, but I end up like telling the agent to do essentially the same thing in every one of these tools. Maybe I should just build my own developer tool. And it seems silly at first because you're like, I can ⁓ I don't know what would a tool, maybe observable might be what ⁓ tool that you would go get that would ha that would that would have this functionality. But then you're I mean you end up paying a lot for that versus what do I actually need to pay for? I want to pay for the data itself and then I'm in control of it. I can ditch it when I need to ditch it. And if I need new features, it's not like I've gotta go touch every I gotta pull up every repo open pull request. It's like that has to happen. I don't have to do that. I can just say, agent, go do a go update. Taylor MacDonald: You know, I'm struck. Like you you just kinda it's all about self-sufficiency, right? Like part of the problem we have is that in order to implement a highly complicated product like Contemple, you need an engineer to do that. That costs a lot of money, a lot of years of training to do it effectively or whatever. One of the coolest things about AI, and I'm str and the what I'm struck by is the core like okay, so I I mentioned this iron forge that I want to build. And I know that sounds silly, but I've always been interested in like Sean C Davis: Mm-hmm. Taylor MacDonald: I took this class on what I thought was going to be on welding a long time ago because I really wanted to learn like the you know, wanted to learn how to lay a stack of dimes, as they say. And turned out that class was like seventy five percent blacksmithing. And so I got ⁓ it was it was kinda cool. It was another example. Like I walked in there not knowing what I was going to take away, and I actually find blacksmithing to be so much more compelling in the long run. Sean C Davis: Mm-hmm. Taylor MacDonald: And so one of the cool things about blacksmithing is that everything that you work with, everything that you can do in blacksmithing is in some ways ⁓ self-supportive of the the task. So like an example, hammers, like a like a cross pin hammer, which is basically like a big old chunk of st of you know carbon steel or whatever. ⁓ Sean C Davis: Mm-hmm. Taylor MacDonald: has been hammered down in a way like it's it's heated up in a in a in a forge and then you hammer it down and you put the angles on it and then you drive a hole through it and you spread that out and all that crap and then that becomes a hammerhead. And so many tools that you use in blacksmithing can be forged by a blacksmith. So it's this kind of like self-sufficient another example. So if you have a piece of rebar or whatever you're ⁓ melting down in your forge That's weird. I just got a phone call from a company called Up Forge. Look at that. That is bananas. That's weird. ⁓ my God. That's terrifying. anyway, everything that you all your tools, like if you're gonna put a piece of of a steel inside your iron forge there and like heat it up really good, you obviously can't stick your hand in there and pull it out. So you need a pair of tongs, right? Well, you can go up to the store and buy a fancy Sean C Davis: ⁓ someone's listening to you. Taylor MacDonald: pair of tongs that was manufactured for you. Or you can literally make your own tongs. Like there are all sorts of YouTube videos about how you take, you know, your whatever your different type of material that you're working with, and here's how you kind of hammer it down. Here's how you angle this piece and this piece and you put together and then you drive a rivet through. And it's how the world used to work. It was this measure of self-sufficiency, right? Before everything was manufacturable, we It's just so interesting, right? Like so many things in this life that I'm like, well how do you make that? Like how do you mold rubber? How do you, you know, fashion this project? Like we have things like 3D printers today, we have AI, we've got, you know, all the rote level material opportunities, like you know, if you want to be a blacksmith and start manipulating metal, you can be a woodworker. It just there's i it it's crazy how self sufficient we could theoretically be if we just leaned into that. And in a way, AI takes some of that from us, but it also gives us ⁓ it's like a path to that self sufficiency. That's a very optimistic take. You know, if my wife was in the room, she'd be really challenging me on that opinion. And rightly so. But I don't know. What what do you agree? I think this has so much to do with Sean C Davis: Yeah. ⁓ Taylor MacDonald: Like the same question of being able to use the AI to generate your tools that solves your specific problem and then if it breaks, you can just recreate it. That is the exact same premise of what it means to to forge iron into different shapes and tools. Sean C Davis: Yeah, yeah. I I had this idea about a year ago or so and I was like I'm so I'm a I'm a big plant nerd and I've ⁓ forgot the garden outside, but I've got really into houseplants during the pandemic. And what I wanna do is I I basically wanna get to a point where I don't have to buy anything other than raw materials and and maybe soil to be able to maintain my plants and and maybe even sell them someday. And so I wanna learn how to make my own planters. And my wife is a great potter, but she also thinks making planters is boring. So I had this thought of well I've always wanna we don't I have I have thought about doing that, but ⁓ we're not there yet. We're not there yet. Taylor MacDonald: Do you own a kiln? Do you guys? You should build one. You should build one. Sean C Davis: We should build one. I I okay. S tangent, I have this theory that it is possible to build a device. This is a ridiculous theory, okay? It is possible to build a device that can that can be a campfire stove and a pottery kiln and a pizza oven. And I wanna invent that thing. Taylor MacDonald: Like a all in one, like three purpose device that does all of those things. Sean C Davis: Yeah, I mean I think there would probably be some moving of bricks around to go from one mode to the other. But but yeah. Yeah. A anyways. Taylor MacDonald: Okay. Yeah. Dude, I wanna build a I wanna build a ⁓ tandoury oven in my backyard. Like I wanna dig a big old hole and build like a clay oven. How weird would that be if I'm out in the yard on my knees like slapping tan like non bread on the inside of this subterranean oven. Sean C Davis: Yes, that would be amazing too. Yeah. I mean just make it Just non non all day every day. Yeah, I love that idea. ⁓ anyways, I I had wanted to get into woodworking for a while and I just I don't it was hard to find the right class and ⁓ so I finally I was like, you know what? I'm just going to do it. I'm just gonna get started. Just start it. It doesn't matter, you don't have to even make something like just just get started. So I went a couple of weeks ago and I took a ⁓ Taylor MacDonald: Anyway, I interrupted, sorry. Got excited about non. Sean C Davis: I took a safety course at some local woodworking shop, but it's same sort of thing is is like, Okay, now I've got step one and what I can do is the next time I need a planter, I can be like, Snap photo, here's the space this planter is going to go. Tell me what ⁓ Claude, tell me what kind of board I should buy, what tools I'm gonna need to work with, and then I'll rent some time at the shop and It'll suck, but you know, it'll get better over time. So yes, it resonates. Yes. Taylor MacDonald: I know. It's liberate it's a liberating time to be alive for some things. ⁓ Sean C Davis: Yes, yes. I mean it does in in a way it's like okay, here's a weird ⁓ I've been I've been trying to think a lot about all the ways that I am hypocritical lately. And one of the one of the things that I've been thinking about is ⁓ okay, so we're we're essentially destroying the earth with what we do all day, every day. But in a way, it's also taking me Taylor MacDonald: Do tell. Sean C Davis: closer to the earth and like making me appreciate all the things that come from it in the same way and trying to be more sustainable with the other with with all of what I consume. So like yeah, I'm consum I'm burning a bunch of water because I'm consuming all these AI tokens, but are they helping me do things where I then consume less of other things and become more self sufficient? Does it offset? I don't know. I don't know that it's possible to actually do that math. But ⁓ it feels good. Taylor MacDonald: So weird, man. I I don't know. Humans as a species, they just consume and destroy and lay in waste. Like that sounds so bleak and discouraging, but I it is true. And it it's this shame. Like I I don't I hate that ⁓ I mean I really do f I feel uncomfortable and hypocritical about many, many things. This one is definitely one of ⁓ And I don't know how to square it. Sean C Davis: Yeah, but it's true. Taylor MacDonald: It's part of the value of these community groups. Like the Centi AI for humans group that we talked about earlier, they do like responsibility and ethics and AI. I think a lot of that has more to do with like PII and ⁓ bias and inferring ⁓ like an inference bias and all sorts of stuff like that. And I I appreciate that they're talking about it. I and I'm sure they're also talking about the environmental Impact, I don't know how to square that circle. I I I don't feel I feel powerless as an individual. Granted, that's a very self-serving statement. Like I run a company, I have influence over the outputs of that company, over the choices of ⁓ you know, the decisions that we make along the way. And I know that a lot of other people don't have that same privilege. And I I I don't want to lose sight of that, but in the big picture. I don't know how I mean we recycle dutifully every week. Like we I don't unfortunately drive an electric car because for reasons. I need to upgrade my my electrical box for one. ⁓ but for two, like I I would like to do that. I would like to install geothermal. I would love it if my entire roof was covered in solar panels and I'd even put a dang windmill in the backyard. I don't think they look that ugly. I know everybody hates ⁓ but I think it's a fascinating technology. It's just a it's it's in it feels insurmountable and existential when you read these posts about the end of the you know, the climate crisis and stuff and I ⁓ I don't know how to approach it. Like I don't know what to do, you know? I feel really conflicted on this. ⁓ but I also recognize a fiduciary responsibility to to figure it out, to help my clients figure it out. I mean Sean C Davis: Mm, mm-hmm. Taylor MacDonald: We're capitalists at the end of the day, for better or for worse, and I have to pay the bills. So I don't know. It's ⁓ complicated emotions on that one. Sean C Davis: It is true. Yeah. I I agree, 'cause we yeah, we live in this we have to exist in this capitalist society, so we gotta figure out how to do that in ideally a sustainable way. ⁓ and we probably so we we don't have enough time probably to get into it today, but I I think let's ⁓ let's use that as a jumping off point for next week because I wanna talk about the ⁓ Leveraging, you know, we we have we have touched on token consumption here and there and different models that you can use for different strategies and and cost management. And I think ⁓ starting to see a trend in how some of that pricing is changing and how that might affect behavior. And I think that could be something to go really deep on next week, but it's gonna take more than five minutes. Taylor MacDonald: Yeah. You mentioned Fable and this has been a I really have appreciated this conversation when we're getting close on time. You've almost talked n like nothing about actual tools or ⁓ demos or or the work that you've been working on. But what in the last couple of minutes here, tell me what your experience with Fable is. Like do you feel good or bad or have you kicked the tires or I don't I I'm conflicted on this one too. Sean C Davis: Yeah. Yeah, it's I so I started to tinker actually I d I just started playing with it this morning and Taylor MacDonald: ⁓ and we should mention Fable is the new model. Sorry, if you're listening and you don't know, Fable is the new model that was recently released by Anthropic, and it is kind of a a sort of stripped down, security mindful, light version of Mythos, which is the really dangerous, scary model that apparently all you have to do is to have like a relationship with Anthropic to get access to at a corporate level. Go ahead. Sean C Davis: ⁓ yes. Yeah, so what did I do? I got this I have this skill that runs and it we go through this specific process where we we use linear for project management at ⁓ or and task management at Netlify. And so I'm I generally interact with linear through Claude. And this morning I was like, I gotta update a couple of my projects and th the the process is the same every time. It's like Okay, cool. This is what I found. Here's my suggestion. Do you want me to post this? And this morning I put Fable on it. And it was like, boom, company wide project update, overview. And I and and within within 30 seconds, there was a VP who was like, I think we should be going a different direction with this project. And I was like, Yeah, I know. That was not my intention to publish that. It's fine. I'm gonna edit it. I got my emoji of my my old ⁓ grandpa Simpson yelling at the the Claude logo. That's one of our most used ones. Taylor MacDonald: Hilarious. You you need to share that with us so I can cop right click that and save it into our slack. Sean C Davis: ⁓ man, I'll see you. I think we've got a Figma file somewhere where it's basically just Grandpa Simpson copied and ⁓ there's like a hundred different logos, so you can always just be yelling at some service that you hate. Taylor MacDonald: That is so funny. ⁓ All right. Well, ⁓ not a great first you know, inaugural voyage on there, it sounds like. Sean C Davis: Yeah, I'm gonna tinker with it. I will say that okay, so here's my cliffhanger for next week. I am I'm concerned of the the posturing of here's this thing that is really, really powerful and it's gonna change your life and it costs twice as much and here it is for free for a little bit. It's like we're d we're what are we selling cocaine? I you know I Taylor MacDonald: Yeah. So this is an extremely cynical take, but I so this guy who I know from a job, you know, we work together, he's a contractor at another company. I think he was speaking it ⁓ since the a ⁓ week, but I saw this this morning on on LinkedIn and I completely agree with him for what it's worth. ⁓ hold on one sec. I'm I'm looking for it. He hit the statement was they want you to s they want you to loop and stop thinking and then they can pull the plug and charge what they really want. So this is this is what dustin Coker wrote that and it came across ⁓ just a like a LinkedIn notification or something. So it must have been a c I haven't even clicked through. But like yeah, that's an interesting thought, man. That's awfully cynical. And he's probably not wrong. Sean C Davis: Yeah. Mm-hmm. Taylor MacDonald: That's the problem with capitalism, man, is that there's always some sort of motive to pull the wool over your eyes because profit is more important necessarily than user experience, or more necessarily than, you know, ethics or, you know, your values. And and again, cynical. And I know that not all investors feel that way, and not all companies operate that way, but boy, it sure does feel that way sometimes. And that is scary. That might be scarier to me than the, you know, than well, as scary as any of these other things. That like if we know there's a fatigue that takes place when you adopt these tools and it's an atrophy of your ability to critically think, in some ways that's good because it gives us the opportunity to build stronger relationships. But in other ways, that's absolutely horrifying because. Sean C Davis: Mm-hmm. Yeah, yeah. Taylor MacDonald: You know, once you relinquish that general awareness and you seed the behaviors to the calc like you if you seed your ability to do math to a calculator, what happens when that calculator, you know, jacks up its subscription rate tomorrow? Like, what's gonna happen? What's gonna happen to us? That's that's pretty scary. Sean C Davis: So I I had to buy I had to buy a dryer this week. Fun stuff, right? All right. I'm like, I might be moving soon. I don't care what dry I want the cheapest dryer from Costco that works. So I sign into Costco, I I search electric dryer, and then I sort the page price low to high. ⁓ I I I need to remember my numbers. I think it was Fourteen of the first eighteen results. Fourteen of the first eighteen results were sponsored and not ordered by price, which means I had to scrape through f through eighteen of the top results to find actually one of the four least expensive drivers. And I'm like, at some point, like this is pure inchidification for sure, but like at some point we go from This is a bad experience too. Like, is this actually a bug? Right? Is this a is this a software bug? You're you're not ordering the page from low to high. No, I mean I mean I just mean like, you know, from a from a product standpoint. It's obviously an intentional choice, but from a user standpoint. Yeah, yeah. Yes, yes. Taylor MacDonald: Doubt. I doubt it. That's the sad reality. Right. I was saying I doubt it's a bug. It's very intentional. Which is is is you know, and the worst part is like the first time that happens, the first time you run into that very real reality, you can dismiss it and wave it off. It's like, well, that's just a bug. That's just a mistake. That's just that one company or whatever. It's freaking everywhere. It's literally all corporations. It's every product. Like if you stopped and looked around at your room, you're like, okay, that one, that one, that one, that one, that one, that one. These are all Sean C Davis: Yeah, it's everywhere. Taylor MacDonald: consumer packaged goods, they're all, you know, consumer level products. We paid money for them. All these corporations exist primarily to put more of those products on your desk. And that's about it. Like that's the value proposition so that the halves can enjoy their yachts and the rest of the world can can basically sit and worry about the impact that they're having on the environment. Cause I'm pretty sure the guy who's done the frickin' Sean C Davis: Mm, mm. Taylor MacDonald: the the yacht could give two shits about the environment. That's I'm painting with a broad brush. I'm sure that's not exactly the case. I know King Charles is a very ⁓ environmentally meant ⁓ environmentally minded human being. So I don't know. I or gilded palaces with frickin', you know, expensive teacups, fancy outfits Sean C Davis: Who hangs out on yachts all the time? Yeah, we got we got we gotta cut it off or we're gonna end up in ⁓ rabbit hole of like empiricism and stuff like that. Yeah. Imperi imperialism. Taylor MacDonald: It's true. All right, Sean. Thank you. I have enjoyed this conversation today. ⁓ next week let's take another look at I know you've made some progress on your ⁓ prototype. But we'll get back to natural language soon. ⁓ but thanks. Thanks for the ⁓ Sean C Davis: Yes. Yeah. Yeah. And I I should give another teaser to folks that ⁓ we didn't cover it today because you and I had a really good chat on Tuesday afternoon. and kind of really like dug in and I think rethought the whole system. And so it's it might be a rewrite from the ground up, but we know those things can go fast. So my intention is that we're looking at kind of something sort of brand new next Thursday. Taylor MacDonald: Next week. Okay. Well let's ⁓ next week we'll talk about some of those architectural concerns and some of the the ideas that we kind of that bubbled out of that conversation that we had. And then yeah, maybe you can show us show us where you are now. Awesome. All right, thanks, man. Really appreciate you. I hope you have a great week and ⁓ we'll see you next time. Peace. Sean C Davis: Love it. All right. Great. See ya.